1
00:00:00,000 --> 00:00:02,500
3C

2
00:00:30,000 --> 00:00:33,000
Hey Chris, how you doing?

3
00:00:33,000 --> 00:00:35,000
Good, how you doing, John?

4
00:00:35,000 --> 00:00:37,000
I'm doing good.

5
00:00:37,000 --> 00:00:46,000
We are one step in front of the other, you know, just trying to figure it out.

6
00:00:46,000 --> 00:00:52,000
I feel like I say that every single week, but it's a mantra, man.

7
00:00:52,000 --> 00:00:56,000
Just keep moving forward.

8
00:00:56,000 --> 00:01:01,000
This would be interesting tonight to see how this all works doing it remotely.

9
00:01:01,000 --> 00:01:03,000
Yeah, we're doing it remotely again.

10
00:01:03,000 --> 00:01:06,000
It seems to be working just fine right now.

11
00:01:06,000 --> 00:01:20,000
I think the real winner here is that you're on that audio Sigma device over there at your house and we don't seem to be having nearly the problems we were having before.

12
00:01:20,000 --> 00:01:28,000
I'm excited. I mean, just buy gear that works. I guess that's the...

13
00:01:28,000 --> 00:01:38,000
Yeah, I'd still love to know why I have all the trouble that I've had when I've had my roadcaster here on this side of it.

14
00:01:38,000 --> 00:01:45,000
Yeah, that's one of those things that I can't like.

15
00:01:45,000 --> 00:01:57,000
It's hard to be a problem solver for a problem solver.

16
00:01:57,000 --> 00:02:03,000
I feel like all of my friends whenever I'm like, you're having trouble with something?

17
00:02:03,000 --> 00:02:06,000
You're not supposed to have trouble with something.

18
00:02:06,000 --> 00:02:09,000
I have trouble with something. You tell me how to fix the thing I'm having trouble with.

19
00:02:09,000 --> 00:02:16,000
I'm not used to you having trouble.

20
00:02:16,000 --> 00:02:17,000
I'm only human.

21
00:02:17,000 --> 00:02:22,000
Yeah, I mean, I'm grateful because most of the time it doesn't seem like it.

22
00:02:22,000 --> 00:02:26,000
Not in the technology side of things.

23
00:02:26,000 --> 00:02:30,000
I guess that's the fun we all get to deal with.

24
00:02:30,000 --> 00:02:32,000
I have got new toys.

25
00:02:32,000 --> 00:02:33,000
Yes, you do.

26
00:02:33,000 --> 00:02:44,000
My wife gave me new toys and because I got new toys, I started messing around with my setup here at the house.

27
00:02:44,000 --> 00:02:51,000
So I've actually got a roadcaster duo now over here on this side.

28
00:02:51,000 --> 00:02:58,000
So hopefully everything's fine and we don't have any weirdness because of that.

29
00:02:58,000 --> 00:03:03,000
But everything's working okay right now.

30
00:03:03,000 --> 00:03:13,000
And then I have an old Hero 9 that I have plugged into my computer as a webcam now.

31
00:03:13,000 --> 00:03:17,000
So I'm actually getting...

32
00:03:17,000 --> 00:03:20,000
Is that the same mic arm that you've had or is that different as well?

33
00:03:20,000 --> 00:03:23,000
This is an XTM 100.

34
00:03:23,000 --> 00:03:24,000
Okay.

35
00:03:24,000 --> 00:03:29,000
So it's running into the roadcaster, which is kind of funny.

36
00:03:29,000 --> 00:03:32,000
Oh, so you're running USB-C into it?

37
00:03:32,000 --> 00:03:35,000
Yeah, I'm running USB-C into the duo.

38
00:03:35,000 --> 00:03:37,000
So I have two more.

39
00:03:37,000 --> 00:03:46,000
The idea would be that I use my Aston as one of the microphones.

40
00:03:46,000 --> 00:03:51,000
The real reason, I don't think I said this while we were recording,

41
00:03:51,000 --> 00:03:56,000
the reason that my wife got me this duo is I have a plan to...

42
00:03:56,000 --> 00:04:06,000
I started a plan a long time ago or a few years ago to start recording my parents telling stories

43
00:04:06,000 --> 00:04:13,000
because there's a lot of stories that my grandparents told that I'm starting to forget a little bit

44
00:04:13,000 --> 00:04:17,000
or just not remember the exact details of.

45
00:04:17,000 --> 00:04:22,000
And they have memories of their parents, my grandparents telling those stories.

46
00:04:22,000 --> 00:04:25,000
They also have memories of their grandparents telling stories.

47
00:04:25,000 --> 00:04:37,000
So I'm going to try to get as many of those family history anecdotes down as I can.

48
00:04:37,000 --> 00:04:40,000
And so the roadcaster duo is going to help me do that.

49
00:04:40,000 --> 00:04:45,000
I'm going to put an XLR microphone on my mom and an XLR microphone on my dad.

50
00:04:45,000 --> 00:04:48,000
And then I'm going to use this guy.

51
00:04:48,000 --> 00:05:01,000
Or if I can squeeze the $250 together, I'll put a wireless mic the road has that can connect to this guy.

52
00:05:01,000 --> 00:05:03,000
And that'd be pretty fun.

53
00:05:03,000 --> 00:05:10,000
You can do just the normal, like, little microphones, but they also have like an actual, you know,

54
00:05:10,000 --> 00:05:18,000
larger diaphragm capsule on an actual mic looking thing that you would use for like interviews and stuff.

55
00:05:18,000 --> 00:05:22,000
I think it's their purpose behind it, but also for this.

56
00:05:22,000 --> 00:05:26,000
Yeah, you're looking at like the, I think they call it the go me.

57
00:05:26,000 --> 00:05:27,000
Something like that.

58
00:05:27,000 --> 00:05:36,000
And then, no, I think the me is one of the, it's one of the little ones that you can buy like a,

59
00:05:36,000 --> 00:05:40,000
you can buy two of them and then the squares.

60
00:05:40,000 --> 00:05:42,000
Yeah.

61
00:05:42,000 --> 00:05:43,000
Oh, sorry.

62
00:05:43,000 --> 00:05:47,000
No, I'm talking about, this is interesting.

63
00:05:47,000 --> 00:05:53,000
I mean, maybe we haven't talked about this.

64
00:05:53,000 --> 00:05:57,000
Let me find the thing real quick.

65
00:05:57,000 --> 00:06:01,000
It's the, it's the road interview pro wireless handheld condenser microphone.

66
00:06:01,000 --> 00:06:05,000
I want to put this in the chat for you so you can look at it.

67
00:06:05,000 --> 00:06:13,000
It's a, or you can just look it up.

68
00:06:13,000 --> 00:06:16,000
I've seen that one before.

69
00:06:16,000 --> 00:06:18,000
Okay.

70
00:06:18,000 --> 00:06:27,000
I think they've upgraded it over the last couple of years, but it pretty nice little condenser microphone.

71
00:06:27,000 --> 00:06:32,000
So you don't have to deal with that little exciting bit.

72
00:06:32,000 --> 00:06:40,000
But anyway, it's an idea at least.

73
00:06:40,000 --> 00:06:51,000
And that thing only costs like 250, which, you know, looking at the other microphones that I've looked at buying for this type of thing.

74
00:06:51,000 --> 00:06:54,000
That's not too bad.

75
00:06:54,000 --> 00:06:58,000
So, or maybe I'll just go get a Luit.

76
00:06:58,000 --> 00:07:01,000
Can't go wrong.

77
00:07:01,000 --> 00:07:03,000
I definitely recommend that.

78
00:07:03,000 --> 00:07:19,000
There's, there's a bunch of different options there for the other microphone, but the idea would be that I would be able to use the nice microphone that I always use and then a nice microphone that I haven't used in a little while and then another nice microphone that I'm yet to buy.

79
00:07:19,000 --> 00:07:23,000
And I'd be able to record all these things for posterity.

80
00:07:23,000 --> 00:07:28,000
Yeah, looks like what I was thinking of is the road wireless go to.

81
00:07:28,000 --> 00:07:38,000
And that gets you the two small square mics with the single receiver, but you can get their love mics that would pop onto that.

82
00:07:38,000 --> 00:07:50,000
Yeah. And I thought about that. I thought about, I thought about just getting those are just getting the DJI or just getting road has a new one that'll plug into your phone.

83
00:07:50,000 --> 00:08:02,000
Sorry, the, the receiver will plug into your phone and then it has two smaller microphones that I mean clip on there even smaller than the squares.

84
00:08:02,000 --> 00:08:11,000
And you can clip those on and then record using their wireless technology to your phone.

85
00:08:11,000 --> 00:08:14,000
And those looked, those look pretty good.

86
00:08:14,000 --> 00:08:25,000
I, you know, it's just kind of one of those, I think the duo gives me the most flexibility without having to go all the way up to the pro.

87
00:08:25,000 --> 00:08:30,000
But we'll see.

88
00:08:30,000 --> 00:08:34,000
Maybe I'll have this for a while and go to the next one.

89
00:08:34,000 --> 00:08:39,000
Yes, holy cow.

90
00:08:39,000 --> 00:08:44,000
You're gonna sell something here?

91
00:08:44,000 --> 00:08:47,000
Me? No, I'm not telling anything.

92
00:08:47,000 --> 00:08:49,000
I didn't think so.

93
00:08:49,000 --> 00:08:52,000
Okay. Hey, what are you drinking?

94
00:08:52,000 --> 00:08:56,000
I am technically breaking the rules, I would say.

95
00:08:56,000 --> 00:09:03,000
Well, we've had three shows in a row where the beer just hasn't been that great.

96
00:09:03,000 --> 00:09:06,000
I mean, yeah.

97
00:09:06,000 --> 00:09:09,000
I couldn't, I couldn't do a fourth in a row.

98
00:09:09,000 --> 00:09:13,000
Okay. So I brought a go to good old Guinness.

99
00:09:13,000 --> 00:09:15,000
I don't think that's breaking the rules.

100
00:09:15,000 --> 00:09:17,000
Well, we've already done it before.

101
00:09:17,000 --> 00:09:20,000
Well, I don't think we have to do it every time, right?

102
00:09:20,000 --> 00:09:26,000
So, but I had to, I had to go with a guaranteed and I brought an extra just in case.

103
00:09:26,000 --> 00:09:28,000
A Shatterbock.

104
00:09:28,000 --> 00:09:29,000
I like it.

105
00:09:29,000 --> 00:09:35,000
I'm also breaking that rule, I guess, because I have my last dragon's milk.

106
00:09:35,000 --> 00:09:42,000
I am excited to drink this bad boy and then go try to find some more.

107
00:09:42,000 --> 00:09:45,000
That is a good one.

108
00:09:45,000 --> 00:09:50,000
Yeah. So here's to opening.

109
00:09:50,000 --> 00:09:52,000
I think we've read all these, right?

110
00:09:52,000 --> 00:09:55,000
Yeah, we've read the blurb before in a different show.

111
00:09:55,000 --> 00:09:58,000
All of the different...

112
00:09:58,000 --> 00:10:00,000
That did not sound good.

113
00:10:00,000 --> 00:10:02,000
That was a good one.

114
00:10:02,000 --> 00:10:05,000
That did not sound good.

115
00:10:05,000 --> 00:10:08,000
Good job.

116
00:10:08,000 --> 00:10:11,000
Yeah.

117
00:10:11,000 --> 00:10:14,000
So, cheers.

118
00:10:14,000 --> 00:10:15,000
Cheers.

119
00:10:15,000 --> 00:10:19,000
Tink.

120
00:10:19,000 --> 00:10:21,000
Oh, that's the good stuff.

121
00:10:21,000 --> 00:10:23,000
Oh my gosh.

122
00:10:23,000 --> 00:10:24,000
Okay. Yeah.

123
00:10:24,000 --> 00:10:27,000
I mean, when you don't...

124
00:10:27,000 --> 00:10:29,000
Are you okay over there?

125
00:10:29,000 --> 00:10:31,000
Your lights went off.

126
00:10:31,000 --> 00:10:34,000
You got a friend outside?

127
00:10:34,000 --> 00:10:39,000
Who knows what happened there?

128
00:10:39,000 --> 00:10:49,000
I'm picturing... This is not true, but I'm picturing your closet where your recording is having one of those buttons when the door gets closed.

