文字稿 🔗
WEBVTT
1
00:00:19.720 --> 00:00:20.320
Hey Ben.
2
00:00:20.320 --> 00:00:21.080
Hey Matt.
3
00:00:21.080 --> 00:00:22.620
We have a plan.
4
00:00:22.620 --> 00:00:23.150
We do.
5
00:00:23.150 --> 00:00:24.310
The plan is not...
6
00:00:24.310 --> 00:00:28.040
It's all of eight seconds old, but we do have a plan.
7
00:00:28.040 --> 00:00:30.780
Yeah, no, we mentioned it last night over drinks.
8
00:00:30.780 --> 00:00:31.980
Well, that's true. That's a good point.
9
00:00:31.980 --> 00:00:42.560
But I haven't really looked into this. So this is going to be by the... shoot from the hip as always. But the plan, I think, was for us to talk about... a letter that was written a little while ago.
10
00:00:42.560 --> 00:00:50.020
That's right. That's right. So I'm checking my facts with... Wikipedia here, just to make sure that I get this right.
11
00:00:50.020 --> 00:00:54.120
As am I, honestly. I've got it lined up there as well.
12
00:00:54.120 --> 00:00:58.040
And so, you know, dear viewer, if you want to, read along with us here...
13
00:00:58.040 --> 00:01:04.980
But, we, well, let's talk about what... What are we going to talk about? I'm actually going to bring it up so that I can even read it out, if we wanted to do some kind of audiobook version.
14
00:01:04.980 --> 00:01:12.440
Yeah. Yes. No, that's good. We probably... we should... there's maybe some select sections of this. So what we're going to talk about today is Bill Gates'
15
00:01:12.440 --> 00:01:17.870
open letter to hobbyists that he wrote in 1976, I believe.
16
00:01:17.870 --> 00:01:18.090
Right.
17
00:01:18.090 --> 00:01:18.540
February 3rd.
18
00:01:18.540 --> 00:01:22.540
February 3rd. Yeah, February 3rd, 1976, before even I
19
00:01:22.540 --> 00:01:22.540
1976. Yes.
20
00:01:22.540 --> 00:01:25.060
was born, but not by much.
21
00:01:25.060 --> 00:01:26.840
Right. Yeah, well, me too.
22
00:01:26.840 --> 00:01:28.040
Well, yeah, you are, but you're young.
23
00:01:28.040 --> 00:01:33.020
Yeah. Right.
24
00:01:33.020 --> 00:01:34.620
Spry 48.
25
00:01:34.620 --> 00:01:36.260
That's right.
26
00:01:36.260 --> 00:01:38.280
Yeah, that's... I mean, you've got, yeah, exactly.
27
00:01:38.280 --> 00:01:56.700
Which is, you know, a nice round number. So... the open letter to hobbyists, I think, is really interesting. It reflects an era that I sort of lived through the very tail end of, which I would just maybe refer to generally as the pre...
28
00:01:56.700 --> 00:01:56.700
Mhm.
29
00:01:56.700 --> 00:01:58.900
internet era.
30
00:01:58.900 --> 00:01:59.170
Yeah.
31
00:01:59.170 --> 00:01:59.440
Right.
32
00:01:59.440 --> 00:02:04.480
Obviously, the internet was around this time, but it wasn't in the form that, you know, we all see it and know it now, is what you mean.
33
00:02:04.480 --> 00:02:07.960
Sure. The commercial internet. The consumer internet.
34
00:02:07.960 --> 00:02:08.280
Right. Yes.
35
00:02:08.280 --> 00:02:10.200
As opposed to the academic internet.
36
00:02:10.200 --> 00:02:12.020
Right, this is AOL...
37
00:02:12.020 --> 00:02:13.600
Right. Right.
38
00:02:13.600 --> 00:02:16.720
AOL, a kind of generation onwards, and the sort of democratized... yeah, yeah, yeah, CompuServe.
39
00:02:16.720 --> 00:02:28.840
Yeah, CompuServe. Yeah, right. Exactly, exactly. Because certainly when I was a kid, access to the Internet was... a thing that was possible but kind of difficult to come by, maybe expensive.
