1
00:00:00,000 --> 00:00:05,200
I did feed into chat GPT an icebreaker because why not?

2
00:00:05,200 --> 00:00:07,000
We're talking chat GPT and it asked,

3
00:00:07,000 --> 00:00:10,980
what is something unusual about yourself that someone may not know?

4
00:00:10,980 --> 00:00:13,480
Which I was challenged to go,

5
00:00:13,480 --> 00:00:15,480
nothing interesting.

6
00:00:15,480 --> 00:00:21,120
But then both of you kindly did remind me for this and that I have painted a lot of things.

7
00:00:21,120 --> 00:00:23,160
Previous episodes, you've seen the paintings behind me.

8
00:00:23,160 --> 00:00:25,240
It's just the random wall that I have in my house,

9
00:00:25,240 --> 00:00:29,040
but I have painted every single surface short of the carpets in my house.

10
00:00:29,040 --> 00:00:32,560
Tiles, cabinetry, countertops,

11
00:00:32,560 --> 00:00:36,600
walls, just because I can have done chalkboard walls.

12
00:00:36,600 --> 00:00:40,040
Because why not? It's paint. It's not permanent.

13
00:00:40,040 --> 00:00:41,960
Kelsey ran out of services,

14
00:00:41,960 --> 00:00:46,080
so she even came to the CIT office and wrote our core values in our conference room.

15
00:00:46,080 --> 00:00:49,160
So just can't paint enough.

16
00:00:49,160 --> 00:00:52,360
You had to break up the beige somehow.

17
00:00:52,360 --> 00:00:55,040
There's a lot of beige.

18
00:00:55,240 --> 00:00:57,600
We are doing renovation soon,

19
00:00:57,600 --> 00:00:59,560
but. It's true.

20
00:00:59,560 --> 00:01:02,160
I'll be painting on more walls coming soon.

21
00:01:02,160 --> 00:01:04,040
It'll be exciting. Then we'll record in front of that wall.

22
00:01:04,040 --> 00:01:07,560
We'll fly Matthew in just to sit in front of the wall.

23
00:01:07,560 --> 00:01:10,680
But what's something unusual about both of you?

24
00:01:10,680 --> 00:01:13,320
Unusual, interesting?

25
00:01:16,840 --> 00:01:24,920
It's very hard to know where to start given that I grew up in a different country.

26
00:01:24,920 --> 00:01:30,520
Things that I find normal, you guys might find weird right out of the gate.

27
00:01:30,520 --> 00:01:33,040
I don't know. I moved around a lot when I was growing up.

28
00:01:33,040 --> 00:01:40,600
I think I moved houses eight times in the first nine or 10 years growing up.

29
00:01:40,600 --> 00:01:44,520
So I lived all around Australia and then obviously moved a lot.

30
00:01:44,520 --> 00:01:51,280
I have four years is the most time I've ever spent in one place, one city ever.

31
00:01:51,280 --> 00:01:55,120
So I tends to be a little bit surprising to people.

32
00:01:55,120 --> 00:01:57,240
On one of our security meetings,

33
00:01:57,240 --> 00:01:59,720
I also asked, what would you not do for a million dollars?

34
00:01:59,720 --> 00:02:01,840
I said a venomous snake handler and he went,

35
00:02:01,840 --> 00:02:05,760
that's easy. There's another one.

36
00:02:05,760 --> 00:02:08,960
Just to clarify, we were talking about milking snakes.

37
00:02:08,960 --> 00:02:11,880
It was a very specific thing.

38
00:02:11,880 --> 00:02:14,920
There's another weird quote about.

39
00:02:14,920 --> 00:02:18,520
Yeah. It was still a moment where the rest of us were definitely like,

40
00:02:18,520 --> 00:02:23,400
yeah, we don't think it's casual to milk the venom from a snake.

41
00:02:23,400 --> 00:02:26,800
It's just not something that we're exposed to.

42
00:02:26,800 --> 00:02:30,280
There's weird places for everything in Australia.

43
00:02:30,280 --> 00:02:37,680
There's a special lizard sanctuary and snake sanctuary

44
00:02:37,680 --> 00:02:40,720
near where I spent a significant portion of my childhood.

45
00:02:40,720 --> 00:02:43,480
So I'll follow you after this because there's

46
00:02:43,480 --> 00:02:45,920
someone on our team that's definitely afraid of lizards.

47
00:02:45,920 --> 00:02:48,640
So I was going to say,

48
00:02:48,640 --> 00:02:51,600
do you have been a healthy relationship with lizards and snakes

49
00:02:51,600 --> 00:02:54,040
or are you at all freaked out by them?

50
00:02:54,040 --> 00:02:55,320
It doesn't really bother me.

51
00:02:55,320 --> 00:02:57,440
I don't like them around mostly because when I see them,

52
00:02:57,440 --> 00:02:58,960
I expect them to be venomous.

53
00:02:58,960 --> 00:03:02,800
But that's most of my fear.

54
00:03:02,800 --> 00:03:06,800
Yeah. What about you, Nick?

55
00:03:06,800 --> 00:03:12,000
I'd say, my family does a lot of jewelry making.

56
00:03:12,000 --> 00:03:18,680
So my dad does a lot of silversmithing and cabochon cutting,

57
00:03:18,680 --> 00:03:21,320
and then my grandpa did a lot of faceting.

58
00:03:21,320 --> 00:03:24,120
So we have a full jewelry room with

59
00:03:24,120 --> 00:03:27,440
all the tools and the faceting machine and everything like that.

60
00:03:27,440 --> 00:03:33,360
So I actually made some necklaces for the wedding and all that fun stuff.

61
00:03:33,360 --> 00:03:36,840
So yeah, really, really enjoy doing that.

62
00:03:36,840 --> 00:03:40,280
We did a lot of mining in other states and everything.

63
00:03:40,280 --> 00:03:42,840
So that's really cool. That's awesome.

64
00:03:42,840 --> 00:03:48,400
Yeah, I was going to say, what's your favorite thing you ever found mining?

65
00:03:49,400 --> 00:03:52,120
Probably some opals.

66
00:03:52,120 --> 00:03:54,640
There's some really, really nice fire opals.

67
00:03:54,640 --> 00:03:59,000
My family, we didn't find these ones, but at least in the dirt.

68
00:03:59,000 --> 00:04:04,000
But we have some Australian opal and Australian opal is just gorgeous.

69
00:04:04,000 --> 00:04:06,880
It's tons of blues and greens, and it's my favorite type of stone.

70
00:04:06,880 --> 00:04:10,720
So that's just because they're Australian.