129
00:10:49,000 --> 00:10:51,000
It turns off sometimes.

130
00:10:51,000 --> 00:10:52,000
It does?

131
00:10:52,000 --> 00:10:54,000
It does, really. It does.

132
00:10:54,000 --> 00:10:57,000
Yeah, if I go pull on that door enough, that light goes out.

133
00:10:57,000 --> 00:10:58,000
That's hilarious.

134
00:10:58,000 --> 00:11:04,000
Okay, well, yeah, not bad.

135
00:11:04,000 --> 00:11:11,000
I think probably for this show, we need to include in the notes a picture of at least your setup.

136
00:11:11,000 --> 00:11:14,000
Mine's not really anything that you want to look at, but...

137
00:11:14,000 --> 00:11:20,000
Well, mine looks like I know what I'm doing.

138
00:11:20,000 --> 00:11:23,000
Yeah. That's why I want a picture of it in the show.

139
00:11:23,000 --> 00:11:26,000
I'm very much still like, is this how this works?

140
00:11:26,000 --> 00:11:31,000
I guess the one thing I have done that I hadn't talked to you about,

141
00:11:31,000 --> 00:11:35,000
and so that maybe I'll have to take a whole new picture from the last one I sent you.

142
00:11:35,000 --> 00:11:36,000
Okay.

143
00:11:36,000 --> 00:11:39,000
But I had some of the equipment up on the desk.

144
00:11:39,000 --> 00:11:40,000
Uh-huh.

145
00:11:40,000 --> 00:11:46,000
And it was like, I really want to get that off the desk because I brought in the new monitors and the monitor controller.

146
00:11:46,000 --> 00:11:50,000
So I actually bought a vertical rack.

147
00:11:50,000 --> 00:11:51,000
Nice.

148
00:11:51,000 --> 00:11:53,000
Got it off here to the side of me.

149
00:11:53,000 --> 00:11:54,000
Nice.

150
00:11:54,000 --> 00:11:56,000
So the switch is in there.

151
00:11:56,000 --> 00:11:58,000
My power conditioner is in there.

152
00:11:58,000 --> 00:12:04,000
And then I've got a, I guess you'd say power distribution,

153
00:12:04,000 --> 00:12:10,000
but it has the toggles for the individual plugs on the back.

154
00:12:10,000 --> 00:12:11,000
Awesome.

155
00:12:11,000 --> 00:12:14,000
So I can turn those off and on.

156
00:12:14,000 --> 00:12:16,000
Depending on what you're...

157
00:12:16,000 --> 00:12:17,000
Depending on what I need.

158
00:12:17,000 --> 00:12:18,000
Huh.

159
00:12:18,000 --> 00:12:19,000
So...

160
00:12:19,000 --> 00:12:20,000
That's a good idea.

161
00:12:20,000 --> 00:12:23,000
I'll take an updated picture for what it's worth.

162
00:12:23,000 --> 00:12:24,000
So...

163
00:12:24,000 --> 00:12:28,000
But I think I'm set for a while now.

164
00:12:28,000 --> 00:12:33,000
Well, I mean...

165
00:12:33,000 --> 00:12:34,000
It's always amazing to me.

166
00:12:34,000 --> 00:12:36,000
This happens to me as a musician.

167
00:12:36,000 --> 00:12:39,000
I'm like, yes, that is my dream setup.

168
00:12:39,000 --> 00:12:43,000
I have my dream set up and then like 15 minutes later, I'm like...

169
00:12:43,000 --> 00:12:45,000
But...

170
00:12:45,000 --> 00:12:47,000
Yep.

171
00:12:47,000 --> 00:12:49,000
I really want that too.

172
00:12:49,000 --> 00:12:54,000
Yeah, the only thing that I would eventually change, but I'm in no rush,

173
00:12:54,000 --> 00:13:00,000
is I thought about either building something simple to go in here,

174
00:13:00,000 --> 00:13:06,000
or I could actually take this rack gear and actually have it mounted into the desk.

175
00:13:06,000 --> 00:13:10,000
Because I've looked at some of the purpose built furniture and it's nice, but it's expensive.

176
00:13:10,000 --> 00:13:13,000
Yeah, no, it for sure is.

177
00:13:13,000 --> 00:13:19,000
I was looking at a rack for underneath our TV and the living room.

178
00:13:19,000 --> 00:13:31,000
I used to just find 19 and a quarter inches apart from each other and drill those ears in.

179
00:13:31,000 --> 00:13:33,000
And then no big deal.

180
00:13:33,000 --> 00:13:39,000
But anyway, seems like it's harder than it used to be.

181
00:13:39,000 --> 00:13:42,000
Yep, I would agree.

182
00:13:42,000 --> 00:13:45,000
What was I going to say?

183
00:13:45,000 --> 00:13:48,000
Oh, okay.

184
00:13:48,000 --> 00:13:58,000
I'm wanting to jump in, but also I feel like I should just say for anybody listening who wants to talk more about the beer,

185
00:13:58,000 --> 00:14:01,000
this beer is really good.

186
00:14:01,000 --> 00:14:03,000
Yes.

187
00:14:03,000 --> 00:14:06,000
Because these beers are so good, we don't have anything else to say.

188
00:14:06,000 --> 00:14:09,000
We just want to drink the beer and enjoy it.

189
00:14:09,000 --> 00:14:12,000
Just moments of silence.

190
00:14:12,000 --> 00:14:18,000
A lot of just sitting here looking at each other, drinking beer.

191
00:14:18,000 --> 00:14:22,000
Well, not in that way though, the way you said that.

192
00:14:22,000 --> 00:14:24,000
Just looking at each other.

193
00:14:24,000 --> 00:14:28,000
Okay, well, we don't have to make it weird.

194
00:14:28,000 --> 00:14:30,000
Fine.

195
00:14:30,000 --> 00:14:34,000
Yep.

196
00:14:34,000 --> 00:14:39,000
All right, but let's go ahead and jump in then.

197
00:14:39,000 --> 00:14:43,000
Since we're going to make that weird, let's jump into the...

198
00:14:43,000 --> 00:14:45,000
What we're talking about today.

199
00:14:45,000 --> 00:14:47,000
This is a fun one.

200
00:14:47,000 --> 00:14:50,000
Yeah, we've got a lot here.

201
00:14:50,000 --> 00:14:55,000
There's probably at least two shows here if we tried to go through it all.

202
00:14:55,000 --> 00:14:57,000
So I don't think we're going to get it all tonight.

203
00:14:57,000 --> 00:15:00,000
I think this could be a series.

204
00:15:00,000 --> 00:15:05,000
Well, it could be, yeah, but we are going to kind of dip our toes into the world of AI.

205
00:15:05,000 --> 00:15:07,000
Yeah.

206
00:15:07,000 --> 00:15:09,000
So much AI.

207
00:15:09,000 --> 00:15:15,000
On another podcast that AI is anonymous Indians.

208
00:15:15,000 --> 00:15:19,000
Okay.

209
00:15:19,000 --> 00:15:23,000
I appreciate the humor in that.

210
00:15:23,000 --> 00:15:24,000
Yeah.

211
00:15:24,000 --> 00:15:29,000
Dots or feathers.

212
00:15:29,000 --> 00:15:31,000
Okay.

213
00:15:31,000 --> 00:15:36,000
So you got some definitions of AI to start us off here.

214
00:15:36,000 --> 00:15:37,000
Yeah.

215
00:15:37,000 --> 00:15:38,000
Let's go with that.

216
00:15:38,000 --> 00:15:40,000
Pulled it from a couple of different places.

217
00:15:40,000 --> 00:15:52,000
So from Britannica, they define AI as the ability of a digital computer or computer controlled robot to perform tasks commonly associated with intelligent beings.

218
00:15:52,000 --> 00:15:54,000
And I would agree with that.

219
00:15:54,000 --> 00:15:56,000
I don't have any issues with that definition.

220
00:15:56,000 --> 00:16:04,000
Oxford says the capacity of computers or other machines to exhibit or simulate intelligent behavior.

221
00:16:04,000 --> 00:16:05,000
No issues with that one.

222
00:16:05,000 --> 00:16:06,000
Sorry.

223
00:16:06,000 --> 00:16:25,000
Cambridge, the use or study of computer systems or machines that have some of the qualities that the human brain has such as the ability to interpret and produce language in a way that seems human, recognizable or create images, solve problems and learn from data supplied to them.

224
00:16:25,000 --> 00:16:30,000
I think that one's getting closer to what I would call it.

225
00:16:30,000 --> 00:16:37,000
And because it references qualities of the human brain, instantly do comes to mind.

226
00:16:37,000 --> 00:16:38,000
Yeah.

227
00:16:38,000 --> 00:16:44,000
For me, because of the way that he approached it, it wasn't the terminator version.

228
00:16:44,000 --> 00:16:49,000
It was humans using AI to control other humans.

229
00:16:49,000 --> 00:16:50,000
Yeah.

230
00:16:50,000 --> 00:16:59,000
And so it became illegal to create a machine in the likeness image or the ability to think like a human.

231
00:16:59,000 --> 00:17:19,000
And then the last one here comes from Google and they say artificial intelligence is a set of technologies that enable computers to perform a variety of advanced functions, including the ability to see, understand and translate spoken in written language, analyze data, make recommendations and more.

232
00:17:19,000 --> 00:17:22,000
And so that's, to me, that's very Google.

233
00:17:22,000 --> 00:17:24,000
That's very Google.

234
00:17:24,000 --> 00:17:25,000
Yes.

235
00:17:25,000 --> 00:17:31,000
So that's kind of the core definitions.

236
00:17:31,000 --> 00:17:34,000
I agree with these for the most part.

237
00:17:34,000 --> 00:17:40,000
I think I take more of a old school purist of what I would define AI.

238
00:17:40,000 --> 00:17:48,000
So if I take an AI system and I train it on all things football, history, players, types of plays,

239
00:17:48,000 --> 00:17:55,000
what's the best play for defense against this offense, everything you could ever want to know about football.

240
00:17:55,000 --> 00:18:03,000
And all of a sudden it starts teaching me basketball, baseball, tennis, talking about history of players.

241
00:18:03,000 --> 00:18:06,000
Okay, something happened because I didn't train it on that data.

242
00:18:06,000 --> 00:18:08,000
I didn't give it access to that data.

243
00:18:08,000 --> 00:18:09,000
Right.

244
00:18:09,000 --> 00:18:11,000
But somehow I pulled it all together.

245
00:18:11,000 --> 00:18:17,000
So to me, that would be how I would define artificial intelligence.

246
00:18:17,000 --> 00:18:18,000
Okay.

247
00:18:18,000 --> 00:18:25,000
It truly learned somehow on its own in a manner that I didn't give it the ability to do.

248
00:18:25,000 --> 00:18:31,000
You may have given it the root function of it, but you didn't specify that behavior or

249
00:18:31,000 --> 00:18:36,000
correct, provide it with that information.

250
00:18:36,000 --> 00:18:42,000
So, okay, let's talk about neural networks.

251
00:18:42,000 --> 00:18:43,000
Yeah.

252
00:18:43,000 --> 00:18:45,000
This is where it gets kind of weird.

253
00:18:45,000 --> 00:18:53,000
And again, we'll have this in the show notes, but neural networks kind of ties into machine

254
00:18:53,000 --> 00:18:57,000
language and deep learning.

255
00:18:57,000 --> 00:19:03,000
So the concept of a neural network, it's trying to mimic neurons in the human brain.

256
00:19:03,000 --> 00:19:04,000
Okay.

257
00:19:04,000 --> 00:19:07,000
And so you have three main layers.

258
00:19:07,000 --> 00:19:08,000
There's the input layer.

259
00:19:08,000 --> 00:19:15,600
So whether that's something that you're feeding into a video, a picture, audio, text of some

260
00:19:15,600 --> 00:19:17,920
kind, whatever that input may be.

261
00:19:17,920 --> 00:19:21,040
So you can imagine that coming in from the left.

262
00:19:21,040 --> 00:19:25,560
And then there is a layer next that's referred to as the hidden layer.

263
00:19:25,560 --> 00:19:32,000
So on a simple neural network, you may have one or two layers of there.