40
00:02:28.840 --> 00:02:29.620
Not...
41
00:02:29.620 --> 00:02:30.400
Right.
42
00:02:30.400 --> 00:02:59.620
So just as I was in the middle of high school was when some of my... more well-off friends who had all the gear and whatnot would get a modem, and, you know, 14.4k. And I, for example, I had an acoustic coupler, if you remember these things, they were like actual things you put the handset into, that I'd found, and I tried to get working. I never actually got it working, but that was, like, the level of, like, attempts to get on to any form of BBS or internet at that time. So, yeah, yeah.
43
00:02:59.620 --> 00:02:59.760
You know...
44
00:02:59.760 --> 00:03:00.120
Yeah, uh-huh. Right.
45
00:03:00.120 --> 00:03:01.210
We're going to come back to this. BBSes in particular is going to be...
46
00:03:01.210 --> 00:03:01.300
But anyway.
47
00:03:01.300 --> 00:03:03.180
Yeah, yeah, yeah. Wait, we're getting distracted already because we're all just...
48
00:03:03.180 --> 00:03:04.480
Part of this whole story arc.
49
00:03:04.480 --> 00:03:04.640
Old men.
50
00:03:04.640 --> 00:03:06.040
I can tell it's going to happen already.
51
00:03:06.040 --> 00:03:11.000
Old men talking about their how they had it on the internet when they were kids.
52
00:03:11.000 --> 00:03:18.560
Right. But I just wanted to start off with the qualifier of, obviously, Matt and I were not heavily involved in the software industry in 1976.
53
00:03:18.560 --> 00:03:19.180
Yeah.
54
00:03:19.180 --> 00:03:21.570
And so, like, our perceptions on this are...
55
00:03:21.570 --> 00:03:21.700
But a gleam.
56
00:03:21.700 --> 00:03:25.730
are backward looking even from our younger days.
57
00:03:25.730 --> 00:03:26.380
Aha.
58
00:03:26.380 --> 00:03:31.350
But with those caveats in place, I think I can make some sort of interesting arguments.
59
00:03:31.350 --> 00:03:31.940
Even then.
60
00:03:31.940 --> 00:03:35.140
about this letter and how we might want to reflect on it today.
61
00:03:35.140 --> 00:03:49.390
Just in case our listener is unaware, as indeed one of the podcasters is relatively unaware of the contents of this letter, other than it's written in front of me right now, I haven't really read it. What, give me your TL;DR, and, you know, as you said, we can read out little parts of it.
62
00:03:49.390 --> 00:03:49.760
Mm-hmm. Right.
63
00:03:49.760 --> 00:04:02.860
Yeah, so this was a letter that Bill wrote when he sort of realized that everyone was stealing his software. He had written a program for computers available at the
Load full transcript
time.
64
00:04:02.860 --> 00:04:11.840
And he sort of, you know, was like, hey, our sales and the usage of this thing do not match up. So people are clearly just making copies of it.
65
00:04:11.840 --> 00:04:13.660
Right.
66
00:04:13.660 --> 00:04:24.040
And he sort of makes the argument in this letter that that's not good or sustainable. You know, maybe just to read the first paragraph of the letter.
67
00:04:24.040 --> 00:04:32.740
He says, to me, the most critical thing in the hobby market right now is the lack of good software courses, books, and the software itself.
68
00:04:32.740 --> 00:04:45.760
Without good software and an owner who understands programming, a hobby computer is wasted. Will quality software be written for the hobby market? He asks.
69
00:04:45.760 --> 00:04:49.180
And the sort of implication is that unless you pay for it, the answer is no.
70
00:04:49.180 --> 00:04:49.570
It's...
71
00:04:49.570 --> 00:04:49.950
Right.
72
00:04:49.950 --> 00:04:50.720
Right.
73
00:04:50.720 --> 00:04:50.960
But so, that's the thrust of the letter.
74
00:04:50.960 --> 00:04:52.820
Right, not unreasonable.
75
00:04:52.820 --> 00:04:54.260
Right. And you can go read it for yourself.
76
00:04:54.260 --> 00:05:04.840
And, you know, he says things along the lines of, you know, we've written these BASICs. You know, the value of the time that we have used exceeds $40,000, he's saying. So it's costing them.
