1
00:00:00,000 --> 00:00:07,320
Welcome to the Canvas Data Science Meetup Group.

2
00:00:07,320 --> 00:00:09,000
Thank you all for attending.

3
00:00:09,000 --> 00:00:10,800
It's your eyes, your ears,

4
00:00:10,800 --> 00:00:13,720
your attention that's really moving things forward.

5
00:00:13,720 --> 00:00:15,620
Couldn't do it without you.

6
00:00:15,620 --> 00:00:19,220
We've made a lot of progress and I want to really

7
00:00:19,220 --> 00:00:23,200
start formalizing some of the data we've been collecting,

8
00:00:23,200 --> 00:00:25,440
making sure it's accessible to you,

9
00:00:25,440 --> 00:00:27,960
and then let's start crunching it.

10
00:00:27,960 --> 00:00:30,600
Because as I've been alluding to,

11
00:00:30,600 --> 00:00:35,040
there's so many good statistics that can be calculated.

12
00:00:35,040 --> 00:00:41,120
I feel each statistic has a myriad of insights

13
00:00:41,120 --> 00:00:45,600
that you can think about and glean from.

14
00:00:45,600 --> 00:00:49,160
How do you actually value those insights?

15
00:00:49,160 --> 00:00:55,440
Hard to say, but one could argue that knowledge is power,

16
00:00:55,440 --> 00:00:57,560
and what's knowledge,

17
00:00:57,560 --> 00:00:59,920
but maybe knowing about the world,

18
00:00:59,920 --> 00:01:01,420
how do we know about the world?

19
00:01:01,420 --> 00:01:04,760
One could argue we could use statistics.

20
00:01:04,760 --> 00:01:07,520
Potentially, if you follow the logic back,

21
00:01:07,520 --> 00:01:09,760
statistics is power.

22
00:01:09,760 --> 00:01:14,200
What can we do to help out in the cannabis industry?

23
00:01:14,200 --> 00:01:18,840
Can calculate statistics to our heart's content.

24
00:01:18,840 --> 00:01:23,320
Just doing that small, quirky,

25
00:01:23,320 --> 00:01:30,680
sometimes tedious work, even if it seems irrelevant,

26
00:01:30,680 --> 00:01:34,680
I argue is really moving things forward.

27
00:01:34,680 --> 00:01:37,240
Those are my thoughts for today,

28
00:01:37,240 --> 00:01:39,040
but it's a meetup after all,

29
00:01:39,040 --> 00:01:41,760
and I'm not here to steal the stage.

30
00:01:41,760 --> 00:01:45,520
If any of you ever have any good ideas, thoughts, comments,

31
00:01:45,520 --> 00:01:48,440
questions, you're always welcome to put them forth.

32
00:01:48,440 --> 00:01:50,820
Because for example, last week,

33
00:01:50,820 --> 00:01:54,480
Robert had this phenomenal question about lab tests

34
00:01:54,480 --> 00:01:56,320
that really got us thinking.

35
00:01:56,320 --> 00:01:59,560
Then Ruth did an excellent job

36
00:01:59,560 --> 00:02:03,840
about explaining the current state of lab testing,

37
00:02:03,840 --> 00:02:07,120
some of the issues, why they matter,

38
00:02:07,120 --> 00:02:12,440
and we even then began just spitballing solutions.

39
00:02:13,880 --> 00:02:18,160
I wanted to pick up a little bit in that vein this week,

40
00:02:18,160 --> 00:02:20,360
as well as share with some data with you.

41
00:02:20,360 --> 00:02:27,360
But Candice, I know your rig may get put to some good use here

42
00:02:27,360 --> 00:02:30,120
in the short to near future,

43
00:02:30,120 --> 00:02:32,920
as start making some of these algorithms

44
00:02:32,920 --> 00:02:36,280
more generally able to run.

45
00:02:36,280 --> 00:02:40,800
But how are you doing on your work?

46
00:02:43,320 --> 00:02:46,840
I'm doing great. Thank you, Keegan. Hi, everybody.

47
00:02:46,840 --> 00:02:48,840
Yeah, Keegan, that sounds great.

48
00:02:48,840 --> 00:02:53,840
Also too, I could, if I'm going to be filling up my 12 terabyte,

49
00:02:53,840 --> 00:02:56,320
once we get all the data in one place,

50
00:02:56,320 --> 00:03:00,400
I can easily dupe it and then send you the exact,

51
00:03:00,400 --> 00:03:03,440
send you an external hard drive

52
00:03:03,440 --> 00:03:06,640
that you can just attach to your laptop too.

53
00:03:06,640 --> 00:03:09,520
It's so funny you mentioned this.

54
00:03:09,520 --> 00:03:14,120
I should have been keeping my ear a little closer to the ground.

55
00:03:14,120 --> 00:03:19,200
But yesterday, Meta just released Llama 2.

56
00:03:19,200 --> 00:03:20,480
Right.

57
00:03:20,480 --> 00:03:24,000
I've got a good feeling about this.

58
00:03:24,000 --> 00:03:27,320
Because Meta has been in the news lately.

59
00:03:27,320 --> 00:03:30,040
They launched their app,

60
00:03:30,040 --> 00:03:34,720
threads that got to 100 million users in a matter of days.

61
00:03:34,720 --> 00:03:38,800
They're coming out hot.

62
00:03:39,640 --> 00:03:43,000
It's the clash of the Titans.

63
00:03:43,000 --> 00:03:50,480
It's Microsoft has stolen the front stage with OpenAI.

64
00:03:50,480 --> 00:03:55,880
You'd think Meta is probably going to come back strong.

65
00:03:55,880 --> 00:04:02,520
Just from just thinking about the economic incentives,

66
00:04:02,520 --> 00:04:08,480
I could imagine that they may have put a lot of time and effort into Llama 2.

67
00:04:08,480 --> 00:04:19,320
From what I can read, it's licensed under a fairly open license.

68
00:04:21,320 --> 00:04:25,240
I think they've got a custom license, but you can look at it.

69
00:04:25,240 --> 00:04:29,680
I don't know how cannabis friendly they're going to be.

70
00:04:29,680 --> 00:04:33,520
That's how cannabis is in general.

71
00:04:33,520 --> 00:04:35,920
It's such a gray area.

72
00:04:35,920 --> 00:04:40,200
They don't really say for certain in the license.

73
00:04:40,200 --> 00:04:49,760
Because it's basically, you have to do things that are legal.

74
00:04:49,760 --> 00:04:54,360
Then it's, is cannabis operation, are they legal?

75
00:04:54,360 --> 00:04:56,600
The states say yes.

76
00:04:56,600 --> 00:04:59,960
The federal government doesn't really want to,

77
00:04:59,960 --> 00:05:02,400
I guess, technically say no,

78
00:05:02,400 --> 00:05:06,560
but they're not interested in getting involved.

79
00:05:08,120 --> 00:05:13,600
Actually speaking of which, it's the later cake government that we have.

80
00:05:17,360 --> 00:05:20,200
It's really cool Keegan,

81
00:05:20,200 --> 00:05:24,280
because I love it that Meta, Mark Zuckerberg,

82
00:05:24,280 --> 00:05:29,880
he is just rolling over to us this new model.

83
00:05:29,880 --> 00:05:33,400
We don't need an internet connection to run it either.

84
00:05:33,400 --> 00:05:36,600
It just plugs in right into Lang chain.

85
00:05:36,600 --> 00:05:41,640
We should be able to pull in all the external cannabis data,

86
00:05:41,640 --> 00:05:50,040
the CSPs, the XLSs, the YouTube videos, Slack.

87
00:05:50,040 --> 00:05:55,760
All of your data can be put into private GPT.

88
00:05:55,760 --> 00:05:58,560
I'm really excited about this.

89
00:05:58,560 --> 00:06:02,040
I think it's really great that there's so much competitive at

90
00:06:02,040 --> 00:06:06,440
the billionaire level that they're giving us free access.

91
00:06:06,440 --> 00:06:08,200
It's not like an API or anything.

92
00:06:08,200 --> 00:06:12,080
I mean, you just download it and whatever you're doing with it.

93
00:06:12,080 --> 00:06:15,320
I'm sure that if it were getting published or something, it would be different.

94
00:06:15,320 --> 00:06:18,240
But we'll stay within the letter of the law.

95
00:06:18,240 --> 00:06:21,680
But we're just talking about cannabis software.

96
00:06:21,680 --> 00:06:24,640
It's not like we're selling cannabis.

97
00:06:24,640 --> 00:06:30,520
I actually thought about this long and hard the other day.

98
00:06:30,520 --> 00:06:36,760
I think what we're doing is incredibly valuable because whether you use cannabis or not,

99
00:06:36,760 --> 00:06:42,520
you generally want to know what are in these products that people are consuming.

100
00:06:42,520 --> 00:06:45,720
If you're a medical practitioner,

101
00:06:45,720 --> 00:06:54,560
it may behoove you to know what's actually in this cannabis that people may be consuming.

102
00:06:54,560 --> 00:07:00,360
Or there's endless reasons.

103
00:07:00,360 --> 00:07:04,000
In general, like we were saying, knowledge is power.

104
00:07:04,000 --> 00:07:09,520
I generally think it's good at least for the information to be available,

105
00:07:09,520 --> 00:07:11,880
whether you're partaking or not,

106
00:07:12,880 --> 00:07:16,200
what you're partaking or not.

107
00:07:16,200 --> 00:07:19,960
That's good. Oh, yes. Then the other point you raised,

108
00:07:19,960 --> 00:07:22,640
because I always go off on all these tangents,

109
00:07:22,640 --> 00:07:27,920
the principal thing that's important about this is this hopefully is

110
00:07:27,920 --> 00:07:34,560
a comparable model to chat GPT that we can hopefully run on our own computers,

111
00:07:34,560 --> 00:07:38,240
like you said, for the cost of running your computer.

112
00:07:38,240 --> 00:07:42,920
So just a minor amount of electricity.

113
00:07:42,920 --> 00:07:47,840
Maybe sometimes not so minor, but that's all there is.

114
00:07:47,840 --> 00:07:55,760
So make sure to go and get your name in the hat,

115
00:07:55,760 --> 00:07:58,720
because I think you have to request access,

116
00:07:58,720 --> 00:08:02,400
but I did and within maybe two to three hours,

117
00:08:02,400 --> 00:08:08,200
got a download link and I got that last evening.

118
00:08:08,200 --> 00:08:12,440
I always have a thousand things to do.

119
00:08:12,440 --> 00:08:16,440
So today, I will download these.

120
00:08:16,440 --> 00:08:18,640
I think the smaller ones may be around

121
00:08:18,640 --> 00:08:23,160
seven gigabytes and the larger ones around 70 gigabytes.

122
00:08:23,160 --> 00:08:29,320
So hopefully, this is a lot of power in our tool belt.

123
00:08:29,320 --> 00:08:31,680
Basically, why do we need this?

124
00:08:31,680 --> 00:08:36,160
We've got thousands upon thousands of certificates of analysis and just

125
00:08:36,160 --> 00:08:40,280
cannabis data, the cannabis video and audio.

126
00:08:40,280 --> 00:08:44,680
We've had some cool ideas about how we can go about using this data.

127
00:08:44,680 --> 00:08:48,480
So for example, we got all the patents and we found,

128
00:08:48,480 --> 00:08:51,760
oh, you can talk with the patents and you can

129
00:08:51,760 --> 00:08:55,800
parse the data out of the certificates of analysis.

130
00:08:55,800 --> 00:08:58,920
It's pennies on the dollar,

131
00:08:58,920 --> 00:09:02,000
but pennies on the dollar add up.

132
00:09:02,000 --> 00:09:04,000
So we found, okay,

133
00:09:04,000 --> 00:09:09,600
the cost of parsing a certificate may be anywhere from two to 10 cents.

134
00:09:09,600 --> 00:09:18,360
Well, we have like 17,000 certificates that need to be parsed in Florida.

135
00:09:18,360 --> 00:09:23,960
That's not the most money.

136
00:09:23,960 --> 00:09:26,920
Some of these large cultivations and retailers,

137
00:09:26,920 --> 00:09:29,880
they've got much higher operating expenses.

138
00:09:29,880 --> 00:09:33,240
But we're trying to keep our operating expenses to a minimum.

139
00:09:33,240 --> 00:09:36,080
So if we can run Llama on

140
00:09:36,080 --> 00:09:40,040
our own computer and do the exact same thing,

141
00:09:40,040 --> 00:09:41,840
which is what are we after?

142
00:09:41,840 --> 00:09:43,760
We're after the cannabis data.

143
00:09:43,760 --> 00:09:47,400
So if we can get the cannabis data at a lower cost,

144
00:09:47,400 --> 00:09:51,040
then there's more value to be had.

145
00:09:53,800 --> 00:09:57,280
Even with some of the larger models,

146
00:09:57,280 --> 00:10:02,960
because I only have 32 giga RAM and I do have the 3080 Nvidia card.

147
00:10:02,960 --> 00:10:08,600
But if it's in a container too,

148
00:10:08,600 --> 00:10:12,480
there's a way too that you can really instead of just running everything in RAM,

149
00:10:12,480 --> 00:10:15,000
it's pretty interesting.

150
00:10:15,000 --> 00:10:22,640
Even then, if I just get all the external data embedded into the vector database,

151
00:10:22,640 --> 00:10:26,240
then I should just be able to upload because I have two terabyte on the Cloud.

152
00:10:26,240 --> 00:10:31,280
It's just so difficult getting all the data uploaded to Google Drive.

153
00:10:31,280 --> 00:10:37,160
But once I did, then everybody would be able to download it pretty quickly.

154
00:10:37,160 --> 00:10:40,360
Then also too, because it's already embedded,

155
00:10:40,360 --> 00:10:43,680
that maybe you can use,

156
00:10:43,680 --> 00:10:49,080
we'll just keep training it as part of the pipeline.

157
00:10:49,080 --> 00:10:54,640
Then hopefully it won't be as much memory consumption because I really

158
00:10:54,640 --> 00:10:57,800
find the embedding with the GPU,

159
00:10:57,800 --> 00:11:00,000
it just makes it lightning fast.

160
00:11:00,000 --> 00:11:01,480
But not so much,

161
00:11:01,480 --> 00:11:03,600
I haven't really been using it with the prompts yet.

162
00:11:03,600 --> 00:11:09,080
I don't think that it would be such a problem prompting without GPU,

163
00:11:09,080 --> 00:11:12,520
but embedding, you really need GPU.

164
00:11:12,520 --> 00:11:14,640
It really makes that much of a difference.

165
00:11:14,640 --> 00:11:19,760
I love it. All the technical know-how that you've come by,

166
00:11:19,760 --> 00:11:25,000
Candice, is going to be incredibly useful as we,

167
00:11:25,000 --> 00:11:27,360
for lack of a better word,

168
00:11:27,360 --> 00:11:31,000
fine-tune our models and our processes.

169
00:11:31,000 --> 00:11:34,680
Because we've got some proofs of concept,

170
00:11:34,680 --> 00:11:37,080
but I think we're at the stage now where it's

171
00:11:37,080 --> 00:11:41,240
really time to start implementing them in full.

172
00:11:41,240 --> 00:11:47,720
For example, we said, wouldn't it be cool if we could talk with the traceability data?

173
00:11:47,720 --> 00:11:50,160
Let's do that.