264
00:19:32,000 --> 00:19:35,900
And then on the far right, you ultimately have your output layer.

265
00:19:35,900 --> 00:19:40,720
And that's the data, the information, the picture, the video, whatever it ultimately

266
00:19:40,720 --> 00:19:44,520
outputs from the neural network.

267
00:19:44,520 --> 00:19:45,520
Okay.

268
00:19:45,520 --> 00:19:53,560
So, and then when you get into deep neural networks or deep learning, you get into lots

269
00:19:53,560 --> 00:19:57,920
of hidden layers in between.

270
00:19:57,920 --> 00:20:03,120
And that's what allows it to break down the input further and further and further, refine

271
00:20:03,120 --> 00:20:08,440
it, refine it, refine it to try to make sure that you're getting the ideal outcome that

272
00:20:08,440 --> 00:20:10,800
you're looking for in the end.

273
00:20:10,800 --> 00:20:16,680
Whereas just a simple neural network may struggle to give you what you're after.

274
00:20:16,680 --> 00:20:27,000
Well, a simple neural network is more, I mean, this is like the difference between, have

275
00:20:27,000 --> 00:20:31,440
you ever seen somebody who, sorry, this is probably really primitive.

276
00:20:31,440 --> 00:20:42,520
And I mean, we all had to take the computer class in high school, I'm assuming, the computer

277
00:20:42,520 --> 00:20:50,280
literacy class in high school that you had to learn how to use Word and Excel and PowerPoint.

278
00:20:50,280 --> 00:20:57,240
But when you initially started doing like in power or not in PowerPoint in Excel, when

279
00:20:57,240 --> 00:21:12,160
you started using the different cells to create, you know, to do equations and things like

280
00:21:12,160 --> 00:21:19,600
that, you're using formulas and you're seeing that like, okay, and then maybe like, so that's

281
00:21:19,600 --> 00:21:24,600
one hidden layer is it's not just giving you the information back that you're putting

282
00:21:24,600 --> 00:21:30,040
into it, but it is doing something with that information, but it's still a function that

283
00:21:30,040 --> 00:21:32,480
you are telling it to do.

284
00:21:32,480 --> 00:21:33,760
It's just not visible to you.

285
00:21:33,760 --> 00:21:35,240
So that's like one hidden layer.

286
00:21:35,240 --> 00:21:42,840
And then you had, you had to, at some point in the class, you had to take that information

287
00:21:42,840 --> 00:21:50,720
put in there, make sure that it did what it was supposed to do in terms of the formula

288
00:21:50,720 --> 00:21:53,240
or the function that it was supposed to have.

289
00:21:53,240 --> 00:22:00,360
And then you had to, you know, it could create like a pie chart or a diagram for you or something

290
00:22:00,360 --> 00:22:01,360
like that.

291
00:22:01,360 --> 00:22:05,840
That's like another layer of hidden layer or something like that inside of it, but it's

292
00:22:05,840 --> 00:22:09,120
still pretty easy for you to see what's going on.

293
00:22:09,120 --> 00:22:14,960
These deep neural networks are more like, did you ever see Randy Collins Excel sheets?

294
00:22:14,960 --> 00:22:17,600
Oh, yeah.

295
00:22:17,600 --> 00:22:22,480
So if you ever met one of those guys that like, they're just wizards on Excel and all

296
00:22:22,480 --> 00:22:29,280
of a sudden this thing is doing like amortizations for you that you just like, that's referencing

297
00:22:29,280 --> 00:22:32,720
a different sheet or whatever they call it.

298
00:22:32,720 --> 00:22:34,400
I think sheets actually call them sheets.

299
00:22:34,400 --> 00:22:35,400
So sorry about that.

300
00:22:35,400 --> 00:22:36,400
Yeah, it's a whole different Excel.

301
00:22:36,400 --> 00:22:37,400
My bad.

302
00:22:37,400 --> 00:22:42,440
But yeah, like different workbooks and you've got macros going on and things like that.

303
00:22:42,440 --> 00:22:49,480
You've got these massive formulas that are being put in, but you can't see any of it.

304
00:22:49,480 --> 00:22:51,720
That's more like deep neural networks.

305
00:22:51,720 --> 00:22:52,720
Oh, yeah.

306
00:22:52,720 --> 00:23:03,880
So you may like, we've all put some input into a computer and gotten a response.

307
00:23:03,880 --> 00:23:09,040
You know, you've done the input layer and then received the output layer and it feels

308
00:23:09,040 --> 00:23:11,320
really one for one.

309
00:23:11,320 --> 00:23:15,080
Yeah.

310
00:23:15,080 --> 00:23:27,160
It's just the like order of magnitude of those hidden layers is now much greater.

311
00:23:27,160 --> 00:23:38,120
And that's what I've seen with artificial intelligence is this, it's doing a lot that

312
00:23:38,120 --> 00:23:44,080
we can't see, but it has still been programmed to do those things.

313
00:23:44,080 --> 00:23:52,960
Where we actually have artificial intelligence that's further than that.

314
00:23:52,960 --> 00:23:58,480
That's when people are starting to get nervous and you can start to see that now because

315
00:23:58,480 --> 00:23:59,600
it looks like magic.

316
00:23:59,600 --> 00:24:00,600
Yeah.

317
00:24:00,600 --> 00:24:05,200
I like him what they're doing to almost filling a database full of everything you could ever

318
00:24:05,200 --> 00:24:11,200
want and then asking it to go retrieve that data and put it together and it can just do

319
00:24:11,200 --> 00:24:17,240
it way faster than any person could ever attempt to do.

320
00:24:17,240 --> 00:24:19,240
Yeah.

321
00:24:19,240 --> 00:24:20,740
Yeah.

322
00:24:20,740 --> 00:24:23,200
It's because it's not breathing.

323
00:24:23,200 --> 00:24:24,200
Correct.

324
00:24:24,200 --> 00:24:27,800
Don't have time for breathing.

325
00:24:27,800 --> 00:24:28,800
Yeah.

326
00:24:28,800 --> 00:24:29,800
Yeah.

327
00:24:29,800 --> 00:24:34,320
So a few more things here on neural networks.

328
00:24:34,320 --> 00:24:37,680
We call them neural networks for a reason.

329
00:24:37,680 --> 00:24:38,680
Yes.

330
00:24:38,680 --> 00:24:39,680
This is mimicking.

331
00:24:39,680 --> 00:24:41,480
Absolutely.

332
00:24:41,480 --> 00:24:46,160
So each one of these, I know you all can't see the picture, but you will if you look at

333
00:24:46,160 --> 00:24:50,440
the show notes and we'll definitely put this in the chapter order as well.

334
00:24:50,440 --> 00:24:51,440
Once again.

335
00:24:51,440 --> 00:24:52,440
But each of...

336
00:24:52,440 --> 00:24:53,440
Podcast two, baby.

337
00:24:53,440 --> 00:24:54,440
Podcasting 2.0.

338
00:24:54,440 --> 00:24:56,720
Get you a good app.

339
00:24:56,720 --> 00:25:02,080
So each of these are referred, these colored dots are referred to as a node and each node

340
00:25:02,080 --> 00:25:05,920
is meant to try to mimic the neuron on the brain.

341
00:25:05,920 --> 00:25:10,240
So they say here neurons are the basic units that receive inputs.

342
00:25:10,240 --> 00:25:15,040
Each neuron is governed by a threshold and an activation function.

343
00:25:15,040 --> 00:25:20,960
The connections in between are links between neurons that carry information regulated by

344
00:25:20,960 --> 00:25:22,440
weights and biases.

345
00:25:22,440 --> 00:25:25,320
I want to stop real quick on weights and biases.

346
00:25:25,320 --> 00:25:30,280
That's where this kind of gets in interesting and that's some of the pieces that you can

347
00:25:30,280 --> 00:25:34,640
kind of change and define for this neural network.

348
00:25:34,640 --> 00:25:35,640
Okay.

349
00:25:35,640 --> 00:25:40,680
So you have an impact on how it perceives and what it gives you in the end.

350
00:25:40,680 --> 00:25:48,680
So depending on what an individual is doing, they may put their own bias into those, I

351
00:25:48,680 --> 00:25:54,040
know bias is one of the settings, but your personal opinion could determine how you set

352
00:25:54,040 --> 00:25:56,400
the weights and the biases.

353
00:25:56,400 --> 00:26:04,960
And so you could, in theory, get an output that you didn't mean to get, but your own

354
00:26:04,960 --> 00:26:10,680
what you were putting in drove it to that outcome versus another person can do it.

355
00:26:10,680 --> 00:26:15,640
It could be driven by their thoughts and opinions on how they set the weights and biases and

356
00:26:15,640 --> 00:26:20,360
get a completely different outcome from the exact same neural network.

357
00:26:20,360 --> 00:26:25,040
So you've got to try to be careful in what you're doing and how you're setting your weights

358
00:26:25,040 --> 00:26:31,440
and your biases so you don't ultimately skew some of the output that you get.

359
00:26:31,440 --> 00:26:38,560
So that's a part that I think makes it feel more human to everybody because there's a

360
00:26:38,560 --> 00:26:41,720
mirror that's happening inside of it.

361
00:26:41,720 --> 00:26:48,840
And every time somebody uses chat GPT and says, no, I want it a little bit more like

362
00:26:48,840 --> 00:26:50,360
this.

363
00:26:50,360 --> 00:26:53,120
It's moving a weight.

364
00:26:53,120 --> 00:26:54,920
Yep.

365
00:26:54,920 --> 00:26:59,520
It's saying this user prefers this.

366
00:26:59,520 --> 00:27:00,960
Yep.

367
00:27:00,960 --> 00:27:08,360
Which is why next time it may look closer, faster.

368
00:27:08,360 --> 00:27:09,360
Propagation functions.

369
00:27:09,360 --> 00:27:12,640
There you go.

370
00:27:12,640 --> 00:27:17,520
That is propagation functions, mechanisms that help process and transfer data across

371
00:27:17,520 --> 00:27:20,000
layers of neurons.

372
00:27:20,000 --> 00:27:24,560
And that's really going to get into more when you're in a deep learning because you have

373
00:27:24,560 --> 00:27:28,360
so many potential hidden layers in between.

374
00:27:28,360 --> 00:27:34,400
And then the learning rule, the method that adjusts weights and biases over time to improve

375
00:27:34,400 --> 00:27:35,400
accuracy.

376
00:27:35,400 --> 00:27:39,440
Is what we were just talking about.

377
00:27:39,440 --> 00:27:40,440
Absolutely.

378
00:27:40,440 --> 00:27:41,440
Okay.

379
00:27:41,440 --> 00:27:49,160
And so the next as we go through this long dissertation, learning in neural networks

380
00:27:49,160 --> 00:27:53,040
follows a structured three-stage process.

381
00:27:53,040 --> 00:27:58,840
First stage is input computation data is fed into the network, output generation based

382
00:27:58,840 --> 00:28:03,880
on the current parameters, the network generates an output and then iterative refinement.

383
00:28:03,880 --> 00:28:05,920
I think I said that right.

384
00:28:05,920 --> 00:28:06,920
It's iterative.

385
00:28:06,920 --> 00:28:07,920
You did.

386
00:28:07,920 --> 00:28:12,840
The network refines its output by adjusting weights and biases gradually improving its

387
00:28:12,840 --> 00:28:15,960
performance on diverse tasks.

388
00:28:15,960 --> 00:28:18,320
So similar to what we already said.

389
00:28:18,320 --> 00:28:19,320
Yeah.

390
00:28:19,320 --> 00:28:29,680
What's interesting to me is the like, what I seem to be like, man, I hear it.

391
00:28:29,680 --> 00:28:35,520
I think it's one of the things from people that are using AI that they like that they

392
00:28:35,520 --> 00:28:40,360
value the most is that iterative refinement.

393
00:28:40,360 --> 00:28:49,360
But the fact that even in itself, it's been loaded with the correct propagation functions

394
00:28:49,360 --> 00:28:57,560
to be able to understand what you mean when you say, I need it to be less flowery language

395
00:28:57,560 --> 00:29:02,800
or make that simpler or embellish that.

396
00:29:02,800 --> 00:29:12,960
I mean, it's not well embellish and simpler probably easier functions to be able to write

397
00:29:12,960 --> 00:29:13,960
in there.