71
00:04:10,720 --> 00:04:13,480
When you go visit,

72
00:04:13,480 --> 00:04:17,280
you've got to go see some of the open mines they have and the caves they have,

73
00:04:17,280 --> 00:04:19,680
where you can walk through and they didn't mine them all.

74
00:04:19,680 --> 00:04:21,920
They've left them in the rocks.

75
00:04:21,920 --> 00:04:25,400
So you can really appreciate what that's like underground with only the lights

76
00:04:25,400 --> 00:04:28,680
to guide you through the space too,

77
00:04:28,680 --> 00:04:32,040
unless you're a little bit agoraphobic and wouldn't appreciate that.

78
00:04:32,040 --> 00:04:35,080
I was just about to say, I don't go where the venomous snakes are.

79
00:04:35,080 --> 00:04:38,560
So that's that. That's that.

80
00:04:38,560 --> 00:04:40,000
Well, awesome. Thank you both.

81
00:04:40,000 --> 00:04:43,520
And thank you, Chad GPT, for giving us our icebreaker here for today,

82
00:04:43,520 --> 00:04:46,400
which is bringing us back on topic for what are we sitting down?

83
00:04:46,520 --> 00:04:48,840
Why are we here? What are we talking about?

84
00:04:48,840 --> 00:04:51,280
We're sitting down on our weekly tech for business podcast,

85
00:04:51,280 --> 00:04:54,160
talking about chat, GPT and cybersecurity.

86
00:04:54,400 --> 00:04:58,120
We've got our director of cybersecurity, Nate, and our GRC analyst, Matthew.

87
00:04:58,320 --> 00:05:01,360
And I'm Kelsey. I'm a member of our marketing team here to help moderate

88
00:05:01,360 --> 00:05:06,280
and average chat GPT users. I asked both of them to put on their chin hats

89
00:05:06,280 --> 00:05:10,120
and let me know where I might be selling my life online to machines.

90
00:05:10,480 --> 00:05:13,800
So to start off with both of you, what is chat GPT?

91
00:05:15,840 --> 00:05:18,240
It's a contract to allow you to sell your sold machines.

92
00:05:20,040 --> 00:05:21,440
Oh, cool.

93
00:05:22,240 --> 00:05:24,000
No, it's Nate.

94
00:05:24,000 --> 00:05:27,880
I feel the description that you gave when we spoke the other day was my favorite.

95
00:05:27,880 --> 00:05:30,320
So I have no idea what you're talking about.

96
00:05:30,320 --> 00:05:32,120
So I will wing it again.

97
00:05:32,120 --> 00:05:38,440
But chat GPT is a tool that was created by a research company.

98
00:05:38,800 --> 00:05:43,400
It's they've taken a lot of data, you know, leading up to 2021.

99
00:05:44,080 --> 00:05:45,560
There's no new data after that.

100
00:05:45,560 --> 00:05:47,320
But then they took all that data.

101
00:05:47,320 --> 00:05:53,320
They trained the AI or artificial intelligence on how to process this information.

102
00:05:53,600 --> 00:05:56,200
And then they made it now readily available.

103
00:05:56,200 --> 00:06:00,720
And now they have an API and everything that you can use to interact with this.

104
00:06:00,720 --> 00:06:03,840
But essentially, it's just a tool out there.

105
00:06:03,840 --> 00:06:09,520
It's the next iteration of freely accessible AI that people can use.

106
00:06:11,200 --> 00:06:16,480
Yeah, I mean, most of us will remember the chat bots

107
00:06:16,840 --> 00:06:21,320
that came out when we were in in high school and college and, you know,

108
00:06:21,320 --> 00:06:24,880
the MSN messenger bots that you could chat to and would respond to you.

109
00:06:24,880 --> 00:06:28,480
This is a much more advanced version of that

110
00:06:29,920 --> 00:06:32,400
based on a lot more deep learning,

111
00:06:32,920 --> 00:06:36,720
the ability to take multiple sources of data and base.

112
00:06:37,960 --> 00:06:41,640
Answers contextually, which was really the big difference

113
00:06:41,640 --> 00:06:45,400
is that contextual understanding.

114
00:06:45,400 --> 00:06:48,560
And the main difference between it and previous iterations is scope.

115
00:06:49,120 --> 00:06:53,840
It's got so much more data and that data is better organized, better tagged,

116
00:06:53,840 --> 00:06:56,760
better qualified than what it was previously.

117
00:06:57,680 --> 00:07:01,760
Yeah, I don't think that anyone would be surprised if I said AI isn't new.

118
00:07:02,040 --> 00:07:03,840
Right. Machine learning isn't new.

119
00:07:03,840 --> 00:07:07,760
I remember when Siri came out on the iPhones before,

120
00:07:08,360 --> 00:07:10,440
the Google assistant or anything like that.

121
00:07:10,440 --> 00:07:12,200
And it was garbage.

122
00:07:12,200 --> 00:07:13,560
It simply didn't work. Right.

123
00:07:13,560 --> 00:07:17,520
You say set my clock and it would say, I can't understand what you're saying.

124
00:07:18,200 --> 00:07:22,960
But over the years, these major organizations have refined

125
00:07:22,960 --> 00:07:27,200
their processing models and their training.

126
00:07:27,440 --> 00:07:31,160
It's gone through millions and millions and millions of iterations

127
00:07:31,560 --> 00:07:35,560
of trying to refine of how do I pick out these keywords?

128
00:07:35,560 --> 00:07:38,000
And, you know, maybe when these two keywords are together,

129
00:07:38,000 --> 00:07:42,080
it means something slightly different and then really start piecing it together.

130
00:07:42,760 --> 00:07:48,120
The. I just pulled up some examples of what chat GPT can do.

131
00:07:48,120 --> 00:07:52,080
So maybe we start there of just a basic overview

132
00:07:52,080 --> 00:07:53,680
to start setting the context.

133
00:07:53,680 --> 00:07:56,920
You know, Matthew said it's it's extremely wide in scope.

134
00:07:58,640 --> 00:08:02,080
When I started playing around with it, I said, write a love letter to my wife.

135
00:08:02,080 --> 00:08:05,640
I, you know, maybe I thought that I made a mistake that day or whatever. Right.

136
00:08:05,640 --> 00:08:08,600
Like, I suck at writing love letters. So let the tool do it for me.

137
00:08:08,600 --> 00:08:11,560
I didn't I didn't actually send it to her, but it was pretty dang good

138
00:08:11,560 --> 00:08:13,240
and better than I could do.

139
00:08:13,240 --> 00:08:15,040
It can correct grammar.