77
00:05:04.840 --> 00:05:07.680
$40,000. Yes.
78
00:05:07.680 --> 00:05:09.080
No, that's a lot. It is 1976. It is a lot of money. Yeah. Yeah.
79
00:05:09.080 --> 00:05:09.920
I mean, in 1976, 40K is a lot of money.
80
00:05:09.920 --> 00:05:20.120
And then everyone says they like their BASIC that they've written. It's been positive. But, you know, most of these, quote, users never bought it, is what they say.
81
00:05:20.120 --> 00:05:24.440
And they, you know, the air quotes, or not even air quotes, the actual literal quotes around users.
82
00:05:24.440 --> 00:05:24.440
Yeah. Yeah. Yeah.
83
00:05:24.440 --> 00:05:31.060
implies that, you know, they think a user is someone who's paid for their software, not somebody who is just using it, although, you know...
84
00:05:31.060 --> 00:05:32.500
Right.
85
00:05:32.500 --> 00:05:44.800
Yeah, the amount of royalties we have received from sales... makes the time spent on Altair BASIC worth less than two dollars an hour for the work that they've done, they're saying, which is, you know, not obviously different, even in 1976, that's not a living, but not a good wage.
86
00:05:44.800 --> 00:05:45.720
Mm hmm.
87
00:05:45.720 --> 00:05:46.940
Not a lot, especially for something like building software, which is, you know... Yeah.
88
00:05:46.940 --> 00:05:56.510
Is it fair? And so on, you know. So... This is, yeah, I would appreciate people who wants to pay up, he says, at the end, which is rather hopeful.
89
00:05:56.510 --> 00:05:56.540
Okay.
90
00:05:56.540 --> 00:06:02.320
or has a suggestion or comment and gives his actual... like, address that you could write it to.
91
00:06:02.320 --> 00:06:03.320
Right. Send me a check.
92
00:06:03.320 --> 00:06:04.540
Yeah, send me a check.
93
00:06:04.540 --> 00:06:05.000
Yeah, right.
94
00:06:05.000 --> 00:06:05.460
Right.
95
00:06:05.460 --> 00:06:08.720
And Bill Gates, general partner, Micro hyphen Soft.
96
00:06:08.720 --> 00:06:09.440
Right. Capital S.
97
00:06:09.440 --> 00:06:11.740
Which is... Capital S. Yeah, really, kind of, it shows what it is.
98
00:06:11.740 --> 00:06:11.800
Right.
99
00:06:11.800 --> 00:06:13.410
So he's basically saying...
100
00:06:13.410 --> 00:06:13.600
Yeah.
101
00:06:13.600 --> 00:06:20.360
Hey, it's great, but we can't sustain this. It's not okay to not pay for software.
102
00:06:20.360 --> 00:06:23.860
Right. Who is going to make your software if you're not paying people to do it?
103
00:06:23.860 --> 00:06:26.580
Right. And it makes perfect sense. Right.
104
00:06:26.580 --> 00:06:26.700
Mm hmm.
105
00:06:26.700 --> 00:06:36.880
It's not like... I mean, he even makes a point about hardware in this. Right. It's like, it's not like I don't expect to have to pay for a physical computer that I'm then going to put on my desk as a hobbyist and then play with.
106
00:06:36.880 --> 00:06:41.580
Right, I know that. That's why I don't just go into a shop and... You know, you wouldn't just steal a computer.
107
00:06:41.580 --> 00:06:46.000
Right. Yes. You wouldn't just download a computer, right?
108
00:06:46.000 --> 00:06:48.000
Right. Right. And so, you know, yeah, it all makes sense.
109
00:06:48.000 --> 00:06:48.640
Uh-huh. Yeah.
110
00:06:48.640 --> 00:06:49.410
So, kind of...
111
00:06:49.410 --> 00:06:49.620
Yeah.
112
00:06:49.620 --> 00:06:51.200
Yeah, what's your point, Ben?
113
00:06:51.200 --> 00:06:52.040
Well, so...