398
00:29:13,960 --> 00:29:23,920
But if you say less flowery language, the machine learning can actually understand what

399
00:29:23,920 --> 00:29:24,920
you mean by that.

400
00:29:24,920 --> 00:29:33,240
And then it like that iterative refinement is trying to take you through a weight, a

401
00:29:33,240 --> 00:29:39,320
bias that's different in order to find the right one for you.

402
00:29:39,320 --> 00:29:45,720
And everybody like that's what I see people going crazy over is I don't.

403
00:29:45,720 --> 00:29:51,400
And it's probably because it's what humans do worst whenever you've got somebody working

404
00:29:51,400 --> 00:29:55,400
for you, it's really hard for them to break.

405
00:29:55,400 --> 00:30:01,240
It seems like to me, it's really hard for people to break their initial idea in order

406
00:30:01,240 --> 00:30:04,480
to come up with something new as a creative.

407
00:30:04,480 --> 00:30:15,120
I know like that's, that's the there's like significant cost to doing that.

408
00:30:15,120 --> 00:30:20,480
Like it, even when I try to just do it myself, even when I try to say like, oh, now put that

409
00:30:20,480 --> 00:30:23,080
like you've you've created that put that aside.

410
00:30:23,080 --> 00:30:25,240
Let's try it a different way.

411
00:30:25,240 --> 00:30:30,800
Like there's a fatigue involved with that that computers don't have and just the ability

412
00:30:30,800 --> 00:30:36,680
to cycle it to say like, oh, we're going to try a different one.

413
00:30:36,680 --> 00:30:38,120
Oh, we're going to try a different one.

414
00:30:38,120 --> 00:30:40,000
Oh, we're going to try a different one.

415
00:30:40,000 --> 00:30:44,560
So much to the point where that that seems to be the majority of the time what they're

416
00:30:44,560 --> 00:30:50,920
what they give you, especially in the art sides of chat, or not chat, GBT, but in the

417
00:30:50,920 --> 00:30:57,480
art sides of AI, they, they are giving you multiple options from the outset.

418
00:30:57,480 --> 00:30:58,480
Yep.

419
00:30:58,480 --> 00:31:02,480
Anyway, yep.

420
00:31:02,480 --> 00:31:06,200
There's going to be a future and people who are excellent prompt jockeys.

421
00:31:06,200 --> 00:31:08,200
Prompt jockey.

422
00:31:08,200 --> 00:31:09,200
Yep.

423
00:31:09,200 --> 00:31:10,200
Okay.

424
00:31:10,200 --> 00:31:12,200
So next, continuing on.

425
00:31:12,200 --> 00:31:19,560
In an adaptive learning environment, the neural network is exposed to a simulated scenario,

426
00:31:19,560 --> 00:31:24,520
a data set, parameters such as weights and biases are updated in response to the new

427
00:31:24,520 --> 00:31:30,320
data or conditions with each adjustment, the network's response evolves, allowing it to

428
00:31:30,320 --> 00:31:35,920
adapt effectively to different tasks or environments.

429
00:31:35,920 --> 00:31:40,920
So that's the taking those weights and then applying it to a different function.

430
00:31:40,920 --> 00:31:46,800
Yeah, or a new, new data that was input into the model.

431
00:31:46,800 --> 00:31:50,360
So, the more you use it, the better it gets.

432
00:31:50,360 --> 00:31:51,360
Yeah.

433
00:31:51,360 --> 00:31:53,880
So next we go into types of neural networks.

434
00:31:53,880 --> 00:32:00,000
This is where I'm still trying to learn myself.

435
00:32:00,000 --> 00:32:04,160
So there's the feedforward artificial neural network.

436
00:32:04,160 --> 00:32:10,320
It says, as the name suggests, a feedforward artificial network is when data moves in one

437
00:32:10,320 --> 00:32:16,480
direction between the input and output nodes, data moves forward through layers of nodes

438
00:32:16,480 --> 00:32:20,080
and won't cycle backwards through the same layers.

439
00:32:20,080 --> 00:32:25,280
Although there may be many different layers with many different nodes, the one way movement

440
00:32:25,280 --> 00:32:29,640
of data makes feedforward neural networks relatively simple.

441
00:32:29,640 --> 00:32:34,960
Feedforward artificial neural network models are mainly used for simplistic classification

442
00:32:34,960 --> 00:32:36,720
problems.

443
00:32:36,720 --> 00:32:42,440
Models will perform beyond the scope of a traditional machine learning model, but do not meet the

444
00:32:42,440 --> 00:32:48,960
level of abstraction found in deep learning model or in a deep learning model.

445
00:32:48,960 --> 00:32:49,960
Interesting.

446
00:32:49,960 --> 00:32:54,920
So yeah, when I was reading that initially and it was talking about, it won't cycle backwards

447
00:32:54,920 --> 00:32:55,920
through the same layers.

448
00:32:55,920 --> 00:33:00,800
I'm like, oh, I thought it all just left, right, left, right, left, right, keep going

449
00:33:00,800 --> 00:33:01,800
people.

450
00:33:01,800 --> 00:33:02,800
Yeah.

451
00:33:02,800 --> 00:33:03,800
Evidently not.

452
00:33:03,800 --> 00:33:05,320
Apparently not.

453
00:33:05,320 --> 00:33:08,320
So, but that went into me.

454
00:33:08,320 --> 00:33:09,320
That's pretty straightforward.

455
00:33:09,320 --> 00:33:11,440
Goes in on the left side, comes out on the right side.

456
00:33:11,440 --> 00:33:12,440
Oh, no.

457
00:33:12,440 --> 00:33:14,480
No, no, nothing goes backwards.

458
00:33:14,480 --> 00:33:15,480
No vice-versa.

459
00:33:15,480 --> 00:33:17,400
Yeah, it doesn't go backwards.

460
00:33:17,400 --> 00:33:18,400
Yep.

461
00:33:18,400 --> 00:33:19,400
Okay.

462
00:33:19,400 --> 00:33:24,320
So next is the perceptron and multilayer perceptron neural networks.

463
00:33:24,320 --> 00:33:30,040
A perceptron is one of the earliest and simplest models of a neuron.

464
00:33:30,040 --> 00:33:37,480
A perceptron model is a binary classifier separating data into two different classifications.

465
00:33:37,480 --> 00:33:43,960
As a linear model, it is one of the simplest examples of a type of artificial neural network.

466
00:33:43,960 --> 00:33:51,880
Multilayer perceptron artificial neural networks or networks add complexity and density with

467
00:33:51,880 --> 00:33:57,920
a capacity for many hidden layers between the input and output layer.

468
00:33:57,920 --> 00:34:04,400
Each individual node on a specific layer is connected to every node on the next layer.

469
00:34:04,400 --> 00:34:09,720
This means multilayer perceptron models are fully connected networks and can be leveraged

470
00:34:09,720 --> 00:34:12,080
for deep learning.

471
00:34:12,080 --> 00:34:17,640
And they're used for more complex problems and tasks such as complex classification or

472
00:34:17,640 --> 00:34:23,720
voice recognition, cause of the model's depth and complexity, processing and model maintenance

473
00:34:23,720 --> 00:34:26,800
can be resource and time consuming.

474
00:34:26,800 --> 00:34:27,800
Interesting.

475
00:34:27,800 --> 00:34:28,800
Yeah.

476
00:34:28,800 --> 00:34:34,520
So it's intriguing that you've got your initial input.

477
00:34:34,520 --> 00:34:39,560
It'll have each of those nodes connects to every other node in the next layer and so

478
00:34:39,560 --> 00:34:40,560
on.

479
00:34:40,560 --> 00:34:46,240
So from layer to layer to layer, all nodes connect to each other.

480
00:34:46,240 --> 00:34:47,240
Yeah.

481
00:34:47,240 --> 00:34:49,480
So you just, yeah.

482
00:34:49,480 --> 00:34:53,160
There's not just a one for one translation there.

483
00:34:53,160 --> 00:34:54,480
Or one for two.

484
00:34:54,480 --> 00:34:55,480
Yeah.

485
00:34:55,480 --> 00:35:01,280
I think the one for two is probably where that made sense to me before.

486
00:35:01,280 --> 00:35:06,480
But all of them being connected, that's, yeah.

487
00:35:06,480 --> 00:35:11,120
So then you jump into radio basis function artificial neural networks.

488
00:35:11,120 --> 00:35:12,120
Yeah.

489
00:35:12,120 --> 00:35:13,120
Radio basis.

490
00:35:13,120 --> 00:35:14,120
Goodness.

491
00:35:14,120 --> 00:35:15,560
You read it or me?

492
00:35:15,560 --> 00:35:17,200
No, you read it.

493
00:35:17,200 --> 00:35:20,000
I was saying goodness because that is a mouthful.

494
00:35:20,000 --> 00:35:21,000
Yes.

495
00:35:21,000 --> 00:35:27,320
So radio basis function neural networks usually have an input layer, a layer with radio basis

496
00:35:27,320 --> 00:35:32,240
function nodes with different parameters and an output layer.

497
00:35:32,240 --> 00:35:37,320
Models can be used to perform classification regression for time series and to control

498
00:35:37,320 --> 00:35:39,120
systems.

499
00:35:39,120 --> 00:35:44,800
Radio basis functions calculate the absolute value between a center point and a given point.

500
00:35:44,800 --> 00:35:50,240
In the case of classification, a radio basis function calculates the distance between an

501
00:35:50,240 --> 00:35:53,040
input and a learned classification.

502
00:35:53,040 --> 00:35:58,720
If the input is closest to a specific tag, it is classified as such.

503
00:35:58,720 --> 00:36:04,360
A common use for radio basis function neural networks is in system control such as systems

504
00:36:04,360 --> 00:36:07,840
that control power restoration after a power cut.

505
00:36:07,840 --> 00:36:13,720
The artificial neural network can understand the priority order to restoring power, prioritizing

506
00:36:13,720 --> 00:36:19,360
repairs to the greatest number of people or core services.

507
00:36:19,360 --> 00:36:24,640
That's interesting because it sounded like, in that first part, it sounded like estimation.

508
00:36:24,640 --> 00:36:27,960
Yes, it did.

509
00:36:27,960 --> 00:36:36,880
But the fact that it uses estimation to then do prioritization is, that's an interesting,

510
00:36:36,880 --> 00:36:41,840
I don't think I would have made that connection.

511
00:36:41,840 --> 00:36:45,080
Obviously, that's really funny.

512
00:36:45,080 --> 00:36:50,040
And then they reference tags there because that's important because in certain models,

513
00:36:50,040 --> 00:36:55,800
when you're inputting the data, you can associate a tag to that data.

514
00:36:55,800 --> 00:37:01,160
For as it's learning, it tries to determine, is it coming as close to that tag as it needs

515
00:37:01,160 --> 00:37:04,160
to be as it tries to learn and understand.

516
00:37:04,160 --> 00:37:05,160
Right.

517
00:37:05,160 --> 00:37:06,160
Is this close enough?

518
00:37:06,160 --> 00:37:07,160
Correct.

519
00:37:07,160 --> 00:37:11,320
So, the idea of this restoring power is kind of intriguing.

520
00:37:11,320 --> 00:37:12,320
Yeah.

521
00:37:12,320 --> 00:37:19,160
I don't know if good old Encore or TXU do it this way, but it's intriguing that it could

522
00:37:19,160 --> 00:37:21,760
go, okay, we've got a power outage.

523
00:37:21,760 --> 00:37:28,080
This is the order of operations to get us back up and running, minimizing the input,

524
00:37:28,080 --> 00:37:30,680
minimizing the effect on the people who are without power.

525
00:37:30,680 --> 00:37:31,680
Yeah.

526
00:37:31,680 --> 00:37:36,480
And that would obviously be way faster than a human sitting there trying to figure this

527
00:37:36,480 --> 00:37:39,160
out.

528
00:37:39,160 --> 00:37:43,920
I mean, yeah, most humans.

529
00:37:43,920 --> 00:37:44,920
Yes.

530
00:37:44,920 --> 00:37:49,800
There's some humans I trust more than a computer to make those calls.

531
00:37:49,800 --> 00:37:52,680
Yeah, there are not many of them though.

532
00:37:52,680 --> 00:37:53,680
Okay.