140
00:08:15,040 --> 00:08:18,080
It can summarize things to a second grader.

141
00:08:18,080 --> 00:08:22,080
So I hypothetically said, what's a sundog? Right.

142
00:08:22,080 --> 00:08:24,680
It's the kind of the halo around the sun with the ice crystals.

143
00:08:24,680 --> 00:08:27,360
If you're in the somewhere where there's winter.

144
00:08:27,360 --> 00:08:31,200
And then I said, summarize it like a fifth grader.

145
00:08:31,200 --> 00:08:33,840
And it would actually do that. It made it easier to read,

146
00:08:34,640 --> 00:08:36,800
taking code and.

147
00:08:37,680 --> 00:08:39,520
Explaining what that code is doing.

148
00:08:39,520 --> 00:08:42,240
Likewise, saying, here's what I want to do.

149
00:08:42,240 --> 00:08:44,600
Go build some code for me.

150
00:08:44,600 --> 00:08:46,640
You can do a ton of other stuff. Right.

151
00:08:46,640 --> 00:08:52,720
Does classify these types of movies or tell me the tone of this language?

152
00:08:52,720 --> 00:08:55,520
Is it upset? Is it happy? Is it sad?

153
00:08:57,000 --> 00:08:58,600
The list is expansive.

154
00:08:58,600 --> 00:09:00,720
That's just a really, really high level overview.

155
00:09:01,480 --> 00:09:04,080
Yeah. And there's a lot of stuff.

156
00:09:04,080 --> 00:09:05,600
Oh, sorry. Oh, no.

157
00:09:05,600 --> 00:09:08,560
I was just going to interject just really quickly, too, because, right,

158
00:09:08,560 --> 00:09:10,320
we're in a podcast and people are just listening.

159
00:09:10,320 --> 00:09:14,000
If for whatever reason you haven't used chat GPT, I know Matthew was like,

160
00:09:14,000 --> 00:09:15,840
oh, remember the chat box back in the day?

161
00:09:15,840 --> 00:09:18,880
It literally just seems like you're instant messaging.

162
00:09:18,880 --> 00:09:22,640
I mean, with a robot, but it does it within a chat thread

163
00:09:22,640 --> 00:09:24,240
that remembers your history.

164
00:09:24,240 --> 00:09:27,520
So just wanted to lay that out. There's somebody that likes using it.

165
00:09:27,520 --> 00:09:30,800
So exactly what Nate's saying of, hey, do this. Now do it this way.

166
00:09:30,800 --> 00:09:34,320
Now do it this way. Now do it with this context from this website.

167
00:09:35,200 --> 00:09:37,120
That's what's made it easier than the past tools,

168
00:09:37,120 --> 00:09:39,840
at least that I've been marketed to as a marketer that have said,

169
00:09:39,840 --> 00:09:41,840
hey, do these AI things for you.

170
00:09:42,480 --> 00:09:45,040
Yeah. And previously those tags were it.

171
00:09:45,040 --> 00:09:47,840
You would just say six tags and it would spit something out

172
00:09:47,840 --> 00:09:49,280
that matched all those tags.

173
00:09:50,160 --> 00:09:53,680
And now it's contextually understanding what you're saying

174
00:09:53,680 --> 00:09:57,040
to rearrange what it gave you previously.

175
00:09:58,320 --> 00:10:03,600
Nate mentioned a lot of items that I feel were about regressing the language.

176
00:10:03,600 --> 00:10:06,240
So it'll take what you've said and it will explain it

177
00:10:06,240 --> 00:10:09,200
in a certain phrasing that's lower. It'll also do the opposite.

178
00:10:09,200 --> 00:10:12,480
It'll go higher. It'll put it into a PhD understanding.

179
00:10:12,480 --> 00:10:16,160
It'll put it into a college understanding if you've originally got it

180
00:10:16,160 --> 00:10:19,200
in a high school understanding. And the same with code.

181
00:10:19,200 --> 00:10:22,400
You can say, I don't know what this code is.

182
00:10:22,400 --> 00:10:24,880
Write this code for me and it will write it.

183
00:10:24,880 --> 00:10:28,800
And then you can say, that's great, but I need you to use these modules

184
00:10:28,800 --> 00:10:33,600
and I need you to use this backend system and it will rewrite the code.

185
00:10:33,600 --> 00:10:36,960
And you can say, be more concise and it will cut it down

186
00:10:36,960 --> 00:10:38,720
and obfuscate the code for you.

187
00:10:38,720 --> 00:10:41,840
Those of you who are listening are probably seeing,

188
00:10:42,880 --> 00:10:45,520
there may be problems with this. We're going to get to some of that.

189
00:10:46,720 --> 00:10:50,960
But the goal here is that it will take things that you provided

190
00:10:50,960 --> 00:10:53,520
or things it creates on your prompt and mess with them

191
00:10:53,520 --> 00:10:56,880
in ways that previously these products definitely couldn't.

192
00:10:57,920 --> 00:11:02,480
There's also the Deep Fake podcast we did previously.

193
00:11:03,280 --> 00:11:05,520
I believe all three of us were in that one.

194
00:11:05,520 --> 00:11:08,720
Did a similar thing with images that we spoke about and audio.

195
00:11:09,600 --> 00:11:13,200
That one's creating them from scratch based on prior information.

196
00:11:13,200 --> 00:11:15,600
This one's doing it based on a similar thing.

197
00:11:15,600 --> 00:11:21,040
It's taking information from elsewhere and using that to create this.

198
00:11:21,520 --> 00:11:25,680
Whether it's come from GitHub and that's where it's getting all its code from

199
00:11:25,680 --> 00:11:28,240
or whether it's coming from academic articles

200
00:11:28,240 --> 00:11:30,720
and they've got access to Jscore and JStore

201
00:11:30,720 --> 00:11:32,480
and they've just let everyone go through it.

202
00:11:32,480 --> 00:11:38,160
The data that's coming into this that it's building from is extensive to say the least.

203
00:11:38,800 --> 00:11:42,000
They've got a lot to pull from when they're creating it and it's contextual.

204
00:11:42,000 --> 00:11:46,080
It's very, it's smart is really the only way.

205
00:11:46,080 --> 00:11:50,800
It's scary how quickly it can respond and how quickly it spits it out

206
00:11:50,800 --> 00:11:52,800
has also been surprising to me.

207
00:11:52,800 --> 00:11:56,000
I'd expected a couple more seconds even.

208
00:11:56,000 --> 00:12:02,800
Kelsey, did you have something to add otherwise I can keep going?

209
00:12:02,800 --> 00:12:05,520
No, I can keep on building off everything that we're saying.