174
00:11:50,160 --> 00:11:53,120
We've got Llama now.

175
00:11:53,120 --> 00:11:55,240
This is a lot of data,

176
00:11:55,240 --> 00:11:57,040
so it's really cool.

177
00:11:57,040 --> 00:11:59,640
We've got a 70-gigabyte model.

178
00:11:59,640 --> 00:12:05,160
We've got 45 gigabytes of Washington State data.

179
00:12:05,160 --> 00:12:12,680
We've got Candice who knows her way around some hardware and software.

180
00:12:17,760 --> 00:12:25,560
Like I said, this is where your computer science background really shines.

181
00:12:25,560 --> 00:12:28,920
But like you said, somehow,

182
00:12:28,920 --> 00:12:34,080
super-powered Llama, add the traceability data to it,

183
00:12:34,080 --> 00:12:36,480
encode it, somehow add it to the model.

184
00:12:36,480 --> 00:12:42,360
Then ideally, we could just talk.

185
00:12:42,360 --> 00:12:50,760
We could say, it'll be interesting to find out what questions can we even ask.

186
00:12:50,760 --> 00:12:54,640
It may not be super great at doing statistics,

187
00:12:54,640 --> 00:12:55,720
or it may or may not.

188
00:12:55,720 --> 00:13:00,080
We could say who has the highest sales in Washington State.

189
00:13:00,080 --> 00:13:05,240
Then this will just be new ground in new territory,

190
00:13:05,240 --> 00:13:11,640
because it may be really good or maybe really bad at doing statistics like that,

191
00:13:11,640 --> 00:13:17,520
but then it may be really good at doing other tasks.

192
00:13:17,520 --> 00:13:24,080
You could say, here's a list of product names.

193
00:13:24,080 --> 00:13:26,120
Could you find all the strain names?

194
00:13:26,120 --> 00:13:28,440
That's a common task we have.

195
00:13:28,440 --> 00:13:33,640
That's something that I think the model would be really good at.

196
00:13:34,560 --> 00:13:37,800
Like I said, I think we've done this in the past.

197
00:13:37,800 --> 00:13:42,080
I think these models are going to be pretty good at strain name generation.

198
00:13:42,080 --> 00:13:45,000
If you just need a creative name,

199
00:13:45,000 --> 00:13:46,760
that's low hanging fruit.

200
00:13:46,760 --> 00:13:50,760
But this is actually where all of you have opportunities.

201
00:13:50,760 --> 00:13:54,400
This is untrod ground.

202
00:13:54,400 --> 00:13:56,040
You're pioneers.

203
00:13:56,040 --> 00:14:02,200
This is just this brand open wild frontier.

204
00:14:02,200 --> 00:14:08,240
No one's really trained a large language model

205
00:14:08,240 --> 00:14:13,240
on the entire population of traceability data in Washington State,

206
00:14:13,240 --> 00:14:15,960
and then started to pick its brain.

207
00:14:15,960 --> 00:14:22,040
We can scrape PubMed cannabis data just like everybody else's too.

208
00:14:22,040 --> 00:14:24,320
That's something we really haven't included yet,

209
00:14:24,320 --> 00:14:27,400
but that would really help too.

210
00:14:28,400 --> 00:14:33,240
I love it. That's really where we should be going,

211
00:14:33,240 --> 00:14:37,880
is basically making it a cannabis expert.

212
00:14:37,880 --> 00:14:40,440
Like I said, just basically feed it.

213
00:14:40,440 --> 00:14:43,360
That's what I would love to do if we can pull it off.

214
00:14:43,360 --> 00:14:46,160
Just feed it everything cannabis,

215
00:14:46,160 --> 00:14:50,400
and then they'll see what happens,

216
00:14:50,400 --> 00:14:55,560
see if it becomes quote unquote, like a cannabis expert.

217
00:14:55,560 --> 00:15:01,520
Like I said, we let it read all of the cannabis data science,

218
00:15:01,520 --> 00:15:03,080
meet up transcripts,

219
00:15:03,080 --> 00:15:07,280
and let it read everything about cannabis on PubMed,

220
00:15:07,280 --> 00:15:13,680
let it look at all the cannabis traceability data.

221
00:15:13,680 --> 00:15:18,200
Like I said, at the very least,

222
00:15:18,200 --> 00:15:23,440
it'll just be a fun exercise in computer science.

223
00:15:25,080 --> 00:15:27,480
I was saying at the beginning,

224
00:15:27,480 --> 00:15:30,000
if statistics and knowledge is powered,

225
00:15:30,000 --> 00:15:36,240
we may unearth just an extraordinary amount of

226
00:15:36,240 --> 00:15:42,480
knowledge that, let's face it, the cannabis industry.

227
00:15:42,680 --> 00:15:46,800
I think a lot of people were forecasting

228
00:15:46,800 --> 00:15:51,360
that it would be generally permitted by now.

229
00:15:51,360 --> 00:15:54,080
I think a lot of businesses were structured that way.

230
00:15:54,080 --> 00:15:56,000
I think a lot of businesses were thinking like,

231
00:15:56,000 --> 00:16:00,320
okay, we're maybe in a little bit of a gray market now,

232
00:16:00,320 --> 00:16:03,480
like back in 2018 to 2020 or so,

233
00:16:03,480 --> 00:16:09,720
but things are growing and it looks like maybe one day,

234
00:16:09,720 --> 00:16:13,000
we'll at least be permitted at the federal level,

235
00:16:13,000 --> 00:16:15,240
so that way people can use banks,

236
00:16:15,240 --> 00:16:19,400
and just in general, just operate like a normal business.

237
00:16:19,400 --> 00:16:22,040
Some people were taking out loans

238
00:16:22,040 --> 00:16:24,880
and trying to weather the storm.

239
00:16:24,880 --> 00:16:30,800
I don't think anybody anticipated that here in 2023,

240
00:16:30,800 --> 00:16:34,200
the legal, I don't know,

241
00:16:34,200 --> 00:16:35,400
I guess this is debatable,

242
00:16:35,400 --> 00:16:41,200
but in a way, the legal state is similar to 2018,

243
00:16:41,200 --> 00:16:44,720
where the states have permitted it,

244
00:16:44,720 --> 00:16:50,560
but the banks, I think there's some state banks that allow it,

245
00:16:50,560 --> 00:16:52,440
but I think maybe even some of

246
00:16:52,440 --> 00:16:54,760
the larger banks are maybe even becoming

247
00:16:54,760 --> 00:16:58,120
increasingly frowning on cannabis.

248
00:16:58,120 --> 00:17:01,640
I know some of the big companies,

249
00:17:01,640 --> 00:17:04,680
they're slowly coming around.

250
00:17:04,680 --> 00:17:09,640
For example, LinkedIn has been a little bit cannabis friendly,

251
00:17:09,640 --> 00:17:13,040
and then that had an influence on Twitter.

252
00:17:14,040 --> 00:17:18,200
There are cannabis companies on Instagram,

253
00:17:18,200 --> 00:17:21,840
but I know they always have trouble there.

254
00:17:21,840 --> 00:17:29,280
Anywho, I'm just rambling.

255
00:17:29,280 --> 00:17:33,520
I'll pass it off to any of you

256
00:17:33,520 --> 00:17:37,640
because now's the time to get your thoughts out there.

257
00:17:37,640 --> 00:17:40,600
Then I'll actually share with you

258
00:17:40,600 --> 00:17:44,440
what this big body of data looks like

259
00:17:44,440 --> 00:17:49,760
and what insights a human can find by just scraping the surface,

260
00:17:49,760 --> 00:17:53,000
and then we can just wonder what the large language models can do.

261
00:17:53,000 --> 00:17:57,240
But before I ramble any further, Robert,

262
00:17:57,240 --> 00:18:01,440
any thoughts, comments, questions, anything on your mind?

263
00:18:01,640 --> 00:18:05,360
Just building on last week,

264
00:18:05,360 --> 00:18:07,720
I was curious, let's say,

265
00:18:07,720 --> 00:18:14,720
looking at the different scores from the different labs,

266
00:18:14,720 --> 00:18:19,640
I guess it could be interesting to do some analysis of variance,

267
00:18:19,640 --> 00:18:27,240
and to see if certain labs would skew one way or another

268
00:18:27,240 --> 00:18:32,760
just across the board with different metrics that they report on

269
00:18:32,760 --> 00:18:40,120
and different companies or cannabis companies that they test.

270
00:18:40,120 --> 00:18:46,800
That seemed pretty interesting as an open-ended question.

271
00:18:46,800 --> 00:18:50,640
But it's just something to really think about

272
00:18:50,640 --> 00:18:55,240
that when these various labs have this tendency

273
00:18:55,240 --> 00:18:59,400
to over inflate some of their scores,

274
00:18:59,400 --> 00:19:04,480
it makes for interesting reading material

275
00:19:04,480 --> 00:19:07,880
when you see some of the articles that are out there.

276
00:19:07,880 --> 00:19:12,400
I like a point you brought up today,

277
00:19:12,400 --> 00:19:19,200
where you may have to add condition upon condition in this case,

278
00:19:19,200 --> 00:19:26,880
because the variance,

279
00:19:26,880 --> 00:19:33,960
it's going to be all over the board and for some natural reasons,

280
00:19:33,960 --> 00:19:36,360
and that's where a lot of the hand-waving comes in,

281
00:19:36,360 --> 00:19:39,480
because cannabis,

282
00:19:39,480 --> 00:19:43,840
it's this necure old plant.

283
00:19:45,920 --> 00:19:49,800
Just measuring things.

284
00:19:53,000 --> 00:20:01,640
Awesome. Basically, measuring is the name of the game.

285
00:20:01,640 --> 00:20:09,720
Gosh, I was just hearing something about measuring,

286
00:20:09,720 --> 00:20:14,200
but I haven't ingrained it enough to drive it home.

287
00:20:14,200 --> 00:20:19,320
But measurement is basically at the heart of science.

288
00:20:19,320 --> 00:20:23,720
You get a good scientific question,

289
00:20:23,720 --> 00:20:30,040
or even in engineering, you have to measure.

290
00:20:30,040 --> 00:20:33,960
It all comes down to the measurement.

291
00:20:36,200 --> 00:20:39,840
How people are measuring it,

292
00:20:39,840 --> 00:20:44,560
and then there's variation there.

293
00:20:44,560 --> 00:20:47,880
What are they measuring?

294
00:20:47,880 --> 00:20:55,440
The cannabinoids, terpenes, pesticides, the moisture.

295
00:20:55,440 --> 00:20:59,200
That matters.

296
00:20:59,200 --> 00:21:02,760
When are they testing this?

297
00:21:02,760 --> 00:21:07,200
That's something that I thought was interesting is,

298
00:21:07,200 --> 00:21:12,240
say when is a product tested versus when is it sold?

299
00:21:12,240 --> 00:21:21,840
We've talked about this a little in the past where you can inject your cannabinoids naturally to grade.

300
00:21:21,840 --> 00:21:26,880
You could say, something's been on the shelf for eight months,

301
00:21:26,880 --> 00:21:33,200
and you could estimate that maybe it's a certain percentage of its original concentration.

302
00:21:33,200 --> 00:21:37,440
The original point is, can you look at variance by labs?

303
00:21:37,440 --> 00:21:42,280
Yes. Can you look at variances by analytes?

304
00:21:42,280 --> 00:21:44,200
Yes.

305
00:21:44,920 --> 00:21:47,760
I'd like to add something here.

306
00:21:47,760 --> 00:21:51,920
Yes, please. I don't know where I'm going.

307
00:21:51,920 --> 00:21:59,200
I think this is an absolutely fantastic question, and it's an area that I'm really interested in.

308
00:21:59,200 --> 00:22:06,960
Part of the problem with what you're asking is trying to establish a baseline for what can we expect for a certain strain.

309
00:22:06,960 --> 00:22:14,240
If we look at a strain and we say, we think that it's on average 20%.

310
00:22:14,240 --> 00:22:20,560
If we look at that same exact strain and all the different tests that have been done for it,

311
00:22:20,560 --> 00:22:26,800
we're going to see a variance of probably 15% to 25% or perhaps even more.

312
00:22:26,800 --> 00:22:36,480
The question is, how much of that variance is truly variance in the actual plant versus fudging or bad fake stuff?

313
00:22:36,480 --> 00:22:41,280
Then there's things going on where the lab is testing legitimately.

314
00:22:41,280 --> 00:22:45,680
I know someone who's addressing this issue specifically.

315
00:22:45,680 --> 00:22:53,360
The growers have their samples and they send them to a lab to be tested.

316
00:22:53,360 --> 00:23:00,960
Well, this whole process of which samples do they send, they have a batch of product.

317
00:23:00,960 --> 00:23:03,680
There's different limits in different states.

318
00:23:03,680 --> 00:23:08,400
In California, I believe it's 50 pounds is the maximum size per batch.

319
00:23:08,400 --> 00:23:15,040
Now you have 50 pounds of cannabis and you need to pick a sample from those 50 pounds to send to the lab.

320
00:23:15,040 --> 00:23:16,560
Well, how do you get that sample?

321
00:23:16,560 --> 00:23:25,280
Are you going to pick the low-lying leaf that's very low in terpenoids or in cannabinoids, in trichomes rather,

322
00:23:25,280 --> 00:23:27,600
or are you going to pick the primo bud?

323
00:23:27,600 --> 00:23:31,120
A lot of places are even leasing their samples.

324
00:23:31,120 --> 00:23:39,680
You can legitimately send it to the lab and that normally 20% is now testing at say 35%.

325
00:23:39,680 --> 00:23:45,200
There's so many different factors to tease out there.

326
00:23:45,200 --> 00:23:54,720
This is one of the big questions that I've been wondering is if we can hone in on a particular strain

327
00:23:54,720 --> 00:23:57,520
and look at the variance, how much of that is natural?

328
00:23:57,520 --> 00:24:06,240
How much of it is manageable because from a user perspective, if you're looking to have a specific experience,

329
00:24:06,240 --> 00:24:08,800
you don't want that 15% to 25% variance.

330
00:24:08,800 --> 00:24:15,120
You want the 20% number every time you buy a plant in order to get predictable results.

331
00:24:15,120 --> 00:24:21,200
The industry would like to narrow that down, but so many different things in the process,

332
00:24:21,200 --> 00:24:29,920
including good, benevolent types of things, problems in the procedures, but also nefarious activity.

333
00:24:29,920 --> 00:24:34,080
There's so much in there.

334
00:24:34,080 --> 00:24:41,280
You said it incredibly well, Ruth, and you brought up some good points that I couldn't quite articulate.

335
00:24:41,280 --> 00:24:45,600
I think it was basically like, what are you measuring?

336
00:24:45,600 --> 00:24:49,760
That's where sampling is so incredibly important.

337
00:24:49,760 --> 00:25:00,160
The really good scientists really hammer this home when you're talking to good chemists or good lab directors.

338
00:25:00,160 --> 00:25:04,720
They definitely pinpoint this or even good cultivators.

339
00:25:04,720 --> 00:25:14,080
They definitely pinpoint that as the primary weakness, but it's just not like a glamorous part of the supply chain.

340
00:25:14,080 --> 00:25:28,640
At the regulatory meetings, it's just more glamorous to talk about, oh, we need the mass spectrometer,

341
00:25:28,640 --> 00:25:35,440
and oh, actually, we should actually be testing by GC-GCMS.