533
00:37:53,680 --> 00:37:54,680
Recurrent neural networks.

534
00:37:54,680 --> 00:37:55,680
Yes.

535
00:37:55,680 --> 00:38:00,720
Recurrent neural networks are a powerful tool when a model is designed to process sequential

536
00:38:00,720 --> 00:38:01,720
data.

537
00:38:01,720 --> 00:38:06,680
The model will move data forward and loop it backwards to previous steps in the artificial

538
00:38:06,680 --> 00:38:11,520
neural network to best achieve a task and improve predictions.

539
00:38:11,520 --> 00:38:17,760
The layers between the input and output layers are recurrent, and that relevant information

540
00:38:17,760 --> 00:38:20,400
is looped back and retained.

541
00:38:20,400 --> 00:38:25,360
Memory of outputs from a layer is looped back into the input where it is held to improve

542
00:38:25,360 --> 00:38:28,360
the process for the next input.

543
00:38:28,360 --> 00:38:33,200
The flow of data is similar to feed forward artificial neural networks, but each node

544
00:38:33,200 --> 00:38:37,320
will retain information needed to improve each step.

545
00:38:37,320 --> 00:38:41,960
Because of this, models can better understand the context of an input and refine the prediction

546
00:38:41,960 --> 00:38:43,440
of an output.

547
00:38:43,440 --> 00:38:49,480
For example, a predictive text system may use memory as a previous word and a string of

548
00:38:49,480 --> 00:38:53,240
words to better predict the outcome of the next word.

549
00:38:53,240 --> 00:38:58,040
A recurrent artificial neural network would be better suited to understand the sentiment

550
00:38:58,040 --> 00:39:03,400
behind a whole sentence compared to more traditional machine learning models.

551
00:39:03,400 --> 00:39:08,800
Recurrent neural networks are also used within sequence-to-sequence models, which are used

552
00:39:08,800 --> 00:39:11,600
for natural language processing.

553
00:39:11,600 --> 00:39:17,320
Think of Dragon naturally speaking, if anybody remembers that one.

554
00:39:17,320 --> 00:39:22,400
Two recurrent neural networks are used within these models, which consist of a simultaneous

555
00:39:22,400 --> 00:39:24,600
encoder and decoder.

556
00:39:24,600 --> 00:39:31,160
These models are used for reactive chatbots, translating language, or to summarize documents.

557
00:39:31,160 --> 00:39:36,480
I mean, I understand what it's saying inside of what these are used for, but this immediately

558
00:39:36,480 --> 00:39:42,760
made me think of predictive text, which has been one of the greatest examples to me of

559
00:39:42,760 --> 00:39:46,120
why I'm not scared of AI.

560
00:39:46,120 --> 00:39:48,120
Yes.

561
00:39:48,120 --> 00:39:55,440
Super great at saying, oh, there's probably going to be a noun next.

562
00:39:55,440 --> 00:39:57,400
You want this noun?

563
00:39:57,400 --> 00:39:59,320
You want this preposition?

564
00:39:59,320 --> 00:40:06,160
You're not really good at understanding what else is going on inside of it.

565
00:40:06,160 --> 00:40:08,680
The more it gets looped back, the better it's supposed to be.

566
00:40:08,680 --> 00:40:17,320
I do appreciate that it's retaining those sequences and measuring them against the one

567
00:40:17,320 --> 00:40:26,320
before it, the frequency and what then turns into, I guess, priority of language, though

568
00:40:26,320 --> 00:40:27,680
I wouldn't call it priority.

569
00:40:27,680 --> 00:40:35,720
It's just more about what's normal in any given sentence.

570
00:40:35,720 --> 00:40:39,600
The more you sound like a computer, the more the computer can guess what you're going to

571
00:40:39,600 --> 00:40:40,600
say.

572
00:40:40,600 --> 00:40:43,520
True.

573
00:40:43,520 --> 00:40:54,840
I did think that this one was going to start, going to be where we started talking about

574
00:40:54,840 --> 00:41:03,640
I guess because of the fact that it is doing the looping backwards into itself.

575
00:41:03,640 --> 00:41:09,800
This is looping backwards into itself is why we get the horse with the five legs in the

576
00:41:09,800 --> 00:41:10,800
drawings.

577
00:41:10,800 --> 00:41:13,800
Yes, in humans with six fingers.

578
00:41:13,800 --> 00:41:15,800
Yeah, right.

579
00:41:15,800 --> 00:41:19,040
Anyway, modular neural networks.

580
00:41:19,040 --> 00:41:22,680
This is our last one in this grouping.

581
00:41:22,680 --> 00:41:26,920
Modular artificial neural network consists of a series of networks or components that

582
00:41:26,920 --> 00:41:31,080
work together, though independently to achieve a task.

583
00:41:31,080 --> 00:41:35,880
A complex task can therefore be broken down into smaller components.

584
00:41:35,880 --> 00:41:41,080
If applied to data processing or the computing process, the speed of the processing will

585
00:41:41,080 --> 00:41:45,360
be increased as smaller components can work in tandem.

586
00:41:45,360 --> 00:41:50,880
Each component network is performing a different subtask which when combined completes the

587
00:41:50,880 --> 00:41:53,240
overall task and output.

588
00:41:53,240 --> 00:41:59,000
This type of artificial neural network is beneficial as it can make complex processing

589
00:41:59,000 --> 00:42:03,920
more efficient and can be applied to a range of environments.

590
00:42:03,920 --> 00:42:13,240
This is like the idea being, I think, the idea being that instead of just nodes being

591
00:42:13,240 --> 00:42:20,200
connected together in this matrix of nodes that are all talking to each other and everything,

592
00:42:20,200 --> 00:42:27,720
you're getting whole networks connected to each other in the same type of matrix.

593
00:42:27,720 --> 00:42:42,000
So it gets exponentially more complex but also more useful, I guess, inside of that.

594
00:42:42,000 --> 00:42:46,520
Yeah, I kind of liken this to the SETI at home project.

595
00:42:46,520 --> 00:42:47,520
Yeah.

596
00:42:47,520 --> 00:42:53,080
Because you've got all these people who've opted in with the software and so you have

597
00:42:53,080 --> 00:42:59,280
all these computers working together, crunching their piece of the puzzle to all optimally

598
00:42:59,280 --> 00:43:01,480
put it all back together.

599
00:43:01,480 --> 00:43:03,160
Yeah.

600
00:43:03,160 --> 00:43:17,000
And I mean, this is the micro-trip at work in terms of, I guess, the same type of thinking

601
00:43:17,000 --> 00:43:18,000
going into it.

602
00:43:18,000 --> 00:43:22,640
Like if you're breaking these processes apart, if you're saying like, no, I'm going to, you're

603
00:43:22,640 --> 00:43:26,800
going to focus on this and this one's going to focus on this, but we're still going to

604
00:43:26,800 --> 00:43:34,040
be like everything's going to be connected, then you're not running into cross traffic.

605
00:43:34,040 --> 00:43:41,600
You're waiting like everything gets done and then gets put into the same either line of

606
00:43:41,600 --> 00:43:50,720
networks or you are then able to run it through more seamlessly.

607
00:43:50,720 --> 00:43:58,400
All of this to just try to create a human.

608
00:43:58,400 --> 00:44:00,120
Or give me a summary of a book.

609
00:44:00,120 --> 00:44:01,120
Yeah.

610
00:44:01,120 --> 00:44:06,720
You got a couple different definitions here, or not definitions, but like, I guess, takes

611
00:44:06,720 --> 00:44:13,840
on these different parts of AI.

612
00:44:13,840 --> 00:44:14,840
Yeah.

613
00:44:14,840 --> 00:44:22,880
So, in looking into this deep learning, machine learning and neural networks all kind of got

614
00:44:22,880 --> 00:44:28,320
blurred in the way people trying to explain it all.

615
00:44:28,320 --> 00:44:32,520
They all kind of work together.

616
00:44:32,520 --> 00:44:39,080
It's all, I guess you could call it sub layers of the larger context of AI.

617
00:44:39,080 --> 00:44:41,280
Yeah, but they are independent.

618
00:44:41,280 --> 00:44:43,480
They're not out of the same thing.

619
00:44:43,480 --> 00:44:44,480
Correct.

620
00:44:44,480 --> 00:44:47,720
So, next we'll jump in the machine learning.

621
00:44:47,720 --> 00:44:52,960
It's focused on enabling computers and machines to imitate the way that humans learn to perform

622
00:44:52,960 --> 00:44:58,680
tasks autonomously and to improve their performance and accuracy through experience and exposure

623
00:44:58,680 --> 00:45:00,760
to more data.

624
00:45:00,760 --> 00:45:05,560
Supervised machine learning models are trained with labeled data sets, which allow the models

625
00:45:05,560 --> 00:45:08,560
to learn and grow more accurate over time.

626
00:45:08,560 --> 00:45:13,880
For example, an algorithm would be trained with pictures of dogs and other things all

627
00:45:13,880 --> 00:45:19,320
labeled by humans, and the machine would learn ways to identify pictures of dogs on

628
00:45:19,320 --> 00:45:21,120
its own.

629
00:45:21,120 --> 00:45:25,640
Supervised machine learning is the most common type used today.

630
00:45:25,640 --> 00:45:30,720
Unsupervised machine learning, a program looks for patterns and unlabeled data.

631
00:45:30,720 --> 00:45:35,400
Unsupervised machine learning can find patterns or trends that people aren't explicitly looking

632
00:45:35,400 --> 00:45:36,560
for.

633
00:45:36,560 --> 00:45:42,600
For example, an unsupervised machine learning program could look through online sales data

634
00:45:42,600 --> 00:45:47,200
and identify different types of clients making purchases.

635
00:45:47,200 --> 00:45:52,480
And then last, reinforcement machine learning trains machines through trial and error to

636
00:45:52,480 --> 00:45:56,720
take the best action by establishing a reward system.

637
00:45:56,720 --> 00:46:01,040
Reinforcement learning can train models to play games or train autonomous vehicles to

638
00:46:01,040 --> 00:46:06,560
drive by telling the machine when it made the right decisions, which helps it learn

639
00:46:06,560 --> 00:46:10,040
over time what actions it should take.

640
00:46:10,040 --> 00:46:12,840
Yep.

641
00:46:12,840 --> 00:46:18,640
Have you seen the, not the derailist, but have you seen the recruit?

642
00:46:18,640 --> 00:46:20,360
I don't have it.

643
00:46:20,360 --> 00:46:21,360
Okay.

644
00:46:21,360 --> 00:46:27,280
So there's a, now I can't remember which, it's probably Netflix.

645
00:46:27,280 --> 00:46:32,440
There's a Netflix show called The Recruit.

646
00:46:32,440 --> 00:46:42,600
Lawyer, like 20 something lawyer is in the CIA now, but he's in, he's in like legal aid,

647
00:46:42,600 --> 00:46:47,920
I guess they're not legal aid, legal, whatever you call it for the CIA.

648
00:46:47,920 --> 00:46:56,000
And one of the, one of the problems that they, actually one of his colleagues gets sent to

649
00:46:56,000 --> 00:47:04,600
deal with is a interrogation robot that's, they're using machine learning to try to make

650
00:47:04,600 --> 00:47:15,200
a robot that can interrogate and it, because it's part of the CIA has been given some enhanced

651
00:47:15,200 --> 00:47:24,000
interrogation options or whatever, but they forgot to program it with the information

652
00:47:24,000 --> 00:47:32,840
that the people or the individual it is interrogating is not a robot.

653
00:47:32,840 --> 00:47:39,440
So like the first interrogation it does, it rips a guy's arm.

654
00:47:39,440 --> 00:47:45,280
Anyway, it was, it was pretty funny whenever it was like, so you didn't program it to know

655
00:47:45,280 --> 00:47:50,040
that it's not interrogating another robot.

656
00:47:50,040 --> 00:47:54,360
And the guy was like, it's way more complicated than that.

657
00:47:54,360 --> 00:47:55,360
But yes,

658
00:47:55,360 --> 00:48:01,640
They forgot the simple rules, do no harm to people.

659
00:48:01,640 --> 00:48:04,760
You know, it didn't think it was doing harm to a person.

660
00:48:04,760 --> 00:48:07,200
Thought it was interrogating another robot.