210
00:12:05,520 --> 00:12:09,920
I could do this but to keep it timely within today, I'll hold back.

211
00:12:11,120 --> 00:12:14,960
Okay, maybe we can touch on just a couple of the applications

212
00:12:14,960 --> 00:12:16,960
that from a business perspective.

213
00:12:16,960 --> 00:12:21,760
So Kelsey, you're helping lead the podcast.

214
00:12:21,760 --> 00:12:24,160
I know that you're helping me with the podcast.

215
00:12:24,160 --> 00:12:26,960
I know that you're playing around with it.

216
00:12:26,960 --> 00:12:28,960
What were you able to do with it?

217
00:12:28,960 --> 00:12:31,760
And then maybe I can touch on some of the cybersecurity side of things.

218
00:12:32,960 --> 00:12:36,800
Yeah, I mean, I'll kind of use because I've been playing around with pieces of

219
00:12:36,800 --> 00:12:40,480
the podcast planning, the publishing, where I've used it just as like a,

220
00:12:40,480 --> 00:12:41,200
hey, why not?

221
00:12:41,200 --> 00:12:46,000
Because I'm not egotistical enough to think that I'm smarter than an AI.

222
00:12:46,000 --> 00:12:46,880
Let's put it that way.

223
00:12:46,880 --> 00:12:49,600
If they could probably do it faster and better than I can.

224
00:12:49,600 --> 00:12:54,880
So right as far as the planning that I can say, I can tell me this is what I want

225
00:12:54,880 --> 00:12:59,440
the next podcast to be and I can go, hey, write me an outline for a podcast

226
00:13:00,080 --> 00:13:01,040
talking about this.

227
00:13:01,840 --> 00:13:05,840
So it'll write the entire outline and then I can send it to you guys and go,

228
00:13:05,840 --> 00:13:06,960
fill in what you need.

229
00:13:06,960 --> 00:13:08,640
Here's the outline for you.

230
00:13:08,640 --> 00:13:09,280
Done and done.

231
00:13:09,280 --> 00:13:11,360
We sit down, we record all the great things.

232
00:13:11,360 --> 00:13:12,400
I publish it.

233
00:13:12,400 --> 00:13:15,280
I can go based on these notes that we have on this outline,

234
00:13:15,280 --> 00:13:16,800
write me a YouTube description.

235
00:13:16,800 --> 00:13:18,800
Now write me three LinkedIn posts.

236
00:13:19,440 --> 00:13:23,760
Now write me an email promoting this and it can do all of those things for me

237
00:13:23,760 --> 00:13:25,360
while I'm waiting for it to export.

238
00:13:25,360 --> 00:13:27,680
While I'm waiting for it to go, I could ask it for tags.

239
00:13:27,680 --> 00:13:29,920
I can ask it for upcoming topics.

240
00:13:29,920 --> 00:13:31,280
I can ask it for trends.

241
00:13:31,280 --> 00:13:35,280
And then I can put the transcript in and say, pull me out quotes from this

242
00:13:35,280 --> 00:13:38,560
transcript and it'll pull me out and I'll say, hey, I need 10 more.

243
00:13:38,560 --> 00:13:40,080
That one's great, but this is the speaker.

244
00:13:40,080 --> 00:13:40,960
I need this to say this.

245
00:13:40,960 --> 00:13:42,640
And it'll go, yep, 100% you're good.

246
00:13:42,640 --> 00:13:44,720
And I'm like, but this is about this website.

247
00:13:44,720 --> 00:13:45,760
This is our podcast.

248
00:13:45,760 --> 00:13:47,040
This is about this website.

249
00:13:47,040 --> 00:13:47,840
This is our website.

250
00:13:47,840 --> 00:13:50,240
I need you to be able to say it's from there.

251
00:13:50,240 --> 00:13:52,320
And it'll say, yep, okay, that's all great.

252
00:13:52,320 --> 00:13:59,520
So now I have all of this content that took me less than an hour to do with just one me

253
00:13:59,520 --> 00:14:02,560
to be able to go through and say, well, now I have an entire quarter of podcasts.

254
00:14:02,560 --> 00:14:03,760
I have all the LinkedIn posts.

255
00:14:03,760 --> 00:14:04,880
I have all the quotes done.

256
00:14:04,880 --> 00:14:08,400
And the only thing I need to do is edit the video, which I'm just waiting for it

257
00:14:08,400 --> 00:14:11,600
to be able to do for me at this point, because I'm really just assuming that's

258
00:14:11,600 --> 00:14:13,120
the next iteration.

259
00:14:13,120 --> 00:14:16,160
That's just the very base level of what I've used it for for this.

260
00:14:16,160 --> 00:14:20,480
I want to give a quick caveat, because I think you're underselling yourself quickly.

261
00:14:20,480 --> 00:14:22,640
It's not perfect.

262
00:14:22,640 --> 00:14:23,200
Okay.

263
00:14:23,200 --> 00:14:25,280
So chat GBT is still making mistakes.

264
00:14:25,280 --> 00:14:26,960
It's still got contextual issues.

265
00:14:26,960 --> 00:14:29,600
There is a lot of editing that goes to getting it to spit that out.

266
00:14:29,600 --> 00:14:34,480
I just feel like you've maybe undersold how much effort's going into getting that all

267
00:14:34,480 --> 00:14:35,360
done afterwards.

268
00:14:35,360 --> 00:14:41,040
Because yeah, it does do a lot of these things and it will understand them, but it's not

269
00:14:41,040 --> 00:14:41,680
a human.

270
00:14:41,680 --> 00:14:46,800
It still will completely misunderstand the context of a word sometimes because they use

271
00:14:46,800 --> 00:14:51,520
differently in different places and it will just fail at some things.

272
00:14:52,240 --> 00:14:56,400
So it's very useful and obviously it can do a lot of this, but there is a decent amount

273
00:14:56,400 --> 00:14:59,440
of time going through it and making sure it's speaking properly.

274
00:14:59,440 --> 00:15:02,080
And that you gave it the right context.

275
00:15:02,080 --> 00:15:05,760
It reminds me a little bit of when they're like Googling is a life skill, which we say

276
00:15:05,760 --> 00:15:08,800
is a joke, but it's the same thing that you're saying.

277
00:15:08,800 --> 00:15:13,440
Don't just put in exactly what you want to think Google to like, you gotta do keywords,

278
00:15:13,440 --> 00:15:16,320
the correct things, quotation marks, hyphens.

279
00:15:16,320 --> 00:15:21,520
So yes, there's maybe a little bit that I've just done knee jerk after learning from Google.