342
00:25:35,440 --> 00:25:40,160
Those are the highfalutin conversations.

343
00:25:40,160 --> 00:25:48,480
But really, it's okay, who's driving to go pick up the sample?

344
00:25:48,480 --> 00:25:51,360
Are they sampling it themselves?

345
00:25:51,360 --> 00:26:00,640
It's just not like a glamorous conversation, but it's just the most important.

346
00:26:00,640 --> 00:26:03,760
It's like, who's actually getting the sample?

347
00:26:03,760 --> 00:26:08,640
Is it the cultivator or is it someone at the laboratory?

348
00:26:08,640 --> 00:26:14,800
Because as we just pointed out, if it's somebody at the cultivation,

349
00:26:14,800 --> 00:26:20,800
well, they have the incentive to pick, say, the choicest buds or the choicest flowers.

350
00:26:20,800 --> 00:26:26,800
Once again, this is all just hearsay allegations,

351
00:26:26,800 --> 00:26:33,760
but someone in Oregon was complaining that people would pass pesticide testing,

352
00:26:33,760 --> 00:26:39,200
and then they would just send in the sample that passed pesticide testing

353
00:26:39,200 --> 00:26:42,720
over and over and over again for different batches.

354
00:26:42,720 --> 00:26:48,160
Once again, that may be just entirely just fictitious.

355
00:26:48,160 --> 00:26:52,080
They may just be just pulling something out of their hat.

356
00:26:52,080 --> 00:26:57,680
But I think this is a critical point.

357
00:26:57,680 --> 00:27:04,640
I don't know for certain, but I think some states that are passing newer testing regulations

358
00:27:04,640 --> 00:27:06,880
are taking this into consideration.

359
00:27:06,880 --> 00:27:12,080
So double check me on this, but check out the state of New York.

360
00:27:12,080 --> 00:27:18,160
I've got a sneaking suspicion that they may have samplers there.

361
00:27:20,160 --> 00:27:21,360
I believe they do.

362
00:27:21,360 --> 00:27:27,200
I've talked to someone in New York who, what he does is online education,

363
00:27:27,200 --> 00:27:29,360
but also education in general.

364
00:27:29,360 --> 00:27:35,680
And one of the things that he started with was training samplers

365
00:27:36,800 --> 00:27:44,080
to go out and actually gather the samples that are going to be then sent to the labs for testing.

366
00:27:45,360 --> 00:27:48,960
Obviously, some states allow the cultivators to choose the samples,

367
00:27:48,960 --> 00:27:51,920
and others require an independent sampling person.

368
00:27:53,680 --> 00:27:56,560
So now they might be corruptible.

369
00:27:56,560 --> 00:28:00,080
I don't know, but all of this is very interesting.

370
00:28:00,080 --> 00:28:06,560
And it will be interesting to see how it all plays out and how well the system ends up working,

371
00:28:06,560 --> 00:28:10,480
because there's enough variation in the plant itself

372
00:28:11,280 --> 00:28:14,160
that when we're dealing with all this other stuff on top of that,

373
00:28:14,880 --> 00:28:19,200
I think the problem becomes really intractable for consumers who are, as I said,

374
00:28:19,200 --> 00:28:21,040
trying to get a consistent experience.

375
00:28:21,040 --> 00:28:24,160
Yes, I think this is what we need to spend more time thinking about.

376
00:28:24,160 --> 00:28:31,920
And I mean, just my thought just now is, if you are, say, getting it,

377
00:28:31,920 --> 00:28:36,080
the whole point about the lab testing is you're getting a third party

378
00:28:36,080 --> 00:28:39,840
to independently test your products.

379
00:28:40,720 --> 00:28:46,160
Well, the sampling is part of the test, one could think.

380
00:28:46,160 --> 00:28:46,660
Right?

381
00:28:46,660 --> 00:28:52,580
When we were doing statistics, gathering the sample is the very first step.

382
00:28:53,300 --> 00:29:01,220
So when you're doing an analysis of cannabis, the very first step is getting the sample.

383
00:29:01,220 --> 00:29:04,180
And that's, I think, what people argued in Oregon is,

384
00:29:04,980 --> 00:29:12,180
if the very first step is compromised or untraceable,

385
00:29:12,180 --> 00:29:21,380
then basically the argument is the whole chain of custody is no good.

386
00:29:21,380 --> 00:29:21,880
Right?

387
00:29:21,880 --> 00:29:30,020
Because it's like you said, this is what Candice always reminds me, garbage in, garbage out.

388
00:29:30,020 --> 00:29:36,420
So it's particularly true in this case, except it's the opposite.

389
00:29:36,420 --> 00:29:38,420
It's the same thing.

390
00:29:38,420 --> 00:29:44,820
It's a primo in, primo out.

391
00:29:44,820 --> 00:29:51,380
So if you just think about the economics of it, the only other alternative

392
00:29:51,380 --> 00:29:54,660
is what some people have suggested in Washington state.

393
00:29:54,660 --> 00:29:58,260
And that's just, I've heard cultivators make this argument.

394
00:29:58,260 --> 00:30:02,340
And under the current rules, it's the most economically logical,

395
00:30:02,340 --> 00:30:05,380
is just let people pick the choicest buds.

396
00:30:05,380 --> 00:30:10,100
Because it's basically like, if you're actually letting them pick themselves,

397
00:30:10,820 --> 00:30:13,620
some people are going to be picking the choicest buds.

398
00:30:13,620 --> 00:30:16,820
So why not just have everybody pick the choicest buds?

399
00:30:16,820 --> 00:30:22,900
But then that negates the whole point of this testing,

400
00:30:22,900 --> 00:30:27,300
is to have a representative sample for the consumer.

401
00:30:27,300 --> 00:30:33,060
And I mean, this goes back to statistics.

402
00:30:33,060 --> 00:30:37,940
As you begin taking a statistics course, a nice dry dusty statistics course,

403
00:30:37,940 --> 00:30:42,740
unless you get a fun professor, they're going to really drive home.

404
00:30:43,460 --> 00:30:48,900
The very first principle is sample versus population.

405
00:30:50,420 --> 00:30:52,740
We could test the entire population,

406
00:30:52,740 --> 00:30:58,340
of the cannabis plant. We could take the entire plant, grind it up,

407
00:30:58,340 --> 00:31:04,580
and put it on the HPLC. But then there would be nothing left to give to the consumer.

408
00:31:04,580 --> 00:31:11,300
So you have to just take a sample of it. You have to just take a bud.

409
00:31:11,860 --> 00:31:16,500
And then as soon as you start sampling from a population,

410
00:31:16,500 --> 00:31:19,140
you're going to have a sample of it.

411
00:31:19,140 --> 00:31:30,100
Sampling from a population, you're going to have a residual error.

412
00:31:31,620 --> 00:31:35,700
Things are going to become an approximation immediately.

413
00:31:36,660 --> 00:31:39,940
And so then how do you combat this?

414
00:31:39,940 --> 00:31:43,460
Well, one, you want to increase your sample size.

415
00:31:44,260 --> 00:31:47,140
And I think in California, they're trying this.

416
00:31:47,140 --> 00:31:52,420
And so for example, I don't know how I got looking around at this,

417
00:31:52,420 --> 00:31:57,380
but I started looking around at batch size in California.

418
00:31:57,380 --> 00:31:58,420
It's funny you bring that up.

419
00:31:58,420 --> 00:32:02,020
So it was like, oh yeah, the maximum batch size is 50 pounds.

420
00:32:02,900 --> 00:32:08,020
And I think I was looking at some labs website and SC Labs was saying

421
00:32:08,020 --> 00:32:11,460
that they want you to send in 14 grams.

422
00:32:11,460 --> 00:32:19,060
And so the whole idea behind that is you may only test 0.2 grams or so,

423
00:32:19,060 --> 00:32:24,020
but the idea is they want to increase their sample size.

424
00:32:24,740 --> 00:32:27,940
And then they'll just grind that all up, homogenize it.

425
00:32:28,580 --> 00:32:32,900
And then they'll just pick out a tiny little piece from that.

426
00:32:32,900 --> 00:32:37,300
And think about it, out of that 14, think about this.

427
00:32:37,300 --> 00:32:46,500
Out of the 50 pounds of cannabis, 14 grams were sent to the laboratory.

428
00:32:48,020 --> 00:32:56,020
About 0.2 grams is then put into a tiny little,

429
00:32:56,020 --> 00:33:02,740
I think they call them voids, like I think like vial of analysis.

430
00:33:02,740 --> 00:33:09,140
So this tiny little bottle, 0.2, and then that's put onto a scientific instrument,

431
00:33:09,620 --> 00:33:21,780
like an HPLC, and then just like the tiniest pieces of this

432
00:33:21,780 --> 00:33:24,340
are then sent through the scientific instrument.

433
00:33:24,340 --> 00:33:29,940
So I mean, it's remarkable because these instruments,

434
00:33:31,940 --> 00:33:34,820
talk about standing on the shoulders of giants,

435
00:33:34,820 --> 00:33:37,460
these instruments are truly remarkable.

436
00:33:37,460 --> 00:33:42,260
And that's why I encourage you all to learn more about the laboratory side.

437
00:33:42,260 --> 00:33:44,980
And I'm truly interested in it.

438
00:33:44,980 --> 00:33:48,420
And that's what I'm learning about because we talked about measuring.

439
00:33:49,060 --> 00:33:51,860
These are the instruments that are used in the laboratory.

440
00:33:51,860 --> 00:33:57,300
We talked about measuring, these are people who are measuring the world.

441
00:33:58,340 --> 00:34:02,420
The chemist, with their mass spectrometers,

442
00:34:03,700 --> 00:34:11,300
just about any organic matter or inorganic, right?

443
00:34:11,300 --> 00:34:17,060
They're testing for heavy metals, just stuff, just matter in the world.

444
00:34:17,060 --> 00:34:19,620
They're measuring it and quantifying it.

445
00:34:19,620 --> 00:34:22,820
And the whole idea is, yes, it's an approximation,

446
00:34:22,820 --> 00:34:26,820
but if you actually let the scientists do their job,

447
00:34:27,380 --> 00:34:29,620
they can do a remarkably good job.

448
00:34:30,420 --> 00:34:38,100
Just sending just particles of this cannabis plant through the instrument,

449
00:34:38,100 --> 00:34:44,340
they can determine, okay, there's about, I don't know, 20% of this

450
00:34:44,340 --> 00:34:55,300
of this petrohydrocannabinol molecule in this plant.

451
00:34:57,140 --> 00:35:01,940
And then, you know, that information can then make it to the consumer.

452
00:35:01,940 --> 00:35:06,180
And then once again, then you're, you know, then it's back to approximation.

453
00:35:06,180 --> 00:35:14,180
Then you, oh, maybe you put about a third of a gram in a bomb and let that rip,

454
00:35:14,180 --> 00:35:19,540
or, you know, put it into some butter and make some cookies or whatever you have.

455
00:35:21,620 --> 00:35:30,100
So, anywho, I kind of went on a long-winded tangent,

456
00:35:30,100 --> 00:35:37,460
but I think this is all really cool, instrumentally important.

457
00:35:38,740 --> 00:35:45,940
And yes, there's going to be tons of variance and tons of ways things get wrong,

458
00:35:45,940 --> 00:35:50,180
can go wrong, but as much as possible,

459
00:35:50,180 --> 00:35:57,060
if we can let the scientists actually, you know, do like,

460
00:35:57,060 --> 00:36:00,660
go through the scientific process without, you know,

461
00:36:00,660 --> 00:36:02,500
tying their hands behind their back and saying,

462
00:36:02,500 --> 00:36:05,860
actually, you can't even, you know, sample the cannabis plant.

463
00:36:08,340 --> 00:36:10,900
You know, I think we could get a better outcome.

464
00:36:10,900 --> 00:36:16,100
But any thoughts, comments, questions to piggyback on that before I

465
00:36:16,100 --> 00:36:24,580
pick up with a curve ball from left field?

466
00:36:24,580 --> 00:36:28,580
Please give this.

467
00:36:28,580 --> 00:36:33,380
Well, this is off base, but you know what, Keegan, I am so excited with this private GPT.

468
00:36:33,380 --> 00:36:36,980
I'm already thinking about the CanLivix hybrid data center.

469
00:36:36,980 --> 00:36:39,940
And I even went and kind of was looking at some use racks,

470
00:36:39,940 --> 00:36:44,020
you know, when they some raid and, you know, lots of memory, nice server.

471
00:36:44,020 --> 00:36:47,700
Rack in there too. But this could be pretty big.

472
00:36:47,700 --> 00:36:50,980
I'm really excited, you know, and I really think too, though,

473
00:36:50,980 --> 00:36:56,500
that keeping the CanLivix maybe a bit privatized and the hybrid data center

474
00:36:56,500 --> 00:36:58,260
may not be a bad idea either.

475
00:36:58,260 --> 00:37:02,340
And then, you know, you have your API for access and but I don't know,

476
00:37:02,340 --> 00:37:04,340
you know, but I'm excited about it.

477
00:37:04,340 --> 00:37:06,420
I'm really excited about this llama.

478
00:37:06,420 --> 00:37:09,300
Bravo, Mark Zuckerberg.

479
00:37:10,340 --> 00:37:12,180
I love healthy competition.

480
00:37:12,180 --> 00:37:14,820
I love healthy competition with the billionaires.

481
00:37:14,820 --> 00:37:17,780
I love it when they throw us surf some crumbs, you know,

482
00:37:18,820 --> 00:37:22,740
just makes life so much better for us being able to, you know,

483
00:37:22,740 --> 00:37:27,380
stay in the game without, you know, having to spend so much money to with GPT-4.

484
00:37:27,380 --> 00:37:31,220
And then somebody was telling me on LinkedIn also that GPT-4 actually,

485
00:37:31,220 --> 00:37:36,420
they are using reinforcement learning because, you know, I was kind of thinking of chat GPT-4

486
00:37:36,420 --> 00:37:42,420
is never really having any conscious, right, consciousness because of the fact that I didn't

487
00:37:42,420 --> 00:37:44,500
think reinforcement learning was used.

488
00:37:44,500 --> 00:37:48,340
So, you know what, that's something we're going to need too to add into our pipeline

489
00:37:48,340 --> 00:37:49,700
as we start training our model.

490
00:37:49,700 --> 00:37:51,460
I'm just so excited about it.

491
00:37:51,460 --> 00:37:55,940
And I have some friends too in that technical skills shares group and they do DevOps.

492
00:37:55,940 --> 00:37:57,700
And there's this guy in Boston.

493
00:37:57,700 --> 00:38:02,340
He is a genius with DevOps and Kubernetes, Docker.

494
00:38:02,340 --> 00:38:06,660
Oh my goodness, I'm so excited because I know he's going to want to come on board

495
00:38:06,660 --> 00:38:08,740
with the ML dev ops, right?

496
00:38:08,740 --> 00:38:10,100
And I don't know.

497
00:38:11,220 --> 00:38:13,220
The future just seems so bright today.

498
00:38:13,220 --> 00:38:14,660
I just can't stop smiling.

499
00:38:17,220 --> 00:38:19,940
I love all the things you just brought up, Candice.

500
00:38:19,940 --> 00:38:25,860
One, I love your go get them attitude about setting up some servers and getting this data

501
00:38:25,860 --> 00:38:31,460
to people because I've seen a bright, awesome future there too,

502
00:38:31,460 --> 00:38:36,340
because like you said, maybe of course you may have to earn a few bucks

503
00:38:38,260 --> 00:38:40,660
if you put it together in awesome API and putting everything together.