661
00:48:07,200 --> 00:48:08,200
Yeah.

662
00:48:08,200 --> 00:48:10,200
See, made a mistake.

663
00:48:10,200 --> 00:48:11,200
Yeah.

664
00:48:11,200 --> 00:48:12,200
Deep learning.

665
00:48:12,200 --> 00:48:13,200
Deep learning.

666
00:48:13,200 --> 00:48:14,840
Here we go.

667
00:48:14,840 --> 00:48:18,680
Deep learning networks are neural networks with many layers.

668
00:48:18,680 --> 00:48:24,240
A layered network can process extensive amounts of data and determine the weight of each link

669
00:48:24,240 --> 00:48:25,560
in the network.

670
00:48:25,560 --> 00:48:31,080
For example, in an image recognition system, some layers of the neural network might detect

671
00:48:31,080 --> 00:48:37,320
individual features of a face like eyes, nose, or mouth, while another layer would be able

672
00:48:37,320 --> 00:48:42,960
to tell whether those features appear in a way that indicates a face.

673
00:48:42,960 --> 00:48:48,480
Like neural networks, deep learning is modeled on the way the human brain works and powers

674
00:48:48,480 --> 00:48:54,600
many machine learning uses like autonomous vehicles, chatpots, medical diagnostics.

675
00:48:54,600 --> 00:49:02,960
The more layers you have, the more potential you have for doing complex things well.

676
00:49:02,960 --> 00:49:08,960
Yeah.

677
00:49:08,960 --> 00:49:11,120
That's the one that kind of freaks me out.

678
00:49:11,120 --> 00:49:15,840
I think it's the one that has the most potential for what you hear.

679
00:49:15,840 --> 00:49:21,640
Someone like Elon Musk who's running around saying that we should be afraid of AI and

680
00:49:21,640 --> 00:49:24,720
what it's capable of.

681
00:49:24,720 --> 00:49:30,920
I'm not seeing anything that backs up his fear, at least not yet.

682
00:49:30,920 --> 00:49:34,600
That always makes me nervous because I feel like guys like that see the things that I

683
00:49:34,600 --> 00:49:35,600
don't see.

684
00:49:35,600 --> 00:49:36,600
Correct.

685
00:49:36,600 --> 00:49:41,800
Meaning, not meaning that, well, they probably do understand things that I don't understand.

686
00:49:41,800 --> 00:49:49,520
I'm not on the same level, but that they have access to things that I don't have access

687
00:49:49,520 --> 00:49:50,520
to.

688
00:49:50,520 --> 00:49:51,520
Yes.

689
00:49:51,520 --> 00:50:00,760
It definitely, the way that some of our friends were talking about the AI that they were experiencing

690
00:50:00,760 --> 00:50:06,800
when they went to a conference and got to, they were in the, is it the sphere?

691
00:50:06,800 --> 00:50:09,320
Oh, yeah, in Vegas.

692
00:50:09,320 --> 00:50:10,320
In Vegas.

693
00:50:10,320 --> 00:50:11,320
Yeah.

694
00:50:11,320 --> 00:50:13,560
And you too, I think, opened it up with that concert.

695
00:50:13,560 --> 00:50:14,560
Yeah.

696
00:50:14,560 --> 00:50:22,520
And they had the AI that was talking to them and it was like making conversation with people.

697
00:50:22,520 --> 00:50:31,600
And it was, I think it was Cody that was talking about how he thinks that they actually programmed

698
00:50:31,600 --> 00:50:42,440
to act a little more robotic because it would have freaked everyone out if it did what it

699
00:50:42,440 --> 00:50:47,520
did and didn't seem robotic.

700
00:50:47,520 --> 00:50:48,760
Potentially.

701
00:50:48,760 --> 00:50:55,760
So there was, there was definitely part of that that I was like, I'm intrigued because

702
00:50:55,760 --> 00:50:59,480
I've seen AI robot.

703
00:50:59,480 --> 00:51:00,720
Some scary stuff there.

704
00:51:00,720 --> 00:51:01,720
Oh yeah.

705
00:51:01,720 --> 00:51:04,160
And then there's.

706
00:51:04,160 --> 00:51:05,640
Terminator scary for a different reason.

707
00:51:05,640 --> 00:51:06,640
And then you got Johnny.

708
00:51:06,640 --> 00:51:10,120
I robot was scary because it's starting to look real.

709
00:51:10,120 --> 00:51:11,120
Yep.

710
00:51:11,120 --> 00:51:13,600
Can't forget Johnny five though.

711
00:51:13,600 --> 00:51:16,120
Man, Johnny five is alive.

712
00:51:16,120 --> 00:51:19,440
Johnny five is alive folks.

713
00:51:19,440 --> 00:51:23,920
Love that show.

714
00:51:23,920 --> 00:51:26,080
That's how we all thought it was going to happen.

715
00:51:26,080 --> 00:51:30,360
We put some programming into something and then lightning would strike it and then it

716
00:51:30,360 --> 00:51:31,360
would become human.

717
00:51:31,360 --> 00:51:32,360
Yeah.

718
00:51:32,360 --> 00:51:33,360
The new version of Frankenstein's monster.

719
00:51:33,360 --> 00:51:34,360
There you go.

720
00:51:34,360 --> 00:51:35,360
Yep.

721
00:51:35,360 --> 00:51:36,360
Okay.

722
00:51:36,360 --> 00:51:42,400
And so the last bit I've got here is challenges of artificial neural network models.

723
00:51:42,400 --> 00:51:47,640
Although there is huge potential for leveraging artificial neural networks and machine learning.

724
00:51:47,640 --> 00:51:50,240
The approach comes with some challenges.

725
00:51:50,240 --> 00:51:53,880
Models are complex and it can be difficult to explain the reasoning behind a decision

726
00:51:53,880 --> 00:51:59,520
and in what in many cases is a black box operation.

727
00:51:59,520 --> 00:52:06,840
This makes the issue of explainability a significant challenge in consideration.

728
00:52:06,840 --> 00:52:11,640
That one is the one that kind of just strikes me is odd.

729
00:52:11,640 --> 00:52:12,920
You built the model.

730
00:52:12,920 --> 00:52:14,080
You're training the model.

731
00:52:14,080 --> 00:52:18,960
You know the weights and measures that you're, I'm sorry, weights and biases that you're

732
00:52:18,960 --> 00:52:21,440
putting into the model.

733
00:52:21,440 --> 00:52:27,240
So why would you run into so much that you couldn't explain?

734
00:52:27,240 --> 00:52:31,320
I think it's because we don't really understand the way that we do things.

735
00:52:31,320 --> 00:52:33,120
Perhaps.

736
00:52:33,120 --> 00:52:34,120
That's one of those things.

737
00:52:34,120 --> 00:52:44,040
It's always been, what I've seen as the potential for a Pandora's box here is that we consider

738
00:52:44,040 --> 00:52:49,240
ourselves experts on so much now.

739
00:52:49,240 --> 00:52:57,760
Like, we got this, we know this and there's just too many things that we don't know.

740
00:52:57,760 --> 00:53:08,200
And you start messing around with stuff and you know, that not to sound too paranoid or

741
00:53:08,200 --> 00:53:14,840
whatever, but it is not surprising to me when there are things that, I mean, I've been working

742
00:53:14,840 --> 00:53:23,760
in IT for too long to not, or to be surprised when I do everything right and it doesn't

743
00:53:23,760 --> 00:53:25,760
work right.

744
00:53:25,760 --> 00:53:26,760
Sure.

745
00:53:26,760 --> 00:53:32,760
I mean, maybe in this regard, I take too much of the traditional programming approach to

746
00:53:32,760 --> 00:53:40,680
this and going, if you're not getting out what you expected, you coded something wrong.

747
00:53:40,680 --> 00:53:46,160
And maybe there's something unique in the way these models function that it does introduce

748
00:53:46,160 --> 00:53:48,680
some uncertainty there.

749
00:53:48,680 --> 00:53:53,560
And maybe that's why sometimes they're surprised on what they get.

750
00:53:53,560 --> 00:53:54,960
Yeah.

751
00:53:54,960 --> 00:54:00,920
I think it has a lot to do with that and just the complexity that's inherent in it.

752
00:54:00,920 --> 00:54:04,960
You know, like I don't, I think we take a lot of things for granted and I think there's

753
00:54:04,960 --> 00:54:10,240
a lot of things that we don't understand and then we think it's binary.

754
00:54:10,240 --> 00:54:12,240
But it's not.

755
00:54:12,240 --> 00:54:14,240
Fair.

756
00:54:14,240 --> 00:54:17,200
Okay.

757
00:54:17,200 --> 00:54:22,680
With all types of machine learning models, the accuracy of the final model depends heavily

758
00:54:22,680 --> 00:54:26,520
on the quantity and quality of training data available.

759
00:54:26,520 --> 00:54:28,520
Data, data, data.

760
00:54:28,520 --> 00:54:33,080
Good data in, good data out, bad data in, bad data out.

761
00:54:33,080 --> 00:54:34,080
Cali.

762
00:54:34,080 --> 00:54:35,080
Yep.

763
00:54:35,080 --> 00:54:38,240
That's why I still have people emailing me and calling me.

764
00:54:38,240 --> 00:54:44,440
They want me to sign on to their service that I'm already using their service.

765
00:54:44,440 --> 00:54:45,440
Don't do that.

766
00:54:45,440 --> 00:54:47,200
Not a good look.

767
00:54:47,200 --> 00:54:53,000
A model built with an artificial neural network needs even more data and resources to train

768
00:54:53,000 --> 00:54:55,480
than a traditional machine learning model.

769
00:54:55,480 --> 00:55:01,440
This means millions of data points in contrast to hundreds of thousands needed by a traditional

770
00:55:01,440 --> 00:55:06,480
machine learning model.

771
00:55:06,480 --> 00:55:11,040
And then next, the most complex artificial neural networks are often referred to as

772
00:55:11,040 --> 00:55:16,640
deep neural networks, referencing the multi-layered network architecture.

773
00:55:16,640 --> 00:55:21,840
Deep learning models are usually trained using labeled training data, which is data with

774
00:55:21,840 --> 00:55:24,480
a defined input and output.

775
00:55:24,480 --> 00:55:29,640
This is known as supervised machine learning, unlike unsupervised machine learning, which

776
00:55:29,640 --> 00:55:32,600
uses unlabeled raw training data.

777
00:55:32,600 --> 00:55:36,760
A model will learn the features and patterns within the labeled training data and learn

778
00:55:36,760 --> 00:55:42,040
to perform an intended task through the examples in the training data.

779
00:55:42,040 --> 00:55:45,960
Artificial neural networks need a huge amount of training data, more so than more traditional

780
00:55:45,960 --> 00:55:48,240
machine learning algorithms.

781
00:55:48,240 --> 00:55:54,000
This is the realm of big data, so many millions of data points may be required.

782
00:55:54,000 --> 00:56:01,720
And this makes sense for why they keep trying to get as many data points from you as they

783
00:56:01,720 --> 00:56:02,720
can.

784
00:56:02,720 --> 00:56:03,720
Correct.

785
00:56:03,720 --> 00:56:07,280
It's important to all of them that you keep plugging in.

786
00:56:07,280 --> 00:56:11,040
It's important to all of them that you keep clicking on things.

787
00:56:11,040 --> 00:56:21,680
It's important that they are allowed to have access to the app failures or whatever information

788
00:56:21,680 --> 00:56:23,480
they can get off of any app.

789
00:56:23,480 --> 00:56:34,240
They want to know more about you so that they can program better or better, I guess, predict

790
00:56:34,240 --> 00:56:38,560
all that waiting and biasing.

791
00:56:38,560 --> 00:56:41,240
And that brings me back to Johnny Five.

792
00:56:41,240 --> 00:56:43,120
Need more input.

793
00:56:43,120 --> 00:56:44,920
Need more input.

794
00:56:44,920 --> 00:56:45,920
Yeah.

795
00:56:45,920 --> 00:56:47,160
Such a good movie.

796
00:56:47,160 --> 00:56:48,160
Yep.

797
00:56:48,160 --> 00:56:49,160
All right.

798
00:56:49,160 --> 00:56:50,160
So next, the need...

799
00:56:50,160 --> 00:56:51,160
Way before it's time.