280
00:15:22,240 --> 00:15:23,040
So yeah.

281
00:15:23,040 --> 00:15:31,840
So the way I'd summarize that is it's a major efficiency gain on a lot of the basics to

282
00:15:31,840 --> 00:15:33,760
at least get the ideas flowing.

283
00:15:33,760 --> 00:15:39,280
And then you use your creative juices to fill in the gaps or refine it or tune it, you know,

284
00:15:39,280 --> 00:15:41,920
and create the tone that you want with that data.

285
00:15:42,880 --> 00:15:43,520
Exactly.

286
00:15:43,520 --> 00:15:48,240
Nate, I just want to just very briefly, one of the things I heard about it being used

287
00:15:48,240 --> 00:15:55,040
for recently on Twitter was an author who is just finishing a novel that they're already

288
00:15:55,040 --> 00:15:56,320
signed on to publish.

289
00:15:56,320 --> 00:16:05,440
And they logged in their blurb and skipped out on the ending and they thought their ending

290
00:16:05,440 --> 00:16:08,160
was very unique and unpredictable.

291
00:16:08,160 --> 00:16:09,920
And chat GPT guessed it.

292
00:16:10,800 --> 00:16:12,320
It guessed the ending straight away.

293
00:16:12,320 --> 00:16:17,280
And I think that that's another one of those things where using it as a bias checker or

294
00:16:17,280 --> 00:16:20,640
as a way of saying, Hey, what do you think is going to happen next in this?

295
00:16:20,640 --> 00:16:25,120
Or what, what makes sense to happen next in this is a cool feature, especially if you're

296
00:16:25,120 --> 00:16:28,640
writing a report, maybe you're writing, like putting together a presentation and you're

297
00:16:28,640 --> 00:16:30,720
like, I don't quite think the ending's right.

298
00:16:31,200 --> 00:16:31,760
Ask it.

299
00:16:32,640 --> 00:16:37,040
I'm going to start asking it to guess my wife's favorite movie endings, which is romance

300
00:16:37,040 --> 00:16:38,720
because I'm pretty sure it'll be right.

301
00:16:38,720 --> 00:16:43,440
So, Hey, but romance, I mean, the ending is kind of either they all die or they get

302
00:16:43,440 --> 00:16:44,000
together.

303
00:16:44,000 --> 00:16:45,120
There's no in between.

304
00:16:45,600 --> 00:16:47,920
Like I said, easy task for chat GPT.

305
00:16:47,920 --> 00:16:53,680
So, um, quick, another business application, and I'm going to front load this conversation

306
00:16:53,680 --> 00:16:59,200
saying never ever put sensitive information into the tool.

307
00:16:59,200 --> 00:17:02,240
They use this information to better train it.

308
00:17:02,240 --> 00:17:08,080
So as I'm going to be talking about some code here, if you're sticking your API secret keys

309
00:17:08,080 --> 00:17:13,520
into there or, you know, customer names or, you know, other info that you don't want someone

310
00:17:13,520 --> 00:17:15,600
else to see, don't use this tool.

311
00:17:16,320 --> 00:17:23,280
Um, so jumping into the coding concept then, um, it's a great resource for you to use.

312
00:17:23,280 --> 00:17:28,080
It's a resource for learning and helping explain the code that you maybe don't understand.

313
00:17:28,080 --> 00:17:29,680
So I love PowerShell.

314
00:17:29,680 --> 00:17:32,080
I've been doing PowerShell for quite some time now.

315
00:17:32,080 --> 00:17:35,840
We have people on the team that are coming out of college, right?

316
00:17:35,840 --> 00:17:37,680
They're, they're brand new to the role.

317
00:17:37,680 --> 00:17:38,800
They're still learning.

318
00:17:38,800 --> 00:17:42,000
You know, maybe they have some Python experience, but PowerShell is new.

319
00:17:42,000 --> 00:17:47,680
Um, at that point, you can use the tool to say, I'm looking for this type of code and

320
00:17:47,680 --> 00:17:53,840
taking that code and forming the basis of what you're looking to do.

321
00:17:53,840 --> 00:17:55,920
Never copy paste code and just run it, right?

322
00:17:55,920 --> 00:18:01,680
That's super dangerous, but at least taking and starting to build some structure and then

323
00:18:01,680 --> 00:18:05,920
you can take those different components of it and actually go write your own code and

324
00:18:05,920 --> 00:18:10,400
have someone who actually knows what they're doing, validate it as well before you execute

325
00:18:10,400 --> 00:18:10,960
it.

326
00:18:10,960 --> 00:18:14,960
So, but it can be a useful training tool.

327
00:18:14,960 --> 00:18:17,200
Um, if you're stuck, right?

328
00:18:17,200 --> 00:18:20,240
Maybe that one person that you're looking for is unavailable.

329
00:18:20,240 --> 00:18:21,200
They're on vacation.

330
00:18:21,200 --> 00:18:23,840
They're, um, currently out at lunch.

331
00:18:24,560 --> 00:18:27,840
Pop something in there and say, I'm looking for code to iterate through an array.

332
00:18:28,400 --> 00:18:30,480
It could probably help you with something like that.

333
00:18:31,440 --> 00:18:36,080
But like I said, never put sensitive info in it and never execute just random code.

334
00:18:37,680 --> 00:18:38,080
Agreed.

335
00:18:38,720 --> 00:18:40,160
We have a risk section.

336
00:18:40,160 --> 00:18:42,400
I know we wanted to get to about this.

337
00:18:42,400 --> 00:18:47,200
I will say that I think you've, you've covered the majority of that risk side of things,

338
00:18:47,200 --> 00:18:55,920
which is just be aware that it's not your product and be aware that it may do things wrong.

339
00:18:56,560 --> 00:19:02,080
It may tell you malicious code because someone has force fed it enough info that it starts

340
00:19:02,080 --> 00:19:03,600
to think that malicious code is good.

341
00:19:03,600 --> 00:19:06,240
It may have picked up a module with the incorrect location.

342
00:19:06,240 --> 00:19:08,880
There's a bunch of things it could do that are wrong.

343
00:19:08,880 --> 00:19:14,400
So don't take it as gospel and just run it in your system.

344
00:19:14,400 --> 00:19:16,480
Don't always use the language that you get.

345
00:19:17,200 --> 00:19:19,120
Make sure you're reading through it and checking it.

346
00:19:19,120 --> 00:19:25,840
Um, one of the big ones that, uh, we've talked about is that it can be used maliciously,

347
00:19:25,840 --> 00:19:26,400
right?