504
00:38:40,660 --> 00:38:43,060
I'll have to start working for money, yeah.

505
00:38:44,420 --> 00:38:47,060
Because I started looking at the pricing.

506
00:38:47,060 --> 00:38:49,540
Even with used equipment, it's still pretty pricey.

507
00:38:49,540 --> 00:38:54,980
But I would start with a small rack and then just have a server, RAID, the network.

508
00:38:54,980 --> 00:38:56,980
And then we can always build from there.

509
00:38:57,700 --> 00:39:00,260
But I think my laptop will be good for a bit.

510
00:39:00,260 --> 00:39:01,380
We have 12 terabytes.

511
00:39:01,380 --> 00:39:02,500
We have the Nvidia.

512
00:39:02,500 --> 00:39:03,940
We have the 32 gig of RAM.

513
00:39:05,940 --> 00:39:08,260
And it's good to get this into people's hands, right?

514
00:39:08,260 --> 00:39:10,020
This is what people want.

515
00:39:10,020 --> 00:39:17,860
And they need to be able to digest this data because it's undigestible at the moment.

516
00:39:18,980 --> 00:39:21,700
And I can start showing you some of that.

517
00:39:21,700 --> 00:39:25,460
But a quick analogy here and a principle from economics.

518
00:39:25,460 --> 00:39:30,900
And I wish there was a better comparison because I'm not like a soda drinker.

519
00:39:30,900 --> 00:39:33,140
You know, I drink tea and coffee.

520
00:39:33,140 --> 00:39:38,500
But in economics, I think it's the Bertrand model.

521
00:39:38,500 --> 00:39:45,780
But a lot of these industrial organization economists study market structure.

522
00:39:45,780 --> 00:39:50,180
And that's what we started out about was just looking at how many firms are in the market.

523
00:39:50,180 --> 00:39:57,060
But for, you know, say like Facebook and Microsoft and these big technology companies,

524
00:39:58,980 --> 00:40:09,860
the one bright side about this is it's kind of like Coke and Pepsi, where it's not going to be

525
00:40:09,860 --> 00:40:13,540
great if you go and try to sell another soda.

526
00:40:13,540 --> 00:40:21,860
So it's not going to be profitable necessarily to go try to sell a large language model,

527
00:40:21,860 --> 00:40:29,860
in my opinion, unless you're Google or Microsoft or Facebook.

528
00:40:30,580 --> 00:40:39,620
But the good thing about this is the Bertrand model is it argues that it'll push the price down to cost

529
00:40:39,620 --> 00:40:48,580
because it's basically just these two large companies competing for market share.

530
00:40:48,580 --> 00:40:52,180
And they basically both have the same product.

531
00:40:52,180 --> 00:40:55,540
So all they can really compete on is price.

532
00:40:55,540 --> 00:41:01,540
And what ends up happening is they just push price down to cost, which, like I said,

533
00:41:01,540 --> 00:41:06,820
it's not great if you're trying to sell soda, but say, hey, if you're a restaurant

534
00:41:06,820 --> 00:41:14,340
and you sell food and you know, you know, sometimes you put soda in a glass and give it to people.

535
00:41:14,900 --> 00:41:18,260
Well, it's good for you because so it is going to be super, super cheap.

536
00:41:19,060 --> 00:41:22,820
And so it's the age of the restaurants.

537
00:41:22,820 --> 00:41:28,100
But in this case, it's going to be like sort of the age of the data restaurants where

538
00:41:29,380 --> 00:41:34,500
it's all about you kind of packaging this up and delivering it to consumers.

539
00:41:34,500 --> 00:41:39,060
So like I said, I wouldn't recommend you go try to create your own llama.

540
00:41:39,780 --> 00:41:48,020
But I think if we start to use them, we start to use llama and that'll put the emphasis on

541
00:41:48,020 --> 00:41:50,340
chat GPT to lower their costs.

542
00:41:51,060 --> 00:41:55,780
The cost of using these large language models may.

543
00:41:57,220 --> 00:42:02,820
Well, the cost to us may go really, really low, just $12,000.

544
00:42:02,820 --> 00:42:07,620
Really, really low just towards the actual cost of actually running the thing.

545
00:42:07,620 --> 00:42:07,860
Great.

546
00:42:07,860 --> 00:42:13,540
Like in the llama case, it's hopefully just going to cost us the power it takes to power our computer.

547
00:42:15,540 --> 00:42:17,460
And so the long story short is.

548
00:42:19,780 --> 00:42:22,180
I think it'll be it's it'll be good for us.

549
00:42:22,180 --> 00:42:27,460
Great competition is always good, even competition amongst the giants.

550
00:42:27,460 --> 00:42:32,020
So, so anywho, I'm excited, I'm excited.

551
00:42:32,740 --> 00:42:39,860
But before we conclude, I guess.

552
00:42:42,420 --> 00:42:44,740
Do you want to look at some data?

553
00:42:44,740 --> 00:42:50,100
I, I, this is a little bit of a weird insight, but.

554
00:42:50,100 --> 00:42:55,860
If you want, I can show you like what having all this data in your pocket can can do.

555
00:42:59,220 --> 00:43:01,540
So once again, I.

556
00:43:06,420 --> 00:43:14,340
Basically, I'll just share with you some of the data that I've been working with and then show show you where you can help out.

557
00:43:14,340 --> 00:43:28,420
So basically, you know, I've been doing these public records requests for the Washington state data, and I think we can finally start to to curate this and make it accessible.

558
00:43:28,420 --> 00:43:31,620
So basically what I'm doing is.

559
00:43:32,580 --> 00:43:38,500
There is say these 22 giant inventory files.

560
00:43:38,500 --> 00:43:41,860
And what I'm doing is, you know, I'm just adding.

561
00:43:43,460 --> 00:43:45,860
All these data points together.

562
00:43:46,740 --> 00:43:54,500
So that way, you know, we have these pretty complete observations about cannabis products.

563
00:43:55,540 --> 00:44:00,900
You know, so here's some fresh frozen flour and clones.

564
00:44:00,900 --> 00:44:08,580
Some mature plants. I saw something with Mars, some some apple tarts.

565
00:44:10,580 --> 00:44:13,780
So we've got tons and tons of data here.

566
00:44:15,140 --> 00:44:17,380
22 inventory files.

567
00:44:18,500 --> 00:44:24,740
Each one may have about half a million to a million observations.

568
00:44:24,740 --> 00:44:32,100
So, yes, so this is starting to add up here.

569
00:44:32,820 --> 00:44:36,980
And just to kind of show you what a human can do.

570
00:44:38,180 --> 00:44:42,660
But then we have to kind of wonder where the large language model can do.

571
00:44:42,660 --> 00:44:48,500
This is just what I was doing this morning, and I thought that this exercise would be worthwhile showing you.

572
00:44:48,500 --> 00:44:55,140
Because this is what we could potentially do with every state, not just Washington.

573
00:44:55,140 --> 00:45:07,460
If they were a bit more forthcoming with their lab results, because once again, you know, say you're a cannabis consumer and you want to learn something about cannabis.

574
00:45:09,460 --> 00:45:17,060
Let's just see if we can't just let you know, you know, we can just let you know, you know, we can just let you know.

575
00:45:17,060 --> 00:45:22,500
Let's just see if we can't just look at the data and and learn something.

576
00:45:24,180 --> 00:45:32,020
And so I was just going to share with you basically what I learned this morning.

577
00:45:32,020 --> 00:45:35,540
I mean, here are actually my my calculations right here.

578
00:45:36,260 --> 00:45:46,820
But basically, the first thing I was curious about was, you know, how many lab tests are happening, you know,

579
00:45:46,820 --> 00:45:53,620
in the whole country? Well, one thing we could potentially do is just do a.

580
00:45:53,620 --> 00:46:04,580
And this is something that a finance professor taught me one time is if you can do a back of the envelope calculation,

581
00:46:04,580 --> 00:46:09,060
there's, you know, they're they can actually kind of get you in the rough.

582
00:46:09,060 --> 00:46:18,820
OK, he said it for this particular he said, one, it'll get you in the rough ballpark and two, they're they're actually super persuasive.

583
00:46:18,820 --> 00:46:21,220
And so basically, like if you're.

584
00:46:22,580 --> 00:46:27,620
In right, this is why I heard it from a finance professor and not an economics professor.

585
00:46:27,620 --> 00:46:30,660
But I think it's a good rule of thought.

586
00:46:30,660 --> 00:46:40,020
I think it's an interesting, an interesting bit of wisdom or perspective.

587
00:46:40,020 --> 00:46:47,940
But he basically said, you know, if you're in the room with some people and you can do a back of the envelope calculation,

588
00:46:47,940 --> 00:46:55,140
it'll be super persuasive and it'll kind of just move things forward because.

589
00:46:55,140 --> 00:47:02,580
Why do we even do statistics? It's just kind of to get a rough mental approximation of something.

590
00:47:02,580 --> 00:47:06,180
Right. We were talking about in the beginning measurement.

591
00:47:06,180 --> 00:47:10,980
Right. So we're just basically trying to measure something.

592
00:47:10,980 --> 00:47:17,300
Sometimes the back of the envelope calculation can be a beginning point, not saying it's.

593
00:47:17,300 --> 00:47:21,620
The end all be all, but it's just a place to begin.

594
00:47:21,620 --> 00:47:26,260
And that's another philosophy that we have is.

595
00:47:26,260 --> 00:47:34,980
If you at least have a place to begin, even if it's bad, then you can start heading in the right direction.

596
00:47:34,980 --> 00:47:48,180
And this is what we I like to call the real life gradient descent in gradient descent is a concept or mathematical principle.

597
00:47:48,180 --> 00:47:57,220
And people, I think, debate who discovered it, but I think you can point some work back to the new new.

598
00:47:57,220 --> 00:48:01,620
But basically, it's the idea is.

599
00:48:01,620 --> 00:48:13,540
Start anywhere and generally head in the better direction, and you'll wind up at least at the locally best.

600
00:48:13,540 --> 00:48:18,180
So so maybe right there is maybe multiple bests, right?

601
00:48:18,180 --> 00:48:28,020
There may be a lot of different good spots, but you'll at least end up at the best spot in the nearby vicinity.

602
00:48:28,020 --> 00:48:33,860
But OK, so anywho, enough of me rambling, what can we discover from this?

603
00:48:33,860 --> 00:48:37,060
OK, quick back of the envelope calculation here.

604
00:48:37,060 --> 00:48:43,940
So this traceability data has been around for about a year and a half.

605
00:48:43,940 --> 00:48:48,420
So this was an act enacted at the beginning of twenty twenty two.

606
00:48:48,420 --> 00:48:55,580
And we have data through the end of June of twenty twenty three.

607
00:48:55,580 --> 00:48:59,300
So we have about 18 months of data here.

608
00:48:59,300 --> 00:49:08,180
And, you know, we've got sixty one, two, four, nine lab tests.

609
00:49:08,180 --> 00:49:14,180
And so I was saying, OK, you know, how many lab tests per month are there?

610
00:49:14,180 --> 00:49:22,420
You know, so there's around, you know, thirty four hundred lab tests per month in Washington state.

611
00:49:22,420 --> 00:49:28,180
And then I was saying, you know, what is the population of Washington?

612
00:49:28,180 --> 00:49:36,300
And once again, this is not going to be apples to apples because all the states have different testing regulations.

613
00:49:36,300 --> 00:49:41,820
But, you know, so I just looked up the population of Washington.

614
00:49:41,820 --> 00:49:51,220
And then the thing we've been doing is just saying, oh, how many dispensaries per one hundred thousand people?

615
00:49:51,220 --> 00:50:02,940
And so I was I miss you've missed this up. So basically what I'm trying to figure out here is.

616
00:50:02,940 --> 00:50:14,820
You know, on average, how many lab tests per month per one hundred thousand people are occurring in Washington?

617
00:50:14,820 --> 00:50:20,540
So, you know, about 44 lab tests per one hundred thousand people.

618
00:50:20,540 --> 00:50:29,700
And then I realized, well, just once again, it'll be an approximation because every state has different testing rules.

619
00:50:29,700 --> 00:50:44,940
But you could kind of, you know, go through and start, you know, just estimating about how many lab tests would be occurring.

620
00:50:44,940 --> 00:50:52,620
In every state per month. So this is California with thirty nine million people.

621
00:50:52,620 --> 00:50:59,740
And they may be doing right. Right. Just any approximation is better than no approximation.

622
00:50:59,740 --> 00:51:07,340
Because I was kind of wondering, you know, what proportion of lab tests are we even observing there?

623
00:51:07,340 --> 00:51:13,620
Because we've observed a few thousand from a few different producers.

624
00:51:13,620 --> 00:51:21,340
And then I think we observed I don't remember how many, maybe five or six thousand from SC labs.

625
00:51:21,340 --> 00:51:24,140
But that may only be like a drop in the bucket. Right.

626
00:51:24,140 --> 00:51:28,780
We've right. We may have only seen five or six thousand lab results,

627
00:51:28,780 --> 00:51:40,460
but there may be more than one hundred thousand lab results per year in California.

628
00:51:40,460 --> 00:51:47,260
That's a lot. And I mean, that data is out there.

629
00:51:47,260 --> 00:51:51,180
One would assume. Right.

630
00:51:51,180 --> 00:51:58,380
And they've been testing in California and maybe since 2018, maybe not.

631
00:51:58,380 --> 00:52:05,980
But, you know, I mean, what if there's five years of one hundred thousand samples?

632
00:52:05,980 --> 00:52:13,740
It could be five hundred thousand lab tests in California alone.

633
00:52:13,740 --> 00:52:18,220
And then, you know, then there's Colorado.

634
00:52:18,220 --> 00:52:22,300
I'm trying to think of some other states where there's been testing for many years.

635
00:52:22,300 --> 00:52:27,220
But, you know, Michigan, Massachusetts.

636
00:52:27,220 --> 00:52:33,780
And then, you know, all the medical states, you know, they're just just kind of picking some random ones out of the hat.

637
00:52:33,780 --> 00:52:40,220
Ohio and so on and so forth. So tons and tons of data out there.

638
00:52:40,220 --> 00:52:46,940
And, you know, why is this? OK, cool. So there's tons of lab results.

639
00:52:46,940 --> 00:52:52,540
You know, why is this even important? Well, because.

640
00:52:52,540 --> 00:52:56,540
Check this out. What can you even do with this data?

641
00:52:56,540 --> 00:53:01,060
So this is sixty two thousand lab results.

642
00:53:01,060 --> 00:53:05,140
And say you're a cannabis consumer in Washington state.

643
00:53:05,140 --> 00:53:14,740
Well, which I am. One thing I may be concerned about or that's not what we want to do.

644
00:53:14,740 --> 00:53:23,820
I thought I would make the screen large. But one thing that I'm concerned about are our pesticides.

645
00:53:23,820 --> 00:53:33,500
So here I've just identified, you know, all the samples with with pesticides in them.

646
00:53:33,500 --> 00:53:47,100
And so, well. Just today, I'm just going to do this kind of interesting exercise where we basically just.