114
00:06:52.040 --> 00:07:08.300
Right. So I think one way to look at... let's say 1976, from when this was written, to, let's say, I don't know, 1996-ish, the mid-nineties. Let's just say the mid-nineties.
115
00:07:08.300 --> 00:07:16.630
And the sort of... easy access of the internet to everyone, right? Like in 1990, let's say. Internet access was possible, doable, not a super common thing.
116
00:07:16.630 --> 00:07:18.510
Right, but not... certainly not common.
117
00:07:18.510 --> 00:07:30.100
You know, it's like... yeah. It was more of a specialty thing. To 2000, 10 years later, it's like, yeah, literally everybody is connected to the internet in some way, even if it's just like dial up.
118
00:07:30.100 --> 00:07:30.410
Mm-hmm.
119
00:07:30.410 --> 00:07:42.320
Literally, everyone is drowning in... computer, sorry, in America Online CDs that they've been sent through the post multiple times and use them as coasters and whatever, and yeah, we've got all the internet we could ever want.
120
00:07:42.320 --> 00:07:43.260
Right. Exactly. Exactly.
121
00:07:43.260 --> 00:07:43.890
Right, right.
122
00:07:43.890 --> 00:07:44.200
Yeah.
123
00:07:44.200 --> 00:07:47.320
So, you know, wherever in the midpoint of... Yes.
124
00:07:47.320 --> 00:07:48.120
Something happened in that decade.
125
00:07:48.120 --> 00:07:52.400
Something happened in that range that was very important, and we're just going to call that the commercial internet.
126
00:07:52.400 --> 00:07:53.500
Yeah. Yeah.
127
00:07:53.500 --> 00:07:54.050
Right.
128
00:07:54.050 --> 00:07:54.600
Mm-hmm.
129
00:07:54.600 --> 00:08:01.030
I think that you could make an argument that basically what happened over that span of time
130
00:08:01.030 --> 00:08:01.300
Mhm.
131
00:08:01.300 --> 00:08:11.520
is you had hobbyists, you had people who really loved working with computers because they were fun. Right, they, like, they really enjoyed it. Maybe it's also their job.
132
00:08:11.520 --> 00:08:12.540
I hear... programming is fun.
133
00:08:12.540 --> 00:08:18.100
Right. Well, and this is part of my inspiration for this is that I'm reflecting on earlier episodes.
134
00:08:18.100 --> 00:08:19.670
I've been told. On earlier episodes, right, right.
135
00:08:19.670 --> 00:08:23.400
They were doing it again. It's a letter to hobbyists.
136
00:08:23.400 --> 00:08:23.700
Yeah.
137
00:08:23.700 --> 00:08:34.340
Right. He's saying, like, hey, guys, like, I realize you're having a great time with all this stuff, and it's fun, and I like it, and you like it, it's great. But someone's got to pay to write this software. Isn't it just come for free?
138
00:08:34.340 --> 00:08:34.820
Yep.
139
00:08:34.820 --> 00:08:49.440
And part of the reason for that, I think, part of Bill's argument here, is that, especially at that time, when you delivered software to somebody, and he makes this point of, like, manuals, courses, and the software itself, well, the software itself is also, like...
140
00:08:49.440 --> 00:08:49.440
Yeah.
141
00:08:49.440 --> 00:08:50.820
Physical media!
142
00:08:50.820 --> 00:08:50.820
Yeah, of course.
143
00:08:50.820 --> 00:08:57.560
That needs to be mailed to someone, or bought in a shop, or, like, you need to, like...
144
00:08:57.560 --> 00:08:58.260
A what?
145
00:08:58.260 --> 00:09:02.210
A store of some kind, like, you know, a mall, perhaps.
146
00:09:02.210 --> 00:09:02.600
Doesn't sound right.
147
00:09:02.600 --> 00:09:03.160
I don't know.
148
00:09:03.160 --> 00:09:03.380
Yeah.
149
00:09:03.380 --> 00:09:04.060
Maybe that's the thing.
150
00:09:04.060 --> 00:09:05.780
That's where like, yeah.
151
00:09:05.780 --> 00:09:06.860
Yeah. Right.