800
00:56:51,160 --> 00:56:52,160
Yes.

801
00:56:52,160 --> 00:56:57,560
Such a large array of labeled quality data is a limiting factor to being able to develop

802
00:56:57,560 --> 00:57:01,240
artificial neural network models.

803
00:57:01,240 --> 00:57:06,320
Organizations are therefore limited to those that have access to the required big data.

804
00:57:06,320 --> 00:57:12,200
The most powerful artificial neural network models have complex, multi-layered architecture.

805
00:57:12,200 --> 00:57:17,480
These models require a huge amount of resources and power to process data sets.

806
00:57:17,480 --> 00:57:22,880
This requires powerful resource-intensive GPU units and system architecture.

807
00:57:22,880 --> 00:57:30,040
Again, the level of resources required is a limiting factor and challenge for organizations.

808
00:57:30,040 --> 00:57:35,000
And that's why so many of these data centers are talking about building their own power

809
00:57:35,000 --> 00:57:36,000
plants.

810
00:57:36,000 --> 00:57:37,000
Yeah.

811
00:57:37,000 --> 00:57:40,000
Because they need more power.

812
00:57:40,000 --> 00:57:43,240
More power, more reliable power, more...

813
00:57:43,240 --> 00:57:45,280
I mean, just every single time.

814
00:57:45,280 --> 00:57:50,640
I don't need to have to worry about whether or not Homeboy flicked on his lights.

815
00:57:50,640 --> 00:57:51,640
Right.

816
00:57:51,640 --> 00:57:52,640
I mean, he doesn't...

817
00:57:52,640 --> 00:57:55,160
All of this needs to work all of the time.

818
00:57:55,160 --> 00:57:56,160
Yep.

819
00:57:56,160 --> 00:58:01,560
And it makes sense why this is being dominated by these big companies.

820
00:58:01,560 --> 00:58:03,280
Oh, yeah.

821
00:58:03,280 --> 00:58:04,280
Deep pockets.

822
00:58:04,280 --> 00:58:05,280
Yep.

823
00:58:05,280 --> 00:58:06,280
All right.

824
00:58:06,280 --> 00:58:12,440
And then last, the method of transfer learning is often used to lower the resource intensity.

825
00:58:12,440 --> 00:58:17,360
In this process, extensive knowledge from other models, an existing artificial neural

826
00:58:17,360 --> 00:58:22,320
network can be transferred or adapted when developing a new model.

827
00:58:22,320 --> 00:58:27,040
The streamlines development as models aren't built from scratch each time, but it can be

828
00:58:27,040 --> 00:58:30,040
built from elements of existing models.

829
00:58:30,040 --> 00:58:32,000
Yeah, that makes sense.

830
00:58:32,000 --> 00:58:36,960
And that, I think, is how the Chinese pulled off what they did with DeepSeek, but they're

831
00:58:36,960 --> 00:58:39,320
just not one to talk about it.

832
00:58:39,320 --> 00:58:41,000
I mean, why would they?

833
00:58:41,000 --> 00:58:48,400
That doesn't help them come out on top, but they didn't start from scratch.

834
00:58:48,400 --> 00:58:53,200
They started from existing data.

835
00:58:53,200 --> 00:58:54,920
They started from existing models.

836
00:58:54,920 --> 00:58:58,120
They started from, oh, but it didn't cost us that much.

837
00:58:58,120 --> 00:59:04,000
I'm like, okay, but probably it would have if you started from zero.

838
00:59:04,000 --> 00:59:06,800
From the ground up, absolutely.

839
00:59:06,800 --> 00:59:12,240
And you know, like there's just, I don't know.

840
00:59:12,240 --> 00:59:20,600
To me, that's why you don't take it seriously when somebody says that it wasn't that hard

841
00:59:20,600 --> 00:59:21,600
to do that.

842
00:59:21,600 --> 00:59:23,920
Well, the idea was already out there.

843
00:59:23,920 --> 00:59:25,120
Yeah.

844
00:59:25,120 --> 00:59:30,080
I don't, it takes a lot of time to come up with the idea and how to actually execute

845
00:59:30,080 --> 00:59:35,800
it when you've got an idea of what it may be or what some of the problems may already

846
00:59:35,800 --> 00:59:41,840
be, then it's easier to tweak it.

847
00:59:41,840 --> 00:59:49,120
But regardless, like I think all of these are just going to keep getting better as long

848
00:59:49,120 --> 01:00:03,680
as we don't let things like that kind of setback mess with us.

849
01:00:03,680 --> 01:00:12,600
Yeah, what was so exciting to me about talking about this is kind of funny.

850
01:00:12,600 --> 01:00:18,040
I was actually thinking about this from a different perspective or a different, yeah,

851
01:00:18,040 --> 01:00:27,240
I guess perspective whenever you brought it up to me, because I realized how much I was

852
01:00:27,240 --> 01:00:45,640
being fed in social media AI solutions to different problems.

853
01:00:45,640 --> 01:00:54,400
I have a friend who was talking to me recently about how he thinks that what young people

854
01:00:54,400 --> 01:01:04,960
need to be putting effort into now is becoming AI gurus or even like AI sherpas, I think

855
01:01:04,960 --> 01:01:12,640
was the term that was used in the conversation that we were having.

856
01:01:12,640 --> 01:01:23,880
Knowing so much about different AI tools and what AI is, which tools are using which, I

857
01:01:23,880 --> 01:01:32,160
guess, functions of AI really well in order to be the guy that when people ask a question

858
01:01:32,160 --> 01:01:37,040
about, well, what does my business need to be using AI for?

859
01:01:37,040 --> 01:01:42,840
This person would be able to say like, oh, these are the best resources for that.

860
01:01:42,840 --> 01:01:50,720
This is maybe a way that you haven't been thinking about using AI that is out there and whatever.

861
01:01:50,720 --> 01:02:03,760
I just took a second, not a second, I took 30 minutes of my scrolling and I wasn't actually

862
01:02:03,760 --> 01:02:11,320
scrolling on Instagram in order to see things that would interest me on Instagram.

863
01:02:11,320 --> 01:02:23,000
I just scrolled through Instagram to see how many AI related advertisements came up.

864
01:02:23,000 --> 01:02:40,880
In 30 minutes, one, two, three, four, five, six, seven, eight, nine, 10, 11, 12, 13, 14,

865
01:02:40,880 --> 01:02:44,760
15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26 different advertisements for different

866
01:02:44,760 --> 01:02:45,920
AI products.

867
01:02:45,920 --> 01:02:51,000
Now, three of those are AI classes, which I thought was really funny considering the

868
01:02:51,000 --> 01:03:00,120
conversation that I had had with a friend of mine about people needing to be like gurus

869
01:03:00,120 --> 01:03:11,640
in AI and to be able to know what tools are out there and how best to use AI in any given,

870
01:03:11,640 --> 01:03:14,000
I guess, sector.

871
01:03:14,000 --> 01:03:20,960
And one of them is straight up from the University of Texas at Austin.

872
01:03:20,960 --> 01:03:28,320
I think the other two were just people's classes on AI, which seemed more homegrown than anything

873
01:03:28,320 --> 01:03:38,160
else, but I mean, there's a class at the University of Texas in Austin.

874
01:03:38,160 --> 01:03:41,320
And I don't know if that's a full degree.

875
01:03:41,320 --> 01:03:49,040
Nature seemed like it from the advertisement, but maybe it's just like an add on.

876
01:03:49,040 --> 01:03:50,640
It didn't seem like a certificate.

877
01:03:50,640 --> 01:03:51,720
It wasn't like that.

878
01:03:51,720 --> 01:03:57,000
But anyway, it's a lot.

879
01:03:57,000 --> 01:04:02,320
You know, I had been looking at VR technology lately for some other things that I was thinking

880
01:04:02,320 --> 01:04:13,640
about and the hand in hand of AI and VR right now is kind of a next five years growth model,

881
01:04:13,640 --> 01:04:15,200
it seems like.

882
01:04:15,200 --> 01:04:24,240
I think both of those will see just a different level inside of both of those in the next

883
01:04:24,240 --> 01:04:26,760
five years is what I'm predicting.

884
01:04:26,760 --> 01:04:35,440
I mean, it may take 10, but I would think five will see pretty serious growth.

885
01:04:35,440 --> 01:04:40,320
Anyway, there was a lot there.

886
01:04:40,320 --> 01:04:49,120
And I mean, it ranges from the huge majority of what I think about whenever I think about

887
01:04:49,120 --> 01:04:56,280
AI right now because of these advertisements are these like handhelds or wearable tech,

888
01:04:56,280 --> 01:05:03,440
whether it's glasses or little lanyards or just a voice recorder.

889
01:05:03,440 --> 01:05:14,440
I kid you not, Chris, there is a little device that's I did.

890
01:05:14,440 --> 01:05:17,480
Again, I didn't spend a whole lot of time looking into each one.

891
01:05:17,480 --> 01:05:23,920
I was just trying to get through how many there were in that period of time, but it's

892
01:05:23,920 --> 01:05:27,440
a stinking breathalyzer for your diet.

893
01:05:27,440 --> 01:05:28,440
Really?

894
01:05:28,440 --> 01:05:35,400
It's analyzing your breath and then telling you what you need to eat next.

895
01:05:35,400 --> 01:05:36,400
That I.

896
01:05:36,400 --> 01:05:40,320
Okay, so I'm just going to say it.

897
01:05:40,320 --> 01:05:45,320
I thought it made me think of a movie.

898
01:05:45,320 --> 01:05:46,320
Half-paked.

899
01:05:46,320 --> 01:05:47,320
Which one?

900
01:05:47,320 --> 01:05:48,320
Huh?

901
01:05:48,320 --> 01:05:49,320
Half-paked.

902
01:05:49,320 --> 01:05:52,240
I think it's, well, maybe I'm wrong.

903
01:05:52,240 --> 01:05:53,240
Maybe, maybe I'm wrong.

904
01:05:53,240 --> 01:05:56,000
Maybe it's actually, no, I'm wrong.

905
01:05:56,000 --> 01:05:58,720
It's, I think it's Biodome.

906
01:05:58,720 --> 01:06:01,760
Oh, yeah, no, it is Biodome.

907
01:06:01,760 --> 01:06:03,320
I know which one you're talking about.

908
01:06:03,320 --> 01:06:06,440
They burp and blow it in each other's faces and they take a breath and tell them what

909
01:06:06,440 --> 01:06:07,440
they ate.

910
01:06:07,440 --> 01:06:08,440
Yes.

911
01:06:08,440 --> 01:06:09,440
Yes.

912
01:06:09,440 --> 01:06:13,800
It's actually one of the Baldwin boys with Polyshore.

913
01:06:13,800 --> 01:06:14,800
Yep, that's it.

914
01:06:14,800 --> 01:06:16,200
I can't remember which one.

915
01:06:16,200 --> 01:06:20,120
Yeah, so, yeah, sorry, wrong movie reference, but we ultimately got there.

916
01:06:20,120 --> 01:06:21,120
The Biodome.

917
01:06:21,120 --> 01:06:22,120
Biodome.

918
01:06:22,120 --> 01:06:23,120
Yes.

919
01:06:23,120 --> 01:06:24,120
Yep.

920
01:06:24,120 --> 01:06:26,960
So, that's what that made me think of.

921
01:06:26,960 --> 01:06:30,320
You're saying that it's a breathalyzer for what you eat.

922
01:06:30,320 --> 01:06:33,200
Yeah, I mean, it's not telling you what you ate.

923
01:06:33,200 --> 01:06:42,600
It's telling you what you need to eat, what you're deficient in based on what you, you

924
01:06:42,600 --> 01:06:45,280
know, what your breath smells like.

925
01:06:45,280 --> 01:06:47,440
I, it was, I mean, it was hilarious.

926
01:06:47,440 --> 01:06:49,800
It's not smell, you know, anyway.

927
01:06:49,800 --> 01:06:50,800
Sure.

928
01:06:50,800 --> 01:06:55,000
That one, that one got me because I was, I was, you know, all of these seem to be, or

929
01:06:55,000 --> 01:07:04,040
not seem to be, most of the like apps and different wearables and glasses, it's pretty

930
01:07:04,040 --> 01:07:13,520
obvious the productivity, sorry, productivity side of this that it's aimed at.