647
00:53:47,100 --> 00:53:57,180
Look at and to just going to look at products and in the pesticides that are in them,

648
00:53:57,180 --> 00:54:03,060
because that's all I'm really interested in. Right. So if I go to the store and, you know,

649
00:54:03,060 --> 00:54:11,340
all I'm going to see are basically, you know, who produced it.

650
00:54:11,340 --> 00:54:16,820
Right. That'll hopefully be on the bag. And then, you know, what's in it?

651
00:54:16,820 --> 00:54:22,260
And then, of course, you know, you'll you'll see the THC, the THC content,

652
00:54:22,260 --> 00:54:27,460
where the THC content and CBD contents on the label.

653
00:54:27,460 --> 00:54:30,740
Well, the pesticide may not necessarily be on the label.

654
00:54:30,740 --> 00:54:42,180
So now, you know, I basically augmented the information that's available at the dispensary.

655
00:54:42,180 --> 00:54:47,780
That's awesome. Right. I'm now like an empowered consumer.

656
00:54:47,780 --> 00:54:51,580
And, you know, basically.

657
00:54:51,580 --> 00:54:55,580
Let's say I want to avoid pesticides.

658
00:54:55,580 --> 00:55:00,700
Well, you can basically just. You can just sort.

659
00:55:00,700 --> 00:55:10,900
So you can basically say, OK. All these products, whether they may have passed or not,

660
00:55:10,900 --> 00:55:15,820
there's there's some trace amount of pesticide there.

661
00:55:15,820 --> 00:55:25,060
So, you know, so you can just immediately say you were really concerned about pesticides.

662
00:55:25,060 --> 00:55:31,580
You can just start looking for products, you know, here and down.

663
00:55:31,580 --> 00:55:37,220
So once again, this isn't going to be super accessible yet.

664
00:55:37,220 --> 00:55:42,740
But that's where the idea behind these say the large language model comes in.

665
00:55:42,740 --> 00:55:47,980
What if you train the model on this data and then you can just say,

666
00:55:47,980 --> 00:55:53,540
say you're at the the dispensary and you're saying, hey, you know,

667
00:55:53,540 --> 00:56:01,500
does this bubble hash by alpha crux have any pesticides in it?

668
00:56:01,500 --> 00:56:08,740
And it'll just say, no, no pesticides were detected. And then you could say, right.

669
00:56:08,740 --> 00:56:11,380
And this could all just be natural language, right.

670
00:56:11,380 --> 00:56:13,420
It's a natural language model.

671
00:56:13,420 --> 00:56:16,820
So this is what we wanted to make available.

672
00:56:16,820 --> 00:56:23,660
We want to just on your phone, just say, hey, I'm at.

673
00:56:23,660 --> 00:56:28,740
Retailer X, looking at.

674
00:56:28,740 --> 00:56:31,740
This ethanol concentrate.

675
00:56:31,740 --> 00:56:34,180
Were there any pesticides detected?

676
00:56:34,180 --> 00:56:38,860
And they're like, there was a trace amount of by Fennel's eight detected,

677
00:56:38,860 --> 00:56:41,460
but it did pass quality control test.

678
00:56:41,460 --> 00:56:46,740
And so then you can say and it was like, oh, and there was also acetone,

679
00:56:46,740 --> 00:56:56,100
ethanol and isopropanol detected in this concentrate, maybe all at all at minor levels.

680
00:56:56,100 --> 00:57:01,500
And then you're like, I'm going to go with the bubble hash, you know,

681
00:57:01,500 --> 00:57:04,020
and that's that's what I would do as a consumer. Right.

682
00:57:04,020 --> 00:57:06,060
There's two products.

683
00:57:06,060 --> 00:57:10,220
That's all the information you see at the retailer.

684
00:57:10,220 --> 00:57:13,700
And if you ask the bud tender, like you said, like, you know,

685
00:57:13,700 --> 00:57:17,100
maybe they'll get out the certificate, but.

686
00:57:17,100 --> 00:57:19,540
You know.

687
00:57:19,540 --> 00:57:23,540
They may or may not be have all this information,

688
00:57:23,540 --> 00:57:28,460
and that's where we kind of talked about last week where these large language models,

689
00:57:28,460 --> 00:57:32,140
it's like asking thousands and thousands of experts.

690
00:57:32,140 --> 00:57:37,940
Right. The bud is a cannabis expert, so to speak.

691
00:57:37,940 --> 00:57:43,100
But, you know, they only have so much knowledge, so much bandwidth.

692
00:57:43,100 --> 00:57:46,260
So we can augment that.

693
00:57:46,260 --> 00:57:51,740
But so that's the main point I wanted to drive home.

694
00:57:51,740 --> 00:57:58,860
There was other stuff here I kind of wanted to talk about.

695
00:57:58,860 --> 00:58:03,980
Is as far as.

696
00:58:03,980 --> 00:58:09,260
These pesticides go, but what do you think?

697
00:58:09,260 --> 00:58:17,940
Are you interested in me pointing out a sticky issue or just maybe had your heart's content for?

698
00:58:17,940 --> 00:58:21,900
For today, so I'll leave it up to you. Do you want.

699
00:58:21,900 --> 00:58:26,100
More or what are your thoughts?

700
00:58:26,100 --> 00:58:30,020
More. OK.

701
00:58:30,020 --> 00:58:39,700
OK, well, I'll just guide you through my my my reasoning here, and you may call me a data miner.

702
00:58:39,700 --> 00:58:44,940
So this is maybe not how you should the best way to go about it.

703
00:58:44,940 --> 00:58:48,260
The scientific process, right?

704
00:58:48,260 --> 00:58:55,860
Typically, you want the research question first before you just start going and pawing through data.

705
00:58:55,860 --> 00:59:00,940
But basically, my my question was, OK.

706
00:59:00,940 --> 00:59:06,940
What like, you know, what pesticides are being used and why?

707
00:59:06,940 --> 00:59:12,940
And unfortunately, I guess, by who and on what products?

708
00:59:12,940 --> 00:59:21,940
And so we started scrolling through these the pyrethrins, I think, are hotly contested.

709
00:59:21,940 --> 00:59:30,940
So I'm not going to poke into those too much yet because I obviously we can see they're.

710
00:59:30,940 --> 00:59:32,940
They're heavily detected.

711
00:59:32,940 --> 00:59:40,940
I've heard some some debate about whether these are natural or in natural and.

712
00:59:40,940 --> 00:59:48,940
Well, I encourage you to look into this because obviously this is showing up in cannabis products.

713
00:59:48,940 --> 00:59:52,940
And but I think there's more to that more to that story.

714
00:59:52,940 --> 00:59:55,940
So pyrethrins will be for another day.

715
00:59:55,940 --> 01:00:06,940
But what I was scrolling through here and we've we've talked about piper, any old butoxide.

716
01:00:06,940 --> 01:00:10,940
And there's tons of piper, any old butoxide.

717
01:00:10,940 --> 01:00:14,940
And then see tons of it.

718
01:00:14,940 --> 01:00:23,940
And see, this is where just kind of visualizing the data can can really get you a an understanding of it.

719
01:00:23,940 --> 01:00:29,940
I want to say piper, any old butoxide may be permitted.

720
01:00:29,940 --> 01:00:32,940
But once again, not at high levels.

721
01:00:32,940 --> 01:00:36,940
But once again, I started scrolling down here and I started saying, OK,

722
01:00:36,940 --> 01:00:41,940
is there another pesticide that's being used at high numbers?

723
01:00:41,940 --> 01:00:49,940
And look at this. This one's maybe not as much as.

724
01:00:49,940 --> 01:00:53,940
Piper, any old butoxide, but I was just curious.

725
01:00:53,940 --> 01:00:58,940
I just kind of picked one out of the hat like this one looked kind of prevalent.

726
01:00:58,940 --> 01:01:04,940
You know, what what what's pack low butrys all.

727
01:01:04,940 --> 01:01:11,940
And so basically I went over to our our handy dandy chat GPT.

728
01:01:11,940 --> 01:01:17,940
And basically, I just said, you know, what's what's pack low butrys all.

729
01:01:17,940 --> 01:01:20,940
And look at this.

730
01:01:20,940 --> 01:01:24,940
This is something that's been.

731
01:01:24,940 --> 01:01:29,940
That's crossed my radar, but I.

732
01:01:29,940 --> 01:01:37,940
A couple more than once, and I maybe hadn't paid it enough heed.

733
01:01:37,940 --> 01:01:42,940
What I think is going on here.

734
01:01:42,940 --> 01:01:50,940
I've just kind of heard like to talk on the town that people will kind of use this disparagingly,

735
01:01:50,940 --> 01:01:56,940
but these they call them PGR's plant growth regulators.

736
01:01:56,940 --> 01:02:00,940
From my understanding.

737
01:02:00,940 --> 01:02:11,940
Heavy application of the plant growth regulators make the cannabis buds like thick and dense.

738
01:02:11,940 --> 01:02:16,940
Once again, I don't know how much truth this there is to that.

739
01:02:16,940 --> 01:02:24,940
But, you know, if you just look at the quick description here and one thing I love to do with chat GPT is, oh, no.

740
01:02:24,940 --> 01:02:28,940
Can you make this super short summer?

741
01:02:28,940 --> 01:02:34,940
So, right, because chat GPT likes to be a little verbose.

742
01:02:34,940 --> 01:02:42,940
Sure.

743
01:02:42,940 --> 01:02:50,940
So that is maybe too short, but basically it just says, oh, it just results in shorter, more compact plants.

744
01:02:50,940 --> 01:02:58,940
And as I said, I think in cannabis, people use them to get really, really dense buds.

745
01:02:58,940 --> 01:03:08,940
And I think they're they're criticized because it's kind of maybe like kind of like like cheating because it's just.

746
01:03:08,940 --> 01:03:27,940
You know, maybe instead of having, you know, proper technique and really good lighting and, you know, all these different environmental controls to grow a healthy plant, you're kind of just spraying it with this with this chemical that just makes the plant.

747
01:03:27,940 --> 01:03:30,940
Just really dense.

748
01:03:30,940 --> 01:03:41,940
And I think once again, this is where, you know, measurement and science kind of take us like, you know, is this a good thing?

749
01:03:41,940 --> 01:03:43,940
Is this a bad thing?

750
01:03:43,940 --> 01:03:46,940
I think that's up to the debate.

751
01:03:46,940 --> 01:03:53,940
But the point being is.

752
01:03:53,940 --> 01:03:59,940
It's it's prohibited.

753
01:03:59,940 --> 01:04:08,940
And so in Washington, because this is kind of what I was wondering is, you know, in Washington state, they're they're screening for pesticides.

754
01:04:08,940 --> 01:04:13,940
But I was wondering, you know, are these pesticides actually permitted?

755
01:04:13,940 --> 01:04:16,940
And.

756
01:04:16,940 --> 01:04:39,940
And so here's like, and this is kind of what I got to wondering is. I wonder, you know, if licensees are like seeing like this pesticide list here and saying like, oh, you know, here are the action levels for pesticides.

757
01:04:39,940 --> 01:04:48,940
And, you know, maybe as long as we stay below zero point four.

758
01:04:48,940 --> 01:04:50,940
Maybe what's this parts per million?

759
01:04:50,940 --> 01:04:57,940
Maybe if we just stay below zero point four parts per million, you know, we'll be we'll be OK.

760
01:04:57,940 --> 01:05:05,940
But from what I can tell, I don't think you're, you know, you're allowed to use these.

761
01:05:05,940 --> 01:05:16,940
And, you know, I just wanted to to kind of just say, OK, you know, what's going on here?

762
01:05:16,940 --> 01:05:21,940
And so if you look at this.

763
01:05:21,940 --> 01:05:29,940
And see, this is what's what's cool about the Excel is, you know, sometimes an underrated tool.

764
01:05:29,940 --> 01:05:43,940
But she look at this. So, you know, it looks like there may be, you know, a handful of companies here who may be using this plant growth regulator.

765
01:05:43,940 --> 01:05:48,940
But it looks like there's the one company in particular.

766
01:05:48,940 --> 01:05:54,940
And there's actually maybe there's one column here.

767
01:05:54,940 --> 01:06:05,940
They had this failure rate column. See, some of these are failing quality assurance testing, but some of them are passing.

768
01:06:05,940 --> 01:06:16,940
And what I'm worried here is that the ones that are passing, you know, they may be using a prohibited.

769
01:06:16,940 --> 01:06:21,940
And once again, I'm just conjecturing here just from what I've seen looking at the data.

770
01:06:21,940 --> 01:06:31,940
But, you know, it's possible that they may be using a prohibited plant growth regulator.

771
01:06:31,940 --> 01:06:47,940
And, you know, they may just be just willing to accept the fact that, you know, some of their tests will fail quality assurance testing as long as, you know, some, you know, some of them pass.

772
01:06:47,940 --> 01:06:59,940
And so maybe the ones that pass, the buds are so much denser that it that it just makes up for the cost of failing.

773
01:06:59,940 --> 01:07:13,940
And if you look at this, so it once again, not to point any fingers, but this one farm keeps coming up for detections.

774
01:07:13,940 --> 01:07:21,940
And I want to.

775
01:07:21,940 --> 01:07:29,940
If you look at the time stamps of these, it's all from, you know, the past year, year and a half.

776
01:07:29,940 --> 01:07:38,940
So this is all since 2022 and up till potentially the present.

777
01:07:38,940 --> 01:07:46,940
And what you see here, so it's a lot of this farm and they're making flour.

778
01:07:46,940 --> 01:07:51,940
And then what you see down here is look at this.

779
01:07:51,940 --> 01:07:59,940
These companies, right, you've got infused doing concentrates.

780
01:07:59,940 --> 01:08:15,940
JPM doing concentrates, triagonal industries concentrates, you know, these companies may just be sourcing their their cannabis from.

781
01:08:15,940 --> 01:08:34,940
And once again, I'm just conjecturing here, but what it looks like to me is these processors may be sourcing their material from flour that.

782
01:08:34,940 --> 01:08:47,940
That may be contaminated with pack load butyra's all in then, as we know, when you make concentrates, sometimes you concentrate down the pesticides.

783
01:08:47,940 --> 01:08:51,940
And so that's why they may also be getting detections here.

784
01:08:51,940 --> 01:09:01,940
And so you see, once again, some of them pass and some of them fail.

785
01:09:01,940 --> 01:09:05,940
So, so that that's the main thing.

786
01:09:05,940 --> 01:09:09,940
And then I think you could use that data because it's there.

787
01:09:09,940 --> 01:09:12,940
They're collecting information from everyone in the supply chain.

788
01:09:12,940 --> 01:09:16,940
And that's a really interesting hypothesis.

789
01:09:16,940 --> 01:09:26,940
And I bet you could track back and look to see who they're buying from because whoever sold the product had to record that transaction.

790
01:09:26,940 --> 01:09:29,940
Exactly. And so all that data is there.

791
01:09:29,940 --> 01:09:32,940
So we could actually confirm.

792
01:09:32,940 --> 01:09:41,940
And what would be really, really, really interesting is to trace it back and see if the grower passed their test.

793
01:09:41,940 --> 01:09:55,940
But as he said, when it got concentrated in the extraction process, then the less than threshold amounts became greater than threshold amounts once the flour was concentrated into extract.

794
01:09:55,940 --> 01:10:03,940
I love how you're thinking, Ruth. And basically, there's almost like a forensic data scientist.

795
01:10:03,940 --> 01:10:07,940
Yeah. Because the data is there.