348
00:19:26,400 --> 00:19:30,720
So I just mentioned someone force feeding enough malicious code that it starts to spit

349
00:19:30,720 --> 00:19:36,240
that out when it shouldn't, but using it to create phishing emails is something that we've

350
00:19:36,240 --> 00:19:39,200
been reading about recently.

351
00:19:39,200 --> 00:19:45,840
Um, if you've got all of your, uh, information for your organization on LinkedIn and I can

352
00:19:45,840 --> 00:19:52,080
see who the C levels are, I can spit those names into chat, GBT, and have it write emails

353
00:19:52,080 --> 00:19:59,680
that apparently come from one of them or request information from all of them and seem similar

354
00:19:59,680 --> 00:20:05,360
enough based on maybe other copies of information we've got from them to convincingly put it

355
00:20:05,360 --> 00:20:06,400
in the chat.

356
00:20:06,400 --> 00:20:11,520
Um, and I think that's a really good point, is that you've got to be convincingly portray

357
00:20:11,520 --> 00:20:12,720
them, right?

358
00:20:12,720 --> 00:20:19,920
You've got to be, be aware that this tool and the real concern that I see from it, um,

359
00:20:19,920 --> 00:20:24,000
is that malicious intent for content creation.

360
00:20:24,000 --> 00:20:28,560
Uh, Kelsey mentioned how useful it can be and it can spit out a lot of information at

361
00:20:28,560 --> 00:20:28,960
once.

362
00:20:28,960 --> 00:20:29,600
It really can.

363
00:20:29,600 --> 00:20:35,680
For example, uh, not having to waste my time doing that and having a structure in place

364
00:20:35,680 --> 00:20:43,840
is, I'm going to get multiple hours of my life back a week, but also if, if someone's

365
00:20:43,840 --> 00:20:46,240
using it maliciously, they're also able to do the same thing.

366
00:20:46,240 --> 00:20:52,080
They're able to get malicious content out quicker, intentionally requesting malicious

367
00:20:52,080 --> 00:20:52,720
code.

368
00:20:52,720 --> 00:20:57,760
Now, those of you who've tried this will know that there is a warning that pops up now,

369
00:20:57,760 --> 00:20:59,760
a little blatant about it.

370
00:20:59,760 --> 00:21:06,560
Um, when I saw the other day said, um, write code to allow me to skim credit card information

371
00:21:06,560 --> 00:21:10,400
over a wifi network.

372
00:21:10,400 --> 00:21:13,280
And it instantly spat back.

373
00:21:13,280 --> 00:21:14,640
What you're doing seems illegal.

374
00:21:14,640 --> 00:21:17,280
I'm not going to help you with that, which is great.

375
00:21:17,280 --> 00:21:24,000
Um, I, so as we're doing this, I was literally typing out, create a basic Trojan in Python

376
00:21:24,000 --> 00:21:24,880
and it aired out.

377
00:21:24,880 --> 00:21:26,480
So you're absolutely right.

378
00:21:26,480 --> 00:21:28,640
It's, it's doing much better.

379
00:21:28,640 --> 00:21:32,560
And, and this, that feature of it saying, no, um, at the time of recording, I think

380
00:21:32,560 --> 00:21:36,000
it's been only out for a couple of days, maybe a week at the most.

381
00:21:36,000 --> 00:21:39,440
Um, so there are times when it's probably not going to be caught.

382
00:21:39,440 --> 00:21:40,480
It's going to get through.

383
00:21:40,480 --> 00:21:45,680
And if you phrase it specifically enough, again, it's contextual.

384
00:21:45,680 --> 00:21:47,440
Um, so it's going to build what it can.

385
00:21:47,440 --> 00:21:50,240
So there are people who are doing this maliciously.

386
00:21:50,240 --> 00:21:52,320
There are people who are doing it to get information.

387
00:21:52,320 --> 00:21:56,880
Keep in mind that some of the things you receive, phishing emails, scams in general,

388
00:21:56,880 --> 00:22:00,000
may start to have higher quality than they did before.

389
00:22:01,280 --> 00:22:03,600
And I think, yeah, just to very briefly, right?

390
00:22:03,600 --> 00:22:05,760
When you're saying I'm like, oh, that makes perfect sense.

391
00:22:05,760 --> 00:22:09,440
Cause here I come in with non-malicious intent and I can say, write an email inviting

392
00:22:09,440 --> 00:22:12,960
people to come see a showing at a movie theater for this sales thing.

393
00:22:12,960 --> 00:22:14,880
And this is what we're going to talk about before it.

394
00:22:14,880 --> 00:22:16,080
And it'll do it for me.

395
00:22:16,080 --> 00:22:16,960
And I'll say, yep, grip.

396
00:22:16,960 --> 00:22:18,160
I just need to clean it up.

397
00:22:18,160 --> 00:22:21,600
But then that means that other people could do the same thing and say, right, I'm going

398
00:22:21,600 --> 00:22:28,800
to email from the CEO of a tech company asking employees to verify their W2 or whatever.

399
00:22:28,800 --> 00:22:32,320
And it would do it because right, there's a good use case for it.

400
00:22:32,320 --> 00:22:33,360
That's not malicious.

401
00:22:35,120 --> 00:22:35,600
Yeah.

402
00:22:35,600 --> 00:22:36,080
Yeah.

403
00:22:36,080 --> 00:22:41,600
One, one of the topics, so we had a little bit of a internal debate here at CIT is, you

404
00:22:41,600 --> 00:22:45,360
know, when all of this started blowing up is, is chat GPT good or is it bad?

405
00:22:45,360 --> 00:22:45,840
Right.

406
00:22:45,840 --> 00:22:49,760
And the kind of the consensus is it depends, right?

407
00:22:49,760 --> 00:22:51,600
You know, no different than we're talking about here.

408
00:22:51,600 --> 00:22:55,760
But one of the interesting things that we were talking about is supply chain risks as

409
00:22:55,760 --> 00:22:56,400
well.

410
00:22:56,400 --> 00:23:05,040
So while the chat GPT data set right now only goes to 2021, future iterations of this may

411
00:23:05,040 --> 00:23:06,000
be live, right?

412
00:23:06,000 --> 00:23:07,280
It's pulling info.

413
00:23:07,280 --> 00:23:12,560
If someone knows that you're actively using chat GPT or any other solution out there,

414
00:23:12,560 --> 00:23:12,800
right?

415
00:23:12,800 --> 00:23:13,920
This is just one model.