152
00:09:06.860 --> 00:09:07.540
I don't get it
153
00:09:07.540 --> 00:09:18.420
So... he's saying, like, the people need to expend their time making this. The manuals need to be printed. You know, the training courses need to be created.
154
00:09:18.420 --> 00:09:24.920
The software needs to be written to disks and tapes of various kinds. And it needs to be delivered to you.
155
00:09:24.920 --> 00:09:25.920
which themselves are, you know, have a cost with it.
156
00:09:25.920 --> 00:09:26.280
Right!
157
00:09:26.280 --> 00:09:36.300
Although, again, you know, obviously if these hobbyists are getting hold of it, they probably are getting hold of... hooky media to then, like, copy to, so there's still some physical cost to them, but just not the... not the part. So...
158
00:09:36.300 --> 00:09:36.780
Right.
159
00:09:36.780 --> 00:09:37.110
Right.
160
00:09:37.110 --> 00:09:37.400
Yeah.
161
00:09:37.400 --> 00:09:46.420
Right, right. And so he's like, there's this cost to this. And I can put myself in the shoes of somebody who's reading this letter, who is one of these hobbyists.
162
00:09:46.420 --> 00:09:46.980
Right.
163
00:09:46.980 --> 00:09:53.960
Who is kind of like... grumble, grumble, grumble. Like, I don't... I don't like this. I mean, you're kind of right about some of these things, but I don't really know.
164
00:09:53.960 --> 00:09:57.780
Listen, you can't see this, but Ben, as he's saying this, is stroking his beard.
165
00:09:57.780 --> 00:10:00.040
Right, exactly.
166
00:10:00.040 --> 00:10:03.620
Yeah, yeah, yeah, in a very... which, it's possibly the kind of person who was, in fact...
167
00:10:03.620 --> 00:10:08.000
Right. Very bearded. Those folks back then.
168
00:10:08.000 --> 00:10:08.560
Very bitter, yeah. Yeah.
169
00:10:08.560 --> 00:10:18.600
And I think I can make the argument that, like, the thing that... the hobbyists were able to refute this with, the refutation of Bill's letter
170
00:10:18.600 --> 00:10:18.600
Mhm.
171
00:10:18.600 --> 00:10:30.360
is open source and the internet. Because once you have... commonly available internet, then the distribution problems essentially evaporate.
172
00:10:30.360 --> 00:10:43.100
Or they become so diminis... diminimously small. That... as to, like... basically say, okay, Bill, like, I get that somebody needs to pay to ship this stuff, but what if we just didn't do that? What if we wrote the software for ourselves?
173
00:10:43.100 --> 00:10:45.710
And then just shared it.
174
00:10:45.710 --> 00:10:46.090
Yep.
175
00:10:46.090 --> 00:10:46.480
Right?
176
00:10:46.480 --> 00:11:03.500
And then the... The physical distribution is covered. The web, or the... or Gopher, or whatever it was back then, was our FTP sites, you know, had... you know, README.txts that would be the documentation, which, you know, I think, you know, quality documentation, as he
177
00:11:03.500 --> 00:11:03.500
Right. Yes. Usenet. Yeah.
178
00:11:03.500 --> 00:11:07.400
puts it here, or quality software, that's a... that's where, that, you know, some of those things... But...
179
00:11:07.400 --> 00:11:07.400
Right. Right.
180
00:11:07.400 --> 00:11:17.970
There are people who said, no, there is another solution to this problem. This is like the kind of, you know, there's a negative solution and a positive solution to the square root kind of thing going on here.
181
00:11:17.970 --> 00:11:18.080
Mm hmm. Mm-hmm, mm-hmm.
182
00:11:18.080 --> 00:11:20.220
It's
like, hey, what if we didn't do that?
183
00:11:20.220 --> 00:11:33.480
Right. Right, and... and I am maybe kind of making the argument here that it's like, open source is only possible because of that commercial access to the internet, whatever it is that happened between the 90s and 2000s.
184
00:11:33.480 --> 00:11:35.400
Like, obviously it was kind of there before, but if you...
185
00:11:35.400 --> 00:11:40.670
I was going to say, we can't... you know, BBSs and things had things like this. It wasn't open.
186
00:11:40.670 --> 00:11:40.960
Right.