796
01:10:07,940 --> 01:10:11,940
But I want to say it's possible.

797
01:10:11,940 --> 01:10:18,940
But sometimes tracing this data, especially between transfers, is ridiculously hard.

798
01:10:18,940 --> 01:10:23,940
But it may be possible. So where there's a will, I'd always like to say there's a way.

799
01:10:23,940 --> 01:10:28,940
But when there's missing data, sometimes that way gets blocked.

800
01:10:28,940 --> 01:10:36,940
But that's I think an interesting... this may have frozen.

801
01:10:36,940 --> 01:10:44,940
But that's I think an interesting... I mean, this is an observation, right?

802
01:10:44,940 --> 01:10:50,940
We've basically observed someone growing flour and it's got Paclobutrazola.

803
01:10:50,940 --> 01:10:54,940
And once again, not pointing any fingers is possible.

804
01:10:54,940 --> 01:10:57,940
It's not their fault. And it's just background.

805
01:10:57,940 --> 01:11:03,940
And then once again, it easily could not be the processor's fault, right?

806
01:11:03,940 --> 01:11:13,940
We could maybe look at the supply chain and say, like Ruth Conjecture, maybe their source flour passed.

807
01:11:13,940 --> 01:11:17,940
And then they just failed their concentrate testing.

808
01:11:17,940 --> 01:11:23,940
And once again, these companies may have already figured this out themselves.

809
01:11:23,940 --> 01:11:31,940
So maybe MFused already figured out why they failed or actually they passed in this case.

810
01:11:31,940 --> 01:11:41,940
But maybe some of these processors like Venns Park, maybe they've already pinpointed, oh, this is the source of our contaminant.

811
01:11:41,940 --> 01:11:45,940
This is why it failed.

812
01:11:45,940 --> 01:11:49,940
But who knows? They may not have figured this out yet.

813
01:11:49,940 --> 01:11:59,940
This is why we like to think that everybody's looking at the data and they know as much as we do.

814
01:11:59,940 --> 01:12:08,940
But a lot of this is untrodden ground.

815
01:12:08,940 --> 01:12:21,940
Technically, the Cannabis Commission should be doing these analyses, but my experience has been that the regulators aren't doing any enforcement of much of anything.

816
01:12:21,940 --> 01:12:29,940
And some of the states are trying to shut down some of the illegal either growers or dispensaries.

817
01:12:29,940 --> 01:12:36,940
But as far as enforcing the regulations like the lab tests and stuff, I don't know how much is going on.

818
01:12:36,940 --> 01:12:39,940
Well, then, too, you're talking about the United States government, right?

819
01:12:39,940 --> 01:12:46,940
You're not talking about a high tech company or even a high tech organization like Cannabis.

820
01:12:46,940 --> 01:12:51,940
But they want to, in Massachusetts, they want to start their own data center.

821
01:12:51,940 --> 01:12:54,940
They want to hire all these data scientists and software engineers.

822
01:12:54,940 --> 01:12:58,940
But you know what? It's the government. It's bound to fail.

823
01:12:58,940 --> 01:13:07,940
And, you know, again, Keegan's really onto something here, Keegan, because, you know, again, if the, what is it, packlobu trazol,

824
01:13:07,940 --> 01:13:15,940
if the boiling temperature is higher than THC, then again, we're going to run into high numbers building up and concentrates, right?

825
01:13:15,940 --> 01:13:20,940
And wow, did you nail it, too, by finding that it's on the do not use list.

826
01:13:20,940 --> 01:13:24,940
And I don't know. I think this is really good data.

827
01:13:24,940 --> 01:13:30,940
And it's interesting, too, because now we have another death in the cannabis industry from trimmer, another trimmer death.

828
01:13:30,940 --> 01:13:34,940
And then they're trying to like blame it on the woman being 60.

829
01:13:34,940 --> 01:13:40,940
But you know what? 60 is a new 40, baby. And, you know, 60 is not elderly.

830
01:13:40,940 --> 01:13:47,940
But she had COPD. Yeah. But you know what? I'm 62. I used to have a pro air inhaler, too.

831
01:13:47,940 --> 01:13:55,940
And, you know, if they had ventilation and that's where those were, because I go to Summit Lounge in Worcester, Massachusetts.

832
01:13:55,940 --> 01:14:00,940
And it's not like I have to even wash my clothes when I get home. The ventilation is not good.

833
01:14:00,940 --> 01:14:07,940
Whereas if I were to go to a bar to watch a Patriots game, my clothes would be disgusting with smoke and I would have to go right into the wash.

834
01:14:07,940 --> 01:14:12,940
Right. That's how good ventilation is. And that's a Worcester, Massachusetts social lounge.

835
01:14:12,940 --> 01:14:22,940
But this is really good stuff. And, you know, since these deaths and the reason I bring it up is because Keegan Florida truly started it.

836
01:14:22,940 --> 01:14:26,940
They're no longer publishing absolute values for pesticides and additives.

837
01:14:26,940 --> 01:14:39,940
And that's since the deaths. I mean, you know, like we can really go and look at it and see like when the death was and then look at when the COAs and compliance certificates changed

838
01:14:39,940 --> 01:14:47,940
to no longer provide absolute values. Right. Because it looks to me, you know, we'll need to get another set of eyes on it.

839
01:14:47,940 --> 01:14:56,940
But it appears to me that they really stopped after Lorna McMurray died. But, again, you know.

840
01:14:56,940 --> 01:15:11,940
I'll just chime in and say the glass half full approaches a never that saying once again, it's a it's a it's a brutal saying.

841
01:15:11,940 --> 01:15:18,940
But it's the brutal saying is never chalk up to malice where you can chalk up to stupidity.

842
01:15:18,940 --> 01:15:31,940
Or in this case, a lack, lack, lack of knowledge. That may be a better way to put it, because it's instead of saying, oh, the say the Washington state trace it.

843
01:15:31,940 --> 01:15:45,940
The Washington state regulators are malicious. Like Candice said, they may just be short of actually they are Washington State.

844
01:15:45,940 --> 01:15:52,940
I think they're malicious because actually Washington State has it all going on by offering open data sets.

845
01:15:52,940 --> 01:16:06,940
I guess maybe I might use this and then the other pesticide, too, that had to build up a concentration concentrates and really try to get Massachusetts to just start providing additives used to patients that purchase state product.

846
01:16:06,940 --> 01:16:15,940
But, you know, I wanted to be safe for patients and workers. But I agree 100 percent.

847
01:16:15,940 --> 01:16:24,940
Basically, you're 100 percent right. I think both of you in that this is a very serious topic and subject.

848
01:16:24,940 --> 01:16:33,940
And that's why we're kind of researching this. Right. We're at the frontier in there's knowledge to be discovered.

849
01:16:33,940 --> 01:16:43,940
Washington State, right. One could argue that, you know, maybe their their regulators should be pawing through this data.

850
01:16:43,940 --> 01:16:51,940
But the governments move slow and, you know, they're they're currently hiring a data consultant.

851
01:16:51,940 --> 01:17:08,940
And, you know, we're just maybe a little headed the curve. We've already got the data set. We've already we've already augmented it and standardized it and have begun analyzing it.

852
01:17:08,940 --> 01:17:13,940
You know, a similar thing in Massachusetts.

853
01:17:13,940 --> 01:17:33,940
Maybe they're just not excited about the public seeing the data because maybe they haven't looked at it themselves. Maybe they're like they have said publicly, Marianne McNeese and that other gal from the one of the other labs in Massachusetts, not the ones that offer.

854
01:17:33,940 --> 01:17:44,940
Coa's right. They were essentially said the public is too stupid to understand the data. And that is just because, you know, they are, you know, the Massachusetts CCC.

855
01:17:44,940 --> 01:17:50,940
They've been I've been seeing the job ads and I really don't know how they expect with the low pay.

856
01:17:50,940 --> 01:17:57,940
It's not even like a true government job either that they're going to get any data scientists worth their salt working for them.

857
01:17:57,940 --> 01:18:09,940
So, you know, all you have to do is just be open with the data. And there's so many people in Massachusetts to do understand data science and high tech that are all raring to go to look at this data, too.

858
01:18:09,940 --> 01:18:15,940
But they choose to keep it close. You know, it's just I don't know. I think it's a way of that.

859
01:18:15,940 --> 01:18:32,940
Everything about cannabis for as far as the regulators goes about control and this is in general, the government has approached cannabis from we're going to if we're going to do this, we're going kicking and screaming and we're going to control everything as much as possible.

860
01:18:32,940 --> 01:18:35,940
So I'm not surprised by any of this.

861
01:18:35,940 --> 01:18:56,940
Well, I've already been telling patients to that you really need to start growing your own or, you know, start finding someone you trust because honestly, the you know, since I've been back in Massachusetts smoking my homegrown and I have no longer been smoking Florida State weed because I'm a snowbird and I have not bought even though I renewed my card.

862
01:18:56,940 --> 01:19:10,940
I have $250 worth of credits. I don't even think I want to buy Massachusetts weed. And the reason why is because my lymph nodes after months of reaction doing Massachusetts cannabis and then Florida cannabis.

863
01:19:10,940 --> 01:19:19,940
I come back here to homegrown that's over a year old, but you know, it's got humidity packs. So it the buds are like I just pulled them on. I just dried them a few weeks ago. Right.

864
01:19:19,940 --> 01:19:30,940
And but I don't really think it's safe. And you know what? It's like I need to I might just start up like a medical marijuana patient group of my own up here in Massachusetts.

865
01:19:30,940 --> 01:19:41,940
And really, because that's who we have to get talking about it, you know, but I don't know. It just doesn't seem right to me because they are that it doesn't feel safe product and it makes me sick when I do it.

866
01:19:41,940 --> 01:19:44,940
So my homegrown doesn't.

867
01:19:44,940 --> 01:19:48,940
I love that you're taking the initiative.

868
01:19:48,940 --> 01:20:06,940
You may want to get in touch with maybe send them a message. The people who put together the Florida medical trees, you could get right when you could join their group and to see, hey, can we start like a chapter in Massachusetts or a similar initiative in Massachusetts?

869
01:20:06,940 --> 01:20:21,940
Because this is what we were talking about last week is, you know, power to the people. Right. In fact, especially now that we've got these large language models, I think it's the opposite. Right. I think it's the public are now the experts.

870
01:20:21,940 --> 01:20:32,940
And so, you know, the public can now fill in the regulators, instead of vice versa.

871
01:20:32,940 --> 01:20:43,940
And the other thing is, I think we're heading in the right direction. Right. And this is one of the reasons why we host this group is.

872
01:20:43,940 --> 01:20:50,940
Even if we can, you know, add one molecule to the picture, you know, today it was.

873
01:20:50,940 --> 01:21:10,940
It was a result, then I think we're moving things in the right direction because, as I said, there's been some talk about this. It's been super informal that just in forums that, oh, you know, you know, people are using plant growth regulators to get dense buds.

874
01:21:10,940 --> 01:21:22,940
And so it makes them harsh. And, oh, you know, some people say that, oh, you know, when they say smoke cannabis flower, they can tell if a plant growth regulator was used.

875
01:21:22,940 --> 01:21:40,940
And so it's not impossible that I mean, once again, I'm just conjecturing. We don't know. But it's not impossible that a large percentage of Florida flower is being treated with plant growth regulators.

876
01:21:40,940 --> 01:21:55,940
And the Wikipedia talk about power of the people. Right. Wikipedia is just a database of human knowledge.

877
01:21:55,940 --> 01:22:10,940
I have a can of Wikipedia. I think there is. I'll have to get some links, but the point being is.

878
01:22:10,940 --> 01:22:15,940
What is my point here? The power to the people.

879
01:22:15,940 --> 01:22:29,940
Well, the pesticides. So does Massachusetts. See, that's the thing, too. See, like at least Florida was telling us the pesticides that they had in their product, you know, up until like around when Lorna died.

880
01:22:29,940 --> 01:22:40,940
And but Massachusetts doesn't provide anything except for like a couple of hokey charts, you know, on their website. And I don't know.

881
01:22:40,940 --> 01:22:48,940
You know, OK, I think basically what I was.

882
01:22:48,940 --> 01:22:54,940
Yes. Just check. OK, I'll just put together my final thoughts and then we can wrap this up.

883
01:22:54,940 --> 01:23:00,940
But the Wikipedia on the pack would be resolved. Right. So chat GPT didn't know yet.

884
01:23:00,940 --> 01:23:11,940
But apparently Wikipedia was saying the boiling point was around four hundred and sixty degrees Celsius. So once again, that's just crowdsourced knowledge from the people.

885
01:23:11,940 --> 01:23:16,940
But once again, that would be too high to boil off for concentrates.

886
01:23:16,940 --> 01:23:24,940
And then, you know, if you do further reading about the substance that, oh, you know, people are spraying it on golf courses.

887
01:23:24,940 --> 01:23:34,940
So it seems to maybe be some sort of. Plants must love it because it seems that plants grow really dense with it.

888
01:23:34,940 --> 01:23:41,940
But I think maybe the research is out about maybe how healthy that is for humans to consume.

889
01:23:41,940 --> 01:23:52,940
But especially medical patients. Right. Because, you know, I don't live on a golf course because also, too, I know that Roundup causes Parkinson's disease is a huge lawsuit over it.

890
01:23:52,940 --> 01:24:00,940
Right. And and then because it's medical. So, you know, I have doctors telling me, you know, start looking for organic, non GMO food.

891
01:24:00,940 --> 01:24:08,940
Be careful what you know you are, what you eat. Even fasting for extended fasting can help, you know, with cancer and and other things, too.

892
01:24:08,940 --> 01:24:13,940
You know, it's just see, it'd be one thing if it were maybe just recreational. Right.

893
01:24:13,940 --> 01:24:18,940
But this is cannabis for medical patients. This is cannabis, too, for cancer patients. Right.

894
01:24:18,940 --> 01:24:23,900
And then you know, I think it's a little bit of a

895
01:24:23,940 --> 01:24:27,940
challenge to have a patient who's got 60 grams in 90 days.

896
01:24:27,940 --> 01:24:31,940
You know, they just really it burns me a bit.

897
01:24:31,940 --> 01:24:36,940
It does. It just burns me a bit, Keegan, that that the state is just ruining it.

898
01:24:36,940 --> 01:24:42,940
It might be hurting patients, you know, all because they want to hoard the data and not share it.

899
01:24:42,940 --> 01:24:50,940
And so you basically raise the point that, oh, you know, you could grow your own and then you could at least control what's in it.

900
01:24:50,940 --> 01:24:58,940
And I think that's that's kind of what it comes down to is you want to know what you're consuming.

901
01:24:58,940 --> 01:25:01,940
Right. We talk, right. You want to measure what you're consuming.

902
01:25:01,940 --> 01:25:08,940
And so if you grow it your own, you can make sure that no plant growth regulators were applied.

903
01:25:08,940 --> 01:25:12,940
But I guess the problem is maybe not everybody can grow their own.

904
01:25:12,940 --> 01:25:15,940
And so exactly. And patients can't grow their own.

905
01:25:15,940 --> 01:25:19,940
I would never have been able to grow my own when I first started it.

906
01:25:19,940 --> 01:25:23,940
I didn't do it until I was 57. And it wasn't until I got better. Right.

907
01:25:23,940 --> 01:25:28,940
That I could actually grow. And but the thing is, it's patient.