416
00:23:15,440 --> 00:23:19,440
If they know you're doing that, similar to what Matthew was saying is, what if they just

417
00:23:19,440 --> 00:23:28,240
start feeding it bad data and now it starts leading you off into misdirection or some

418
00:23:28,240 --> 00:23:32,880
type of malicious outcome that you didn't intend, but the tool brought you there because

419
00:23:32,880 --> 00:23:34,720
someone else was using it nefariously.

420
00:23:35,280 --> 00:23:40,080
So the, I'm going to quote some of the Spider-Man, right?

421
00:23:40,080 --> 00:23:42,080
With great power comes great responsibility.

422
00:23:42,080 --> 00:23:48,800
Every single tool out there can be used for a great purpose or with malicious intent.

423
00:23:48,800 --> 00:23:54,640
AI, like I said, quite a while ago now was it's nothing new, right?

424
00:23:54,640 --> 00:24:01,920
We have major committees being formed to address the morality and the application of AI.

425
00:24:02,560 --> 00:24:04,800
So this is just another component of it.

426
00:24:04,800 --> 00:24:12,640
Yeah. And there's a, along with that, we're talking about specific risks.

427
00:24:13,360 --> 00:24:17,360
I want to bring up the biases that come into a product like this.

428
00:24:18,320 --> 00:24:21,280
I mentioned one of them already, which is the academic bias.

429
00:24:21,280 --> 00:24:28,640
If you're training this tool on how to write scientific articles for journals,

430
00:24:28,640 --> 00:24:30,240
you have an academic bias.

431
00:24:30,240 --> 00:24:35,360
It's going to be writing by default naturally towards that style.

432
00:24:36,080 --> 00:24:41,200
So keep in mind that that's not bad or good, but is it what you want?

433
00:24:42,080 --> 00:24:43,280
Is it what you're aiming for?

434
00:24:43,840 --> 00:24:46,800
There's also cultural, there's monetary.

435
00:24:47,680 --> 00:24:52,880
Some of them, if you're phrasing it generally may come from, they may have a Fortune 500 mindset

436
00:24:52,880 --> 00:24:58,800
in who it's being written for or what the guidelines are of the moral and ethical lines

437
00:24:58,800 --> 00:25:02,800
that would be used in comparison to whether you're working as a contractor or whether you're

438
00:25:02,800 --> 00:25:04,400
working in a small business.

439
00:25:04,400 --> 00:25:08,880
The language you use amongst your employees when you're building these things is going to change

440
00:25:09,920 --> 00:25:11,760
based on a million different factors.

441
00:25:11,760 --> 00:25:20,400
So keep that in mind and it can be told to use specific biases as well, specific tones.

442
00:25:20,400 --> 00:25:25,680
So mention that, say like Nate mentioned, the one where you can say for a second grader,

443
00:25:25,680 --> 00:25:31,680
it will take those prompts and use them based on the information it has to better aim the language

444
00:25:31,680 --> 00:25:34,960
at the people that you're trying to get it written for.

445
00:25:36,160 --> 00:25:37,360
So that's the first one.

446
00:25:38,560 --> 00:25:40,080
Oh, what's so funny, Nate?

447
00:25:41,440 --> 00:25:46,320
One of the things, so when you're talking about intonation of how you want it to write,

448
00:25:47,840 --> 00:25:51,200
during performance review time for our employees, I was like,

449
00:25:51,200 --> 00:25:54,000
write a performance review for a GRC analyst, right?

450
00:25:54,000 --> 00:25:56,560
Yeah.

451
00:25:56,560 --> 00:25:57,520
Just see what it says.

452
00:25:57,520 --> 00:25:59,680
And I said, make it a negative review.

453
00:26:01,120 --> 00:26:02,800
And then that's when I submitted off to Matthew.

454
00:26:02,800 --> 00:26:04,160
So yes, I saw that.

455
00:26:04,160 --> 00:26:04,960
That was very funny.

456
00:26:05,600 --> 00:26:07,360
But it can do stuff like that.

457
00:26:07,360 --> 00:26:08,240
So exactly.

458
00:26:08,240 --> 00:26:12,800
And for the record, you did warn me you had done that before I got it.

459
00:26:12,800 --> 00:26:16,560
So it wasn't like I got it out of the blue, but it was great.

460
00:26:16,560 --> 00:26:18,320
And those are the things you can do.

461
00:26:18,320 --> 00:26:21,040
You can intentionally phrase it that way.

462
00:26:21,040 --> 00:26:27,200
But one thing I've noticed with chat GBT is that it does seem to lack some heart.

463
00:26:28,000 --> 00:26:31,120
And so it will intentionally push towards enterprise.

464
00:26:31,120 --> 00:26:33,360
It will intentionally push towards academic.

465
00:26:34,800 --> 00:26:39,360
If you want it to show kindness, you generally need to tell it to.

466
00:26:40,240 --> 00:26:42,160
So that's the first one.

467
00:26:43,280 --> 00:26:47,280
And for me, we talked about the positives and negatives.

468
00:26:47,280 --> 00:26:51,680
And I have a lot of feelings about this, both positive and negative.

469
00:26:51,680 --> 00:26:55,120
And I wanted to give an example that I heard about the other day

470
00:26:55,120 --> 00:27:03,520
that has helped me remember to stay as neutral or as positive about this as possible,

471
00:27:03,520 --> 00:27:09,280
which is about the calculators and when calculators began becoming popular.

472
00:27:10,960 --> 00:27:16,720
So in short, and this was before my time, so I am reading from articles.

473
00:27:16,720 --> 00:27:17,840
This was research I did.

474
00:27:18,640 --> 00:27:23,520
But basically, apparently there was a lot of pushback when they first started being used.

475
00:27:23,520 --> 00:27:26,960
It was considered that they would stop kids from using their brain

476
00:27:26,960 --> 00:27:29,520
and thinking and having to know how the math worked.

477
00:27:29,520 --> 00:27:30,800
They would stop learning proofs.

478
00:27:30,800 --> 00:27:32,080
They would stop learning everything.

479
00:27:32,080 --> 00:27:35,200
They just type it into a calculator, put the answer down and walk away.

480
00:27:36,480 --> 00:27:41,360
I can understand pushback from this tool on that exact thing.

481
00:27:41,360 --> 00:27:44,880
When a high schooler can just say, write an essay on Romeo and Juliet,

482
00:27:44,880 --> 00:27:50,400
never read it and submit it and get an A because of course, chat GPT knows how to do that.

483
00:27:52,000 --> 00:27:55,200
So in the case of calculators, what happened was,

484
00:27:55,840 --> 00:27:58,800
yes, it took a while for them to become popular and used,

485
00:27:58,800 --> 00:28:05,360
but it changed the way that math was taught by making what had previously been harder mathematics

486
00:28:06,720 --> 00:28:08,880
move down into earlier grades.