187
00:11:40.960 --> 00:11:44.800
I mean, I don't remember it being open source. There were a lot of demos going on. There were lots of, like...
188
00:11:44.800 --> 00:11:44.800
Right.
189
00:11:44.800 --> 00:11:55.670
cracked copies of things for commercial games. There was... there were obviously, I mean, obviously the GNU project is older than, like, the 90s, right, and things of that nature, and the Free Software Federation Foundation.
190
00:11:55.670 --> 00:11:57.420
Yeah, yes. Right. Absolutely. Mm hmm.
191
00:11:57.420 --> 00:12:01.650
So clearly there have been folks who have been thinking this way in this sort of, dare I say it, socialist...
192
00:12:01.650 --> 00:12:01.740
Right.
193
00:12:01.740 --> 00:12:04.380
software.
194
00:12:04.380 --> 00:12:06.300
Right.
195
00:12:06.300 --> 00:12:07.200
Little bit of politics there.
196
00:12:07.200 --> 00:12:18.010
But, yes. But... Fast-forwarding to the state that we are in today, where I think you can very easily make the argument that, like, a good chunk of the technology that we have rests on these open source projects, on this...
197
00:12:18.010 --> 00:12:18.250
Mm hmm.
198
00:12:18.250 --> 00:12:18.600
A hundred percent.
199
00:12:18.600 --> 00:12:18.720
Like.
200
00:12:18.720 --> 00:12:18.840
Yeah.
201
00:12:18.840 --> 00:12:30.560
Software that was basically granted to the world for free. And all of the technology that we have right now, certainly all of the, you know...
202
00:12:30.560 --> 00:12:33.370
LLMs that we have that's writing all this code was trained on
203
00:12:33.370 --> 00:12:34.360
Yeah.
204
00:12:34.360 --> 00:12:36.320
MIT licensed software and probably other, not...
205
00:12:36.320 --> 00:12:36.800
Yeah, yeah.
206
00:12:36.800 --> 00:12:39.510
super legally licensed software, if we're being real about it.
207
00:12:39.510 --> 00:12:39.760
Yeah.
208
00:12:39.760 --> 00:12:48.740
that was just given away by these hobbyists. And the progenitors, or the... not the progenitors, but the successors of the sort of hobbyist mentality.
209
00:12:48.740 --> 00:12:48.740
Yeah.
210
00:12:48.740 --> 00:12:58.200
that Bill was speaking to in this letter, and the reaction to it, as, like, a whole group, was, like, screw you, we'll build our own software.
211
00:12:58.200 --> 00:12:58.540
Yeah.
212
00:12:58.540 --> 00:13:06.940
If it's not economically viable for you to do it, then we don't need you. We'll do it ourselves. And the
thing that allowed them to do that
213
00:13:06.940 --> 00:13:07.160
Right.
214
00:13:07.160 --> 00:13:16.100
to the degree that we have today, is the internet and the ability to distribute the software to literally everyone in the world for basically free.
215
00:13:16.100 --> 00:13:22.560
And... I'm just sort of thinking back through, yeah, I mean, yes, absolutely. I think it's a foundational aspect, but I also think that, like,
216
00:13:22.560 --> 00:14:06.260
stuff like compilers being accessible. Like, so when I was growing up and I was writing stuff in assembly, the reason I didn't learn C, amongst just my... just general grumbliness, is that there wasn't a free version of a C compiler for my platform. And GCC ultimately was ported to my platform eventually. But obviously somebody somewhere said, this is... and it's that is a key unlocking, enabling technology, right at the very bottom of the stack, right, that, and, you know, maybe Linux is the other example of something that is very foundational. And then, you know, like, every academic
217
00:14:06.260 --> 00:14:14.770
wants to add and contribute. You know, you want to get your master's degree in something, well, how about you add a feature to GCC? How about you add something, you know, that sort of compounds from an academic sensibility? So I'm wondering if that's part of it as well, but I do think that, like, the distribution means that, like, it's taken as read that it's everywhere, right? GCC is everywhere.
218
00:14:14.770 --> 00:14:15.520
Mm hmm. Mm hmm. Yeah.