908
01:25:28,940 --> 01:25:33,940
So, you know, I'm going to and I can't get this data out of Massachusetts. I couldn't last year.

909
01:25:33,940 --> 01:25:44,940
But this year, because I did renew my card and because I want to be a patient because I'm going to ask purely, I'm going to ask all these MSOs in Massachusetts because, you know, I asked them last year and they blew me off.

910
01:25:44,940 --> 01:25:47,940
Nobody will tell me if the product is even radiated. Right.

911
01:25:47,940 --> 01:25:57,940
So now I'm going to ask what like, do you have anything that's somewhat organic? Right. Or that's organic or that is not radiated.

912
01:25:57,940 --> 01:26:01,940
That might be my first question to everybody. And then I'm going to make a spreadsheet.

913
01:26:01,940 --> 01:26:07,940
There are companies that do not offer any unreaded, un-remediated product. Right.

914
01:26:07,940 --> 01:26:12,940
And then try to find the few that do. And then maybe go somewhere with that with patients.

915
01:26:12,940 --> 01:26:22,940
And but, you know, it's just see when I became a patient. Right. It's like Keegan, they were like to me, they're like, oh, this is so good for you. It's so healthy.

916
01:26:22,940 --> 01:26:28,940
And, you know, when I was a patient, I didn't realize that they were just they're loading it up with chemicals.

917
01:26:28,940 --> 01:26:36,940
They're, you know, they're radiating, you know, probably powdery mildewed weed is getting radiated so it can pass tests.

918
01:26:36,940 --> 01:26:43,940
And then it's been we are us medical patients are smoking it. It's just it just doesn't seem right to me.

919
01:26:43,940 --> 01:26:50,940
It just doesn't. I get the jar half full.

920
01:26:50,940 --> 01:26:58,940
A view of this is all that Washington state data.

921
01:26:58,940 --> 01:27:04,940
It didn't look like there was a lot of culprits for this pack of butyra's.

922
01:27:04,940 --> 01:27:10,940
All right. Right. We saw there was right. That one farm that was really using a lot of it.

923
01:27:10,940 --> 01:27:20,940
And I've pinpointed a few farms that I once again, actually, we do have pretty good data here in Washington.

924
01:27:20,940 --> 01:27:24,940
But once again, I guess you don't know everything in their environment.

925
01:27:24,940 --> 01:27:30,940
But, you know, at least you can see, OK, no detections for pesticides and so on and so forth.

926
01:27:30,940 --> 01:27:35,940
But I think that's what you got. Every consumer cares about this.

927
01:27:35,940 --> 01:27:48,940
But like you said, try to find some licensees that you think are growing in a healthy manner or in a way that you approve of and stick to those.

928
01:27:48,940 --> 01:28:03,940
And like I would my point earlier is you can either grow it yourself or I think the states need to be a bit more open with their data because they because if you grow it yourself, you can control what's in it.

929
01:28:03,940 --> 01:28:19,940
Or if you have to do rely on, say, a medical provider like a medical licensed grower or just like in Washington state, just just generally licensed grower.

930
01:28:19,940 --> 01:28:28,940
If you do have to rely on them, then the states shouldn't be pretty open with their data that way.

931
01:28:28,940 --> 01:28:33,940
Right. Either they should do both or one or the other. Right.

932
01:28:33,940 --> 01:28:44,940
So it's either allow people to have home grow or else let them see the data that way they can actually know what they're consuming.

933
01:28:44,940 --> 01:29:03,940
I think there's there's a larger point here, and that's that you two are both very well informed cannabis users. And my experience has been and I have a friend who's an advocate and has a huge following.

934
01:29:03,940 --> 01:29:20,940
And she tries to educate patients. And when you have people, especially older people coming in to cannabis as medical patients, they're thinking about medical cannabis the same way that they think about other medications.

935
01:29:20,940 --> 01:29:29,940
And they assume that it's safe. It's tested. It's completely transparent. They have no idea what they don't know.

936
01:29:29,940 --> 01:29:38,940
And we can in the industry, they say, well, you need to be safe. You need to make sure that there's no pesticides. However, that's kind of a map and abstraction.

937
01:29:38,940 --> 01:30:00,940
Now that we actually have the data, we can go out there and we can tell people there's evidence that the products have pesticides in them that theoretically they're supposed to be pesticide free because that's what the regulators state that has to be tested.

938
01:30:00,940 --> 01:30:17,940
But there's actual data showing that there's pesticides. You need to make sure that there's no pesticides. You need to ask for your COAs to make and look specifically at the line items, you know, where it says pesticides, because they have no idea how to do this.

939
01:30:17,940 --> 01:30:26,940
They have no idea of any of these concepts or what's going on. They I mean, you are so far ahead of the average consumer.

940
01:30:26,940 --> 01:30:33,940
Well, actually, we get a lot of information from average consumers. You know, it's Reddit medical groves is what they do.

941
01:30:33,940 --> 01:30:42,940
They just turned us on to 18,000 COAs that we could spray publicly right in Florida. I'm telling you, medical patients are pretty savvy.

942
01:30:42,940 --> 01:30:51,940
But then you get the CCC coming to the Worcester Senior Center and telling everybody how healthy state weed is, right? Because they're not giving full disclosure.

943
01:30:51,940 --> 01:31:00,940
I hear what you say. And I know that there's a lot of people out there. And the problem is, is they've been forced to be very proactive.

944
01:31:00,940 --> 01:31:07,940
And the knowledgeable consumers are at the front of the curve trying to push for everyone else.

945
01:31:07,940 --> 01:31:18,940
But if you look at the masses of people behind them who still don't know what they don't know, I would I think it's there's so many amazing people in cannabis,

946
01:31:18,940 --> 01:31:25,940
just like you, who are savvy, who are standing up for all of the other patients.

947
01:31:25,940 --> 01:31:31,940
But again, in my experience, you are not typical.

948
01:31:31,940 --> 01:31:44,940
That's all I'm going to say. I agree with both of you in that probably the people in the forums in say medical trees on Reddit are also kind of probably people like us,

949
01:31:44,940 --> 01:31:51,940
probably the cannabis experts. And I always say we're at the frontier.

950
01:31:51,940 --> 01:32:02,940
And sometimes it may seem a little scary when you see a bear and you're on the frontier. But that's to be expected.

951
01:32:02,940 --> 01:32:21,940
And so one, I think there is there's a it's a it's a long road uphill ahead of us. But I'm willing to keep trudging up, keep trudging uphill.

952
01:32:21,940 --> 01:32:33,940
Maybe I should go on YouTube for senior 55 plus medical patients, right. And then maybe do it, make a presentation and do a little roadshow down to the Worcester Senior Center after the CCC came through.

953
01:32:33,940 --> 01:32:38,940
It said how healthy marijuana will be for you compared to big pharma drugs. Right.

954
01:32:38,940 --> 01:32:45,940
We are you know, there's so many black box warnings on those big pharma drugs that they keep pushing down seniors throats.

955
01:32:45,940 --> 01:32:54,940
And, you know, it just the medications and telling them to take them.

956
01:32:54,940 --> 01:33:05,940
And really, the average person, I and I'm not trying to be critical, but I don't think you appreciate how how educated you are.

957
01:33:05,940 --> 01:33:16,940
I'm educated too because I'm a patient. Right. And I replace that. Yes. That's at a band. My brother is an MS patient and he actually had a stem cell transplant.

958
01:33:16,940 --> 01:33:24,940
And he's been he knows more than any of his doctors because he cares more about his health than his doctors do.

959
01:33:24,940 --> 01:33:32,940
And he educates other people on MS and the treatments and the side effects and everything else. He is not typical.

960
01:33:32,940 --> 01:33:40,940
How does he do it? Does he have like a YouTube video or you know, because that's a website where he's documented his whole journey.

961
01:33:40,940 --> 01:33:49,940
But how does he share it with other patients? He is on. He has a website, as I said, he's also active.

962
01:33:49,940 --> 01:33:58,940
He's in California and there's the Central Coast MS Society and he's active in that.

963
01:33:58,940 --> 01:34:04,940
Yeah. And yeah, but I'm telling you, he's not the typical person.

964
01:34:04,940 --> 01:34:14,940
And he also he's one of the many who actually took charge of his own care because he's not just being a passive member, you know.

965
01:34:14,940 --> 01:34:21,940
But most people, as I keep saying, I so many people, they don't know any better. They're very intimidated.

966
01:34:21,940 --> 01:34:26,940
My dad was a doctor and I've seen how people respond to doctors. They treat them like gods.

967
01:34:26,940 --> 01:34:36,940
And they don't question them. And and this is again, most people, they're very passive consumers of health care.

968
01:34:36,940 --> 01:34:43,940
And you are you come across as being so articulate and so knowledgeable.

969
01:34:43,940 --> 01:34:49,940
And that is not my experience with the average person. And I just I just want to put that out there for you.

970
01:34:49,940 --> 01:34:58,940
Well, like your brother, right. So I'm on Social Security Disability for Parkinson's disease and there is actually some family history.

971
01:34:58,940 --> 01:35:10,940
So, you know, maybe I'm in early stages. But, you know, I tend to think that it was just that the Mass General Hospital went a little too stir crazy, prescribing me on way too many drugs that I never should have had.

972
01:35:10,940 --> 01:35:15,940
And I think that's what contributed to the tremors. And because I was pretty bad. I mean, I couldn't drive and stuff, too.

973
01:35:15,940 --> 01:35:25,940
And cannabis is amazing, like, you know, so the six milligrams of Ativan, you know how I got off of that, like I was on that for decades. Right.

974
01:35:25,940 --> 01:35:33,940
And and I was cold cut. But, you know, garden remedies came over with RSO. So I was doing a lot of RSO.

975
01:35:33,940 --> 01:35:41,940
And that helped me get through that addiction. And, you know, Percocets, that was, you know, after decades of daily use, three times three X a day.

976
01:35:41,940 --> 01:35:54,940
That was the easy drug to kick. But the Ativan, six milligrams a day and the Ativan, I mean, the Ambien, the sleeping pill at night, when they cold cut me from that, the doctors, I was I was not well.

977
01:35:54,940 --> 01:36:05,940
And that RSO was a miracle drug as far as like I can understand how people get addicted because you don't think of it. You know, your doctor's like, oh, yeah, take this. Oh, yeah, they have no harm.

978
01:36:05,940 --> 01:36:19,940
It won't cause dementia. It won't, you know. But I think that I see I've met a lot of people, too, that, you know, that instead of doing something really crazy, because, you know, my doctors were recommending deep brain stimulation surgery.

979
01:36:19,940 --> 01:36:29,940
And one of my other friends with Parkinson, she had it done. And and but I didn't, you know, and I might still. Yeah, I still need medical marijuana. Right.

980
01:36:29,940 --> 01:36:39,940
But I can like dose so that, you know, it's not like I'm stoned all the time and I still can't really do this at Eva like he likes his fever. Right.

981
01:36:39,940 --> 01:36:46,940
And and I should to help with creativity and stuff, too. But I don't know. It's been a crazy journey. But you know what?

982
01:36:46,940 --> 01:36:51,940
I'm wondering, like with what your brother's doing, maybe I should maybe do the same thing, you know, maybe not like me.

983
01:36:51,940 --> 01:37:00,940
Like, oh, my goodness, I'm 62. I'm all wrinkled. I can't do YouTube. But maybe that's what I need to do to to just try to help people understand because they do.

984
01:37:00,940 --> 01:37:08,940
And they need to ask for COAs because, you know, I thought that the medical marijuana was just some beautiful legacy miracle. Right.

985
01:37:08,940 --> 01:37:16,940
With people that really were passionate and cared about humanity. But, you know, I discovered, you know, and then I started late in life. Right. I'm 62.

986
01:37:16,940 --> 01:37:24,940
I started at 57. And although I never, ever smoked pot when I worked the 20 years I worked the Dell EMC or digital equipment, never, never.

987
01:37:24,940 --> 01:37:32,940
And I don't know. I'm so disappointed because it just seems like people are just rushing and corner cutting the state weed.

988
01:37:32,940 --> 01:37:44,940
They don't care about the health of their workers or the patients, whereas the people I have met in the legacy groups, because I didn't know anybody that's sold pot or anything except for the state.

989
01:37:44,940 --> 01:37:49,940
But, you know, I went to some legacy parties here in Massachusetts, some gorgeous houses, too.

990
01:37:49,940 --> 01:37:57,940
And, you know, sometimes you would just kind of like pay for a ticket and then they might have raffles. Right. But it wasn't like about selling pot. It was about education.

991
01:37:57,940 --> 01:38:05,940
And I don't know. You know, I'm just so disappointed with the state, you know, compared to when I first started five years ago.

992
01:38:05,940 --> 01:38:23,940
It's like that in every state. And I think that the state administrators are all corrupt. And if you look at, I've done a lot of writing and speaking about how rec use is squeezing out medical use, medical has just gone down the tubes.

993
01:38:23,940 --> 01:38:33,940
Ruth, are you on LinkedIn? I am. I'll try to find you and I'll send you a, do you want to connect on LinkedIn? Absolutely.

994
01:38:33,940 --> 01:38:43,940
That would be great. I'd love to see, I'd love to read. I actually just got a post because I know I've been involved in the education, patient education.

995
01:38:43,940 --> 01:38:50,940
And I've talked to so many people who are trying to work with patients to educate them and help guide them in their journey.

996
01:38:50,940 --> 01:38:57,940
And this is when I came to cannabis and my brother, I came because of my brother, he found the cannabis worked. We started using it.

997
01:38:57,940 --> 01:39:04,940
He's like, let's develop a technology to help other people do this. So we started developing a tracking technology.

998
01:39:04,940 --> 01:39:20,940
We hit a wall largely because the COA data weren't available and we needed that. But, shoot, where is I going with this?

999
01:39:20,940 --> 01:39:30,940
Oh, so in the meantime, though, so that, that project kind of got tabled, but I've been working with so many people on this education aspect and also the tracking.

1000
01:39:30,940 --> 01:39:41,940
And I think trying to find good product is essential, you know, and the right product, it's essential. And information needs to be readily accessible and transparent.

1001
01:39:41,940 --> 01:39:50,940
And, OK, so as I said, most patients don't know what they're doing. And I've talked to a lot of people who are setting up organizations to guide patients.

1002
01:39:50,940 --> 01:39:58,940
And that just doesn't work because there's no mechanism out there that will reimburse these people to educate the patients.

1003
01:39:58,940 --> 01:40:05,940
And I just posted an analysis explaining why you can't get paid to help patients.

1004
01:40:05,940 --> 01:40:13,940
And so everyone, many groups have tried this and they've all left the market because they can't get reimbursed for their services.

1005
01:40:13,940 --> 01:40:24,940
And this is a huge problem. I forget where I was going with this, but I also I have a very good friend who's a TBI survivor.

1006
01:40:24,940 --> 01:40:26,940
Yeah, brain injury. I know that one.

1007
01:40:26,940 --> 01:40:40,940
She was a pediatric nurse and she's 50s. And she said before I came to Canvas, if you had told me as a nurse that you were you were using cannabis or giving it to your kids,

1008
01:40:40,940 --> 01:40:48,940
I would have called the Department of Child Protective Services on you. She had her injury.

1009
01:40:48,940 --> 01:40:57,940
She went from being a nurse and a provider in the system to being a patient. And that just completely transformed her.