931
01:07:13,520 --> 01:07:17,000
This is, this is trying to take things off of your plate.

932
01:07:17,000 --> 01:07:26,200
It's trying to make certain decisions easier, you know, I think there's more ADD, ADHD in

933
01:07:26,200 --> 01:07:27,720
our world today.

934
01:07:27,720 --> 01:07:28,720
Yes.

935
01:07:28,720 --> 01:07:39,880
And, and that like executive function paralysis and the, the idea that you, you can offload

936
01:07:39,880 --> 01:07:44,960
some of that is really attractive.

937
01:07:44,960 --> 01:07:53,600
I mean, the number of these apps, especially that are related to calendar management or,

938
01:07:53,600 --> 01:08:00,080
or just transcription that can be turned into either notes or tasks.

939
01:08:00,080 --> 01:08:07,400
I was like, wow, it is, it is for sure making long strides in this.

940
01:08:07,400 --> 01:08:10,840
And I think, you know, the, the nervousness that people have is they see things like that

941
01:08:10,840 --> 01:08:16,080
and they're like, oh, it's stealing my information.

942
01:08:16,080 --> 01:08:23,360
I mean, it is, I don't, I don't know how to tell you like that, that, that iterative

943
01:08:23,360 --> 01:08:30,400
aspect of AI where it has to take what you did, like what you, what it received from

944
01:08:30,400 --> 01:08:37,840
you, what you asked of it, and then what your response to what it gave you, like it

945
01:08:37,840 --> 01:08:49,320
has to keep putting that up against everybody else's to see what's, you know, normal where

946
01:08:49,320 --> 01:08:54,240
that, where that weight needs to be, where the bias needs to be.

947
01:08:54,240 --> 01:08:57,440
So it's really, it's really funny to me.

948
01:08:57,440 --> 01:09:07,800
I think the real, it's not, you know, it's not the danger here in my mind is not machine

949
01:09:07,800 --> 01:09:17,680
is taking over in like a, you know, the, the, you know, Skynet fear that we have, it's more

950
01:09:17,680 --> 01:09:19,160
herd mentality.

951
01:09:19,160 --> 01:09:20,660
Yes.

952
01:09:20,660 --> 01:09:26,760
It's more, it's, I wouldn't even call it, yeah, I wouldn't even call it echo chamber.

953
01:09:26,760 --> 01:09:45,360
It's just the, at some point, I think we choose what the machine gives us in its first attempt.

954
01:09:45,360 --> 01:09:53,000
And when that happens, when that's the normal or even second attempt, whatever the like

955
01:09:53,000 --> 01:09:59,440
predicted option is there, like if, if we always wait until the second attempt or if

956
01:09:59,440 --> 01:10:10,040
we always take the first attempt or whatever, like that becomes the, I mean, it's going

957
01:10:10,040 --> 01:10:12,720
to learn that that's the thing.

958
01:10:12,720 --> 01:10:18,080
It's waiting for the second option or it's, it's going to give us what it thinks the first

959
01:10:18,080 --> 01:10:25,440
time and then they always take it and then it doesn't learn anymore and then we're stuck.

960
01:10:25,440 --> 01:10:26,760
Yep.

961
01:10:26,760 --> 01:10:33,720
If we rely on the computer to think for us, then the computer will think for us.

962
01:10:33,720 --> 01:10:41,760
And that's, that, that to me is more makes me more nervous than the other side of it.

963
01:10:41,760 --> 01:10:48,280
Oh, well, you said you wanted peace, so that means you need to be destroyed.

964
01:10:48,280 --> 01:10:49,280
Yep.

965
01:10:49,280 --> 01:10:50,280
Okay.

966
01:10:50,280 --> 01:10:51,280
Well, yeah.

967
01:10:51,280 --> 01:10:54,320
Yeah, I think for me, we may have given it too much credit there.

968
01:10:54,320 --> 01:10:55,320
It's two things.

969
01:10:55,320 --> 01:11:00,680
It's one, it's, you're going to give up more and more control to the computer than you

970
01:11:00,680 --> 01:11:01,680
should give it.

971
01:11:01,680 --> 01:11:02,680
Right.

972
01:11:02,680 --> 01:11:08,480
And that leads you down a path of a thousand paper cuts until you realize, oh, I've given

973
01:11:08,480 --> 01:11:12,000
up way more and it's too late.

974
01:11:12,000 --> 01:11:13,000
Yep.

975
01:11:13,000 --> 01:11:21,760
Type of a thing or the idea that this starts to become so common that you get a couple of

976
01:11:21,760 --> 01:11:26,000
generations that just grow up and this is just what you do.

977
01:11:26,000 --> 01:11:31,160
This is life, kind of like what's happened with smartphones.

978
01:11:31,160 --> 01:11:34,880
And you just have people who just like, oh yeah, this is just kind of the way it is.

979
01:11:34,880 --> 01:11:42,680
And they just inherently trust the AI and don't question it, don't know how to reason

980
01:11:42,680 --> 01:11:49,320
and essentially fact check and think for themselves.

981
01:11:49,320 --> 01:11:58,000
And you just kind of end up in a, in an odd world that leads to an idiocracy.

982
01:11:58,000 --> 01:11:59,000
Yeah.

983
01:11:59,000 --> 01:12:03,960
I think that those are valid concerns.

984
01:12:03,960 --> 01:12:09,720
We try to make life easier and then you realize by making life easier, you took away the things

985
01:12:09,720 --> 01:12:12,920
that made you better.

986
01:12:12,920 --> 01:12:15,720
That almost sounds like something Jordan Peterson would say.

987
01:12:15,720 --> 01:12:22,120
I'm just saying I'm not Jordan Peterson.

988
01:12:22,120 --> 01:12:24,840
I'm not smart enough to be.

989
01:12:24,840 --> 01:12:27,400
Anyway, super interesting.

990
01:12:27,400 --> 01:12:32,120
We may have another show on this just because of the amount that there is to it.

991
01:12:32,120 --> 01:12:35,560
But thanks for bringing this up.

992
01:12:35,560 --> 01:12:42,320
I'm grateful to get to talk about it and drink this beer while I'm doing it.

993
01:12:42,320 --> 01:12:43,320
Absolutely.

994
01:12:43,320 --> 01:12:46,800
Never hurts to have a good beer with a friend.

995
01:12:46,800 --> 01:12:47,800
Nope.

996
01:12:47,800 --> 01:12:50,040
I'd agree with that.

997
01:12:50,040 --> 01:12:54,040
Well, and not just because you said it.

998
01:12:54,040 --> 01:12:55,040
Yeah.

999
01:12:55,040 --> 01:12:57,200
No booster grants tonight folks.

1000
01:12:57,200 --> 01:13:00,200
Jason Corden didn't keep up his end of the show.

1001
01:13:00,200 --> 01:13:05,480
Jason, I mean, I guess we've got to have him on again.

1002
01:13:05,480 --> 01:13:06,480
Yeah.

1003
01:13:06,480 --> 01:13:12,480
I actually had somebody ask me if we were going to have more people come on the show

1004
01:13:12,480 --> 01:13:13,980
as a guest.

1005
01:13:13,980 --> 01:13:15,480
We just need to recruit.

1006
01:13:15,480 --> 01:13:16,480
Yeah.

1007
01:13:16,480 --> 01:13:20,840
If you want to be on the show, let us know.

1008
01:13:20,840 --> 01:13:21,840
Yeah.

1009
01:13:21,840 --> 01:13:27,120
Anyway, I'm all for it, especially if there's, I mean, this would be a good, this would be

1010
01:13:27,120 --> 01:13:29,560
a good show to have Cody on.

1011
01:13:29,560 --> 01:13:30,560
Yeah.

1012
01:13:30,560 --> 01:13:33,480
He's probably got a lot of thoughts in this area.

1013
01:13:33,480 --> 01:13:34,480
Yeah.

1014
01:13:34,480 --> 01:13:39,240
Him and probably Robbie Jones.

1015
01:13:39,240 --> 01:13:41,400
Oh, Robbie Jones.

1016
01:13:41,400 --> 01:13:43,280
Oh, Brad Wofford.

1017
01:13:43,280 --> 01:13:44,280
Yeah.

1018
01:13:44,280 --> 01:13:54,040
Brad Wofford would love to talk about at least the, what are the two that he uses?

1019
01:13:54,040 --> 01:13:56,240
Oh, for the art.

1020
01:13:56,240 --> 01:13:57,240
Yeah.

1021
01:13:57,240 --> 01:14:06,640
He uses one for art, he uses another for either copy or like even songwriting.

1022
01:14:06,640 --> 01:14:14,760
Jordan shared the song that we used in our D Now video that just got, like our church

1023
01:14:14,760 --> 01:14:18,680
just did Disciple Now this past weekend.

1024
01:14:18,680 --> 01:14:26,240
And the song that they used or that the guy who made the video used for the video was

1025
01:14:26,240 --> 01:14:28,800
completely AI generated.

1026
01:14:28,800 --> 01:14:37,080
I was like, you're listening to the song, you're like, oh, it's like a contemporary

1027
01:14:37,080 --> 01:14:40,240
Christian music.

1028
01:14:40,240 --> 01:14:44,240
And then all of a sudden it said like our church name and I was like, what?

1029
01:14:44,240 --> 01:14:48,880
Oh, no, that, like that happened.

1030
01:14:48,880 --> 01:14:58,120
This is not a song that was independently recorded by an artist.

1031
01:14:58,120 --> 01:14:59,120
Yeah.

1032
01:14:59,120 --> 01:15:08,720
So I wonder if Brad is either using Leonardo, there's Da Vinci and there's Dolly for the

1033
01:15:08,720 --> 01:15:10,440
art at least.

1034
01:15:10,440 --> 01:15:12,920
I didn't think that's the one, I could be wrong.

1035
01:15:12,920 --> 01:15:14,680
I had Chad knows all of this.

1036
01:15:14,680 --> 01:15:16,720
I'll just ask him.

1037
01:15:16,720 --> 01:15:22,400
But yeah, we should definitely, let's put our heads together in the next week, try to

1038
01:15:22,400 --> 01:15:30,360
figure out if there's somebody that can come talk to us about their particular use of AI.

1039
01:15:30,360 --> 01:15:31,360
Yep.

1040
01:15:31,360 --> 01:15:33,480
Especially if they like beer.

1041
01:15:33,480 --> 01:15:34,960
Oh yeah.

1042
01:15:34,960 --> 01:15:35,960
Absolutely.

1043
01:15:35,960 --> 01:15:39,480
That sounds like a good job we could put together.

1044
01:15:39,480 --> 01:15:40,480
I like it.

1045
01:15:40,480 --> 01:15:41,480
Thanks, Chris.

1046
01:15:41,480 --> 01:15:43,480
Enjoyed it, man.

1047
01:15:43,480 --> 01:15:44,480
Cheers.

1048
01:15:44,480 --> 01:15:45,480
Cheers.

1049
01:15:45,480 --> 01:15:47,120
One of these remotely.

1050
01:15:47,120 --> 01:15:48,120
Empty.

1051
01:15:48,120 --> 01:15:49,120
Right?

1052
01:15:49,120 --> 01:15:52,560
I feel so much better about this now that we've done it.

1053
01:15:52,560 --> 01:15:53,560
Yeah.

1054
01:15:53,560 --> 01:15:54,560
And it worked.

1055
01:15:54,560 --> 01:15:55,560
Yep.

1056
01:15:55,560 --> 01:15:56,560
I'm afraid I'm empty too.

1057
01:15:56,560 --> 01:15:57,560
Thanks, Francisco.

1058
01:15:57,560 --> 01:16:01,080
I appreciate you making this possible.

1059
01:16:01,080 --> 01:16:04,200
Do love the Pod Mobile.

1060
01:16:04,200 --> 01:16:06,200
Pod Mobile, get you one.

1061
01:16:06,200 --> 01:16:07,200
Okay.

1062
01:16:07,200 --> 01:16:09,200
See you all later.

1063
01:16:09,200 --> 01:16:10,200
Yep.

1064
01:16:10,200 --> 01:16:17,200
Excellent.

1065
01:16:17,200 --> 01:16:32,280
I could listen to another hour.

1066
01:16:32,280 --> 01:16:42,280
There is no way anything can be this good.