219
00:14:15.520 --> 00:14:28.610
Yeah, I mean, again, maybe, you know, like, looking at the moment, everything is hosted through Git, or, you know, something Git derivative. Obviously there was SVN, yeah, before that was CVS, before that there's all manner of things, but, like, so... Oh, yeah, SourceForge, man.
220
00:14:28.610 --> 00:14:28.760
Yeah.
221
00:14:28.760 --> 00:14:29.060
SourceForge. Yeah.
222
00:14:29.060 --> 00:14:31.910
Oh, gosh, how the mighty have fallen, right? I mean, in a way.
223
00:14:31.910 --> 00:14:32.120
Yeah.
224
00:14:32.120 --> 00:14:56.910
But like that also is an enabler, because, you know, like, having open software, being out, being able to collaborate. And of course, the... the collaboration tools themselves are open source, so there's a kind of bootstrappy feeling to this as well, whereas, you know, I remember using, like, Perforce, which you had to pay a per-seat license for at companies and things like that, so, you know, and Subversion came along and said, like, well, we can be Perforce but we're not Perforce, we are Subversion, we're effectively the same in many, many ways.
225
00:14:56.910 --> 00:14:58.400
Right. Right.
226
00:14:58.400 --> 00:15:05.380
So there's a whole bunch of technologies, but, like, this... Yeah, I don't know where we're going now. Where are we going with this, Ben? Where have you led me?
227
00:15:05.380 --> 00:15:05.860
Well.
228
00:15:05.860 --> 00:15:10.000
I'm looking, it's going like, wow, I hadn't thought of it that way. And now...
229
00:15:10.000 --> 00:15:11.260
But that's... so, so what of it, right?
230
00:15:11.260 --> 00:15:13.080
Why was Bill so wrong? You know, yeah.
231
00:15:13.080 --> 00:15:23.200
Right. Well, so here's the thing is that... as we were discussing in a previous episode about programming and funness, right? So these hobbyists were doing it for fun.
232
00:15:23.200 --> 00:15:23.390
Yep.
233
00:15:23.390 --> 00:15:23.980
Yeah.
234
00:15:23.980 --> 00:15:25.820
They were building computers for fun.
235
00:15:25.820 --> 00:15:28.890
They were building eventually software for fun with the open source movement.
236
00:15:28.890 --> 00:15:29.020
Mm hmm.
237
00:15:29.020 --> 00:15:54.240
They were able to collaborate with other people that were doing it for fun. And then that turned into all of this foundational technology that we have. And it was essentially free to them. It's like once you pay the cost of the computer, you can write your own software. And once you can write your own software and connect that computer to the world, and now you can share the software, and you can collaborate on it. And now you basically have everything that you need to do this as a hobby for fun.
238
00:15:54.240 --> 00:15:59.720
Are we, perhaps, maybe now, coming full circle back with Bill, where...
239
00:15:59.720 --> 00:16:03.700
It's like, well, it's not as fun as it used to be.
240
00:16:03.700 --> 00:16:04.380
Oh, I see.
241
00:16:04.380 --> 00:16:11.200
And you have to pay for it, because you have to pay for tokens or you have to pay for access to LLMs.
242
00:16:11.200 --> 00:16:22.120
Well, that is interesting. So let's take that apart. I mean... There is a certain democratization that anyone who can pay for the tokens can potentially make software now.
243
00:16:22.120 --> 00:16:22.890
We've talked about this before.
244
00:16:22.890 --> 00:16:23.160
True.
245
00:16:23.160 --> 00:16:25.520
But it's... But it's not. It's not free as in beer.
246
00:16:25.520 --> 00:16:37.800
It's not free. As in beer, no. It is... You don't have to do that. No one's making people do that. You know, there's definitely, I mean, within my wider friend group, there are a lot of people we still... wait.
247
00:16:37.800 --> 00:16:38.020
True. That's true.
248
00:16:38.020 --> 00:16:52.220
I'm wearing a demo T-shirt right now. You know, we still make demos, and there is definitely a schism opening amongst the sort of more artistic end of programming for fun, for free, where the idea of using AI is verboten.
249
00:16:52.220 --> 00:16:53.040
Mm hmm.