1010
01:40:57,940 --> 01:41:12,940
She was called a drug seeker. She was she was on dozens and dozens of medications before she came to Cannabis and ended up getting off of most of her for her medications.

1011
01:41:12,940 --> 01:41:18,940
And she is pretty much completely self-taught. She's in Buffalo.

1012
01:41:18,940 --> 01:41:24,940
I thought you were going to say Nikki Lawley because she's amazing. That's how you're talking about.

1013
01:41:24,940 --> 01:41:26,940
That's what I'm talking about. She's a close friend of mine.

1014
01:41:26,940 --> 01:41:30,940
Oh, that's who you are talking about, Nikki? She's one of my contacts. She's wonderful.

1015
01:41:30,940 --> 01:41:44,940
She started she's in New York. She couldn't get good product. She said she's tried hundreds and hundreds, probably over 500 different strains to find trying to find products that work.

1016
01:41:44,940 --> 01:41:50,940
She got a bunch. She's like, I have this product that works. I'm like, what's in it? She's like, I have no idea. I bought it off the black market.

1017
01:41:50,940 --> 01:41:54,940
I'm like, send it to a lab. She's like, there are no labs. I said, let me look for you.

1018
01:41:54,940 --> 01:42:02,940
And I looked. There were no labs where she could send her product to. So here she has no idea what's working for her.

1019
01:42:02,940 --> 01:42:09,940
Now, during COVID, regularly she goes into Canada because she gets better product and better information there.

1020
01:42:09,940 --> 01:42:15,940
She goes to Canada regularly. She's right on the border. COVID came and shut down the border.

1021
01:42:15,940 --> 01:42:23,940
For several years, she couldn't get access to good product. So she was stuck using black market product and she had no idea what was in it.

1022
01:42:23,940 --> 01:42:33,940
And she is out there on the front lines trying to get education, get access, get transparency.

1023
01:42:33,940 --> 01:42:42,940
And I asked her, I said, are you advocating because patients want transparency and they want this information?

1024
01:42:42,940 --> 01:42:52,940
I said, or are you asking for it on their behalf? And she said it's the latter. She said no one has any idea about asking for any of this stuff.

1025
01:42:52,940 --> 01:42:58,940
And she's in Americans for Safe Access, which is, again, trying to spearhead all of this.

1026
01:42:58,940 --> 01:43:08,940
But I talked, you know, I went to a, I went to a conference here in San Francisco with psychedelic people and they were all researchers.

1027
01:43:08,940 --> 01:43:17,940
They were all pharma people, essentially. And I said, yeah, I'm working on this tracking and the transparency issue, making sure that the information is available.

1028
01:43:17,940 --> 01:43:23,940
And they're total blank looks. I'm like, you understand what I'm referring to, right? They're like, no.

1029
01:43:23,940 --> 01:43:32,940
I said, patients have no idea what's in their product. And they gave me this, like this look like it just didn't click with them.

1030
01:43:32,940 --> 01:43:39,940
So no one in the system, no one in the traditional system understands any of these problems.

1031
01:43:39,940 --> 01:43:49,940
And I regularly say, here are all the problems, just basic issues that exist in cannabis that you don't have in the regular health care industry.

1032
01:43:49,940 --> 01:43:59,940
And they're not aware of any of it. They're not aware of any of it. The transparency, the fact that no one knows what's in their product, the fact that the doctor says,

1033
01:43:59,940 --> 01:44:12,940
OK, you need a product with THC, CBD, CBG, and you want some limonene and, you know, terpenolene, and they go to the dispensary and the patient's like, yeah, this is what my doctor told me to get.

1034
01:44:12,940 --> 01:44:21,940
And the budtender's like, yeah, well, we don't have that. Here, take this product and they go home with a couple hundred dollars worth of product that's shipped for them.

1035
01:44:21,940 --> 01:44:35,940
It won't work. It won't do anything. The doctor has no understanding that what he's telling the patient to get, the patient's not going to find because a doctor, when he prescribes the medication, the patient goes to the pharmacy.

1036
01:44:35,940 --> 01:44:42,940
It's all standardized. And he knows the patient's going to walk out with exactly the medication the doctor told the patient to get.

1037
01:44:42,940 --> 01:44:54,940
In fact, the pharmacist isn't allowed to give the patient something different than what was prescribed. That doesn't exist in cannabis.

1038
01:44:54,940 --> 01:45:03,940
Yeah, but cannabis is a plant, too. You know, it's not like a petroleum-based pharmaceutical drug either.

1039
01:45:03,940 --> 01:45:25,940
And then again, too, with the plants, like, you know, Keegan, you know, it's like there have been, I've seen other people talk passionately on YouTube about how strains will differ, how you can actually get like COAs for Blue Dream, but you'll get like a different chemical analysis across producers.

1040
01:45:25,940 --> 01:45:37,940
And then remember, Keegan, we had that wonderful gal with epilepsy that told us that, you know, she bought a strain. It worked great. Then she bought the same strain again, and it was completely different.

1041
01:45:37,940 --> 01:45:52,940
And, you know, so I would love to see like instead of a periodic table with just the strain names on it, right, actually more with the chemical balance, you know, and then or make sure that a particular like, you know, Keegan was doing patent, right?

1042
01:45:52,940 --> 01:46:01,940
But, you know, if we and then John, you know, with his principal component analysis, he had like, you know, fingerprints for all the strains, right?

1043
01:46:01,940 --> 01:46:14,940
So then you can like kind of like look at a Blue Dream and say, well, does it, you know, and see if the principal component analysis is the same with another one and then or actually match it up with a totally different strain and say, oh, this, they're telling Blue Dream, but you know what?

1044
01:46:14,940 --> 01:46:18,940
It's really GG4. It's Gorilla Glue, right?

1045
01:46:18,940 --> 01:46:38,940
So we're missing a point here. And anywhere else in our economy, the way things work, if you go to the store and if you buy Cheerios, and then you go to another store somewhere else, you know, even like five years ago and you buy Cheerios, those products are the same thing.

1046
01:46:38,940 --> 01:46:45,940
In cannabis, anyone can call any product wedding cake. There is no standardization.

1047
01:46:45,940 --> 01:46:52,940
That's what I was saying, right? And even at the same time, though, that has pesticides in it, right?

1048
01:46:52,940 --> 01:47:02,940
But at the same time, you can take, we can take, and you and I can take the seeds from the same exact plant. You grow them in Massachusetts, I grow them in California.

1049
01:47:02,940 --> 01:47:10,940
And what we're going to get is going to be completely different because the soil is different, the environment is different, the climate is different.

1050
01:47:10,940 --> 01:47:22,940
Even the same string grown, you can go different microclimates. You can even go to different parts of the same plant and get different expressions.

1051
01:47:22,940 --> 01:47:33,940
And this is, again, it's, I'm not criticizing cannabis. What I'm saying is it's different from what you're used to in the way that you consume other types of products.

1052
01:47:33,940 --> 01:47:43,940
In the industry, it's trying to apply everything from other products into cannabis and the way they're thinking about cannabis and that fundamentally doesn't work.

1053
01:47:43,940 --> 01:47:56,940
They need to consider cannabis and how that works and design the rules around cannabis to get the outcomes they want because they're trying to shove cannabis into a pigeonhole that doesn't fit.

1054
01:47:56,940 --> 01:48:13,940
Well, it's a plant, you know, it's not a chemical and it's a plant. God gave man kind of cannabis whereas John D. Rockefeller gave mankind petroleum based, harmful pharmaceuticals, right?

1055
01:48:13,940 --> 01:48:22,940
It became a huge industry and you've only seen human health decline and lifespan decline because of the new Western pharmaceutical world.

1056
01:48:22,940 --> 01:48:26,940
I digress, sorry.

1057
01:48:26,940 --> 01:48:43,940
I first want to applaud you both and then I think I can tie all this together. First, I want to applaud you both for working diligently to spread knowledge because this is going to get super meta.

1058
01:48:43,940 --> 01:48:54,940
I think this is necessary. You've raised so many good points here and one was how the governments look corrupt. I think what does this all come down to?

1059
01:48:54,940 --> 01:49:04,940
I was thinking the other week, you know, what exactly, there's a great Monty Python movie, The Meaning of Life.

1060
01:49:04,940 --> 01:49:19,940
What exactly is the meaning of life? And I think some of the politicians or the regulators, they've got more to learn in life because maybe they think the meaning of life is just pursuing money.

1061
01:49:19,940 --> 01:49:36,940
So that's really that base. So that's what some people are doing. So they've got some more growing and learning to do in life and then you can say, oh, you know, maybe the purpose of life is the pursuit of happiness.

1062
01:49:36,940 --> 01:49:50,940
And so that would go to, you know, the medical patients, right? They just want to live a happy life, right? They don't want to be debilitated by a condition or they just want to get over that and get back to enjoying their life.

1063
01:49:50,940 --> 01:49:56,940
So you could argue maybe it's happiness. And then I was thinking, well, what are we doing here?

1064
01:49:56,940 --> 01:50:14,940
Maybe the meaning of life is the pursuit of knowledge. So that's kind of what I kept thinking about as you two were talking or debating this topic is what does this all come down to?

1065
01:50:14,940 --> 01:50:31,940
It's really just all of us trying to pursue knowledge. And I was thinking about this in if we knew everything, right? Like say there was actual perfect information like economists like this assume there is.

1066
01:50:31,940 --> 01:50:49,940
And the world wouldn't be any fun, right? It wouldn't be any fun if everybody knew everything about everything. And there would be nothing left to learn. And so we live in an imperfect world where knowledge is super limited.

1067
01:50:49,940 --> 01:51:03,940
We all have limited knowledge base. And let's face it, different people are knowledgeable about different things. Maybe there's a carpenter who knows a ton about carpentry.

1068
01:51:03,940 --> 01:51:27,940
Maybe they just don't have much knowledge of communication or cannabis or this or that. And so just for what just the way life took its turns, you know, Candice and myself and Ruth, we just happen to have lots of knowledge about data and cannabis and statistics and public health.

1069
01:51:27,940 --> 01:51:50,940
And, you know, what's the meaning of life? You know, maybe it's for us to continue pursuing our knowledge. You know, maybe share it with others because it sure seems that, you know, the more of this knowledge we accumulate and the more that we spread it around to other people, it seems like they're able to live better lives.

1070
01:51:50,940 --> 01:52:04,940
It is a slow, incredibly slow process, right? And it's imperfect, right? We have to go through all these imperfections, right? We have to deal with the regulators who are pursuing money.

1071
01:52:04,940 --> 01:52:17,940
And we have to deal with sometimes the medical professions may have imperfect knowledge. Sometimes the bud tenders may have imperfect knowledge. The consumers often have imperfect knowledge.

1072
01:52:17,940 --> 01:52:28,940
So it's just the imperfect world we live in where nobody's got it right. We're, you know, we're all just trying to gather information and live better lives.

1073
01:52:28,940 --> 01:52:42,940
And that's why I kind of want to applaud you both, because you're both doing just that. Right, Ruth, right? You launched a company that's doing the traceability and tracking for consumers.

1074
01:52:42,940 --> 01:52:53,940
You're basically trying to provide people with knowledge in Candace. You said you were going to start an initiative to start educating medical patients in Massachusetts.

1075
01:52:53,940 --> 01:53:02,940
And you've already been helping enormously with the collection of knowledge by helping with the cannabis data science meetup.

1076
01:53:02,940 --> 01:53:16,940
So I think that's the glass half full is is everything perfect right now? Nowhere close. Is there a lot of improvement to be done? Yes.

1077
01:53:16,940 --> 01:53:25,940
Do consumers need more knowledge? Yes. Do medical providers need more knowledge? Yes. Do bud tenders need more knowledge? Yes.

1078
01:53:25,940 --> 01:53:33,940
Could everybody in the world benefit from learning a little bit more? Probably, myself included.

1079
01:53:33,940 --> 01:53:47,940
So I think we're I think we're on the right path. And I think it does help pointing out the imperfections, because if you don't spotlight them, then there's no way to fix them.

1080
01:53:47,940 --> 01:54:01,940
So I think I think it's great. Right. So, Ruth, right. It's awesome. Right. You can point out things like, you know, a lot of consumers have a worryingly little amount of knowledge about this.

1081
01:54:01,940 --> 01:54:11,940
And, you know, maybe people have the regulators have worryingly little knowledge about the medical side of cannabis.

1082
01:54:11,940 --> 01:54:21,940
So there's and then Candice is right. You know, you know, the consumers worryingly little access to data.

1083
01:54:21,940 --> 01:54:28,940
So, well, Ruth is right. You know, you're right, Ruth. It's like it must be because I'm I was born an engineer.

1084
01:54:28,940 --> 01:54:34,940
So I just have to know everything about everything. Like, you know, I know how to grow. I can make my own RSO and edibles.

1085
01:54:34,940 --> 01:54:41,940
And, you know, I just always you know, it's always curiosity. I'm so curious that, you know, I'm kind of driven to that.

1086
01:54:41,940 --> 01:54:48,940
But, you know, we are connected on LinkedIn and I am going to pay more attention, especially to your latest article.

1087
01:54:48,940 --> 01:54:54,940
And I'm going to look to and if I can't find anything about what your brother's doing, I might ask you if that's OK.

1088
01:54:54,940 --> 01:54:59,940
And because I really do want to learn, I really I'm real interested to and what you're putting out there.

1089
01:54:59,940 --> 01:55:04,940
And because, like you said, but I mean, I've just been talking about right.

1090
01:55:04,940 --> 01:55:08,940
Education patients, but I haven't done anything about it. I haven't done a YouTube video.

1091
01:55:08,940 --> 01:55:16,940
I haven't been doing any blogging. And, you know, I just think it's fabulous what your brother's doing, what you're doing.

1092
01:55:16,940 --> 01:55:22,940
You're participating in Reddit forums. I think that you're doing more than you realize.

1093
01:55:22,940 --> 01:55:29,940
I do more on Facebook. It's done everywhere. You're providing a lot of information that people really need.

1094
01:55:29,940 --> 01:55:34,940
And for that, you're absolutely to be applauded. Thanks.

1095
01:55:34,940 --> 01:55:41,940
I need to do YouTube video really, really stick my foot in the really jump into the stream.

1096
01:55:41,940 --> 01:55:45,940
I love this. I love the direction we're going in.

1097
01:55:45,940 --> 01:55:53,940
And I want to thank you both. So thank you both from the bottom of my heart for coming.

1098
01:55:53,940 --> 01:55:59,940
And we've now spent two hours talking about cannabis data and how to help people.

1099
01:55:59,940 --> 01:56:13,940
So I'll let you go on and enjoy your days. But I think this was an especially fruitful day today because we pinpointed just the current state of access to data, where we can go with it.

1100
01:56:13,940 --> 01:56:22,940
Some of the current pitfalls in the industry. And I think we've got a bright future ahead of us.

1101
01:56:22,940 --> 01:56:25,940
Thank you for putting this together very much.

1102
01:56:25,940 --> 01:56:32,940
Definitely. And as I said, even if we're only moving things a molecule at a time, we're moving forward.

1103
01:56:32,940 --> 01:56:34,940
Moving forward. Thank you.

1104
01:56:34,940 --> 01:56:36,940
Too cool, everyone. Thank you for coming.

1105
01:56:36,940 --> 01:56:44,940
Have a wonderful day.