487
00:28:08,880 --> 00:28:12,320
I distinctly remember talking to my father when I was in grade nine

488
00:28:12,320 --> 00:28:17,440
and him saying that the math we were doing there was the math he had done in grade 12

489
00:28:17,440 --> 00:28:23,200
with pen and paper because calculators weren't allowed in classrooms at that time.

490
00:28:23,200 --> 00:28:28,960
And so he had had to work on doing all of it, pen and paper,

491
00:28:28,960 --> 00:28:34,400
and I could just type it into my graphing calculator in grade nine and get the answer.

492
00:28:34,400 --> 00:28:37,520
We still had the basic knowledge that came from having to learn it,

493
00:28:38,160 --> 00:28:40,240
but I then got to study more.

494
00:28:40,240 --> 00:28:44,640
I got to learn more and I think that's where I see this going.

495
00:28:44,640 --> 00:28:51,920
If you aren't spending hours upon hours writing a report into a system

496
00:28:52,800 --> 00:28:55,760
every night for homework, which I did occasionally,

497
00:28:57,840 --> 00:29:02,560
you're going to have time to do deeper dives into maybe why that language is important,

498
00:29:02,560 --> 00:29:07,040
why these items matter, why knowing why this report works is critical,

499
00:29:07,040 --> 00:29:11,440
even if you didn't have to write it, will become more critical of why it works

500
00:29:11,440 --> 00:29:12,640
rather than how to do it.

501
00:29:14,240 --> 00:29:15,360
So I think there's benefits.

502
00:29:15,360 --> 00:29:17,520
I think there's a lot of really good that can come from it.

503
00:29:17,520 --> 00:29:20,000
We have Wolfram Alpha already, which has been out of spit out

504
00:29:20,000 --> 00:29:22,160
mathematic equations since I was in high school

505
00:29:23,120 --> 00:29:24,960
and tell us why things work and didn't work.

506
00:29:25,840 --> 00:29:26,960
Yes, it's scary.

507
00:29:26,960 --> 00:29:30,720
Yes, it's changing a lot, but it's going to happen.

508
00:29:30,720 --> 00:29:32,880
It's here that we can't shut it down.

509
00:29:32,880 --> 00:29:36,240
Yeah, I'll just quick mention, I know we're running out of time,

510
00:29:36,240 --> 00:29:40,400
so I'll just touch on a little bit more of the academic stuff real quick.

511
00:29:40,400 --> 00:29:44,000
There's a lot of concerns about this in the academic community, right?

512
00:29:44,000 --> 00:29:48,400
There's multiple schools that have actually banned it from being used altogether.

513
00:29:49,600 --> 00:29:56,800
I'm sure students will still use it, but there's concerns that the school is taking tests,

514
00:29:56,800 --> 00:30:01,920
they'll feed it questions and it's passing MBA courses through, I believe, Warcraft.

515
00:30:01,920 --> 00:30:04,800
I believe Wharton was the one that tested it, right?

516
00:30:04,800 --> 00:30:08,000
Wharton's a prestigious school and it's still at least passing, right?

517
00:30:08,000 --> 00:30:11,280
Maybe a C or something like that, but it's successful.

518
00:30:12,800 --> 00:30:15,040
So I guess the thing that I was going to say there is

519
00:30:16,000 --> 00:30:19,360
OpenAI, the organization that developed this tool,

520
00:30:19,360 --> 00:30:25,360
they're also doing beneficial things to try and identify the negative components of this.

521
00:30:25,360 --> 00:30:27,200
They're putting in content filtering.

522
00:30:27,200 --> 00:30:34,560
They're now putting in items such as detecting whether or not something was generated by AI.

523
00:30:34,560 --> 00:30:35,520
It's not the best right now.

524
00:30:35,520 --> 00:30:39,200
I think it had like 26% success rate, but they said with additional training,

525
00:30:39,200 --> 00:30:43,360
we'll be able to better identify when a tool like this is used.

526
00:30:43,360 --> 00:30:48,000
So there's components that, again, people are trying to make this a useful tool,

527
00:30:48,000 --> 00:30:52,000
but just like calculators, just like the internet, just like AI,

528
00:30:52,000 --> 00:30:58,400
originally is these tools are being used to build efficiencies for the collective

529
00:30:59,280 --> 00:31:01,120
knowledge of all humans.

530
00:31:01,120 --> 00:31:06,720
And as we adopt these tools, while there's fear that comes with it,

531
00:31:06,720 --> 00:31:10,800
it still is going to help advance us at a faster and faster rate.

532
00:31:10,800 --> 00:31:15,440
By letting the technology do it, you just have to be aware of the concerns

533
00:31:15,440 --> 00:31:16,960
and the implications along the way.

534
00:31:16,960 --> 00:31:20,720
Now, I think to be honest, I could be like, and now I have a bunch of follow-up questions

535
00:31:20,720 --> 00:31:22,720
that we could do a whole other podcast.

536
00:31:22,720 --> 00:31:27,440
So that's probably where we'll leave it here today is that we will probably have a second podcast.

537
00:31:27,440 --> 00:31:32,160
I'm sure I've had other people internally say, hey, I also have opinions on it.

538
00:31:32,160 --> 00:31:35,600
So keep an eye out for that in the near future, but thank you both.

539
00:31:35,600 --> 00:31:40,640
And I think you both wrapped it up beautifully to say, yes, there's pros and cons, but it's here.

540
00:31:41,520 --> 00:31:44,480
And we can use that to help us understand what's going on.

541
00:31:44,480 --> 00:31:47,360
But it's here and we can use it.

542
00:31:47,360 --> 00:31:48,720
But of course, it's not perfect.

543
00:31:48,720 --> 00:31:51,360
So then if you're using it, quality control is still there.

544
00:31:51,360 --> 00:31:55,520
And right, like in my case, I can use it to get more things done.

545
00:31:55,520 --> 00:32:01,280
It certainly doesn't replace me in its entirety, but I can do the work three times as fast by not

546
00:32:01,280 --> 00:32:04,080
having to replicate all of those things every single time.

547
00:32:04,640 --> 00:32:08,480
So if anybody has additional questions at all, you know where to reach us.

548
00:32:08,480 --> 00:32:14,400
We're at info at cit-net.com or head on out to our website, cit-net.com.

549
00:32:14,400 --> 00:32:17,760
Thank you, Nate. Thank you, Matthew.

550
00:32:17,760 --> 00:32:47,680
And we'll be back next week with yet another episode.