250
00:16:53.040 --> 00:17:01.580
for good reason, for some folks, and, like, and is a huge enabler for other folks who are, like, being able to do things that have never been able to be done before, by them at least.
251
00:17:01.580 --> 00:17:02.400
Right. Yeah, I am.
252
00:17:02.400 --> 00:17:11.420
And, you know, there's a democrat... democratizer. I mean, and... but yeah, is there a little bit of the... the Incredibles when, you know, when everybody's super,
253
00:17:11.420 --> 00:17:22.830
nobody's super. Is that perhaps what people fear, and that in that particular world? You know, like the folks who are incredibly talented are like, well, hey, can I say, my talent is now not as valuable as it was before because everyone else can do it too, right?
254
00:17:22.830 --> 00:17:23.470
Yeah.
255
00:17:23.470 --> 00:17:25.410
You know?
256
00:17:25.410 --> 00:17:30.060
But... where was I going? I can't remember where I was going with it. Why are we talking about demos? Because... AI...
257
00:17:30.060 --> 00:17:35.280
Well, it's not... it's that there are going to be people who don't use LLMs to write code.
258
00:17:35.280 --> 00:17:40.610
Right. You don't have to. Like, you can have people that say, like, look, this is artisanal handcrafted code.
259
00:17:40.610 --> 00:17:41.080
You don't have to. It's not a hard requirement that you do. Right.
260
00:17:41.080 --> 00:17:44.070
Come and marvel at how good I am at programming and how beautiful the code looks.
261
00:17:44.070 --> 00:17:44.180
Right.
262
00:17:44.180 --> 00:17:53.630
And that's great. You know, again, the repair shop, you know, there is still a world in which Steve from The Repair Shop fixes clocks and doesn't just go, well, here's a Casio.
263
00:17:53.630 --> 00:17:53.930
Right.
264
00:17:53.930 --> 00:17:56.980
Yeah, yeah, yeah.
265
00:17:56.980 --> 00:17:59.400
You know, get off you go, here's a new battery, we're done here, right?
266
00:17:59.400 --> 00:17:59.980
Right. Yeah, exactly. Exactly. Yeah.
267
00:17:59.980 --> 00:18:11.800
That will always be the case. But the financial aspect of it has changed. We've kind of litigated, or... it's shifted around, and perhaps that's what you're getting at, you know, for a starting point.
268
00:18:11.800 --> 00:18:23.740
Well, I'm kind of making the case that, like, all of these foundational open source technologies that were created by these hobbyists and given away for free were used by all of humanity to create all of this technology that we have.
269
00:18:23.740 --> 00:18:27.300
And maybe what's about to happen is the faucet for that is going to turn off.
270
00:18:27.300 --> 00:18:27.300
Is it there?
271
00:18:27.300 --> 00:18:33.620
Because the motivations to continue... expanding at the same rate that we have been.
272
00:18:33.620 --> 00:18:34.800
Well...
273
00:18:34.800 --> 00:18:35.950
I'm not gonna say that those projects are gonna disappear.
274
00:18:35.950 --> 00:18:36.240
No.
275
00:18:36.240 --> 00:18:37.500
Well, clearly not.
276
00:18:37.500 --> 00:18:39.140
Well, they're certainly adapting, aren't they?
277
00:18:39.140 --> 00:18:42.080
I mean, like, so some obviously are just carrying on exactly as before.
278
00:18:42.080 --> 00:18:42.920
Yeah. Yeah. Right.
279
00:18:42.920 --> 00:18:50.880
Some... are asking the question, what do we do about well-meaning folks
who have got a token budget to spend and want to help?
280
00:18:50.880 --> 00:18:50.880
Mm-hmm.
281
00:18:50.880 --> 00:18:56.330
Which, you know, is another thing, right? You can get a reward from, like, not being as skilled necessarily.
282
00:18:56.330 --> 00:18:56.680
Right? Yeah. Right.
283
00:18:56.680 --> 00:18:57.750
And not... it's not even skill, right?
284
00:18:57.750 --> 00:18:57.900
Yeah.
285
00:18:57.900 --> 00:18:58.660
It's just...
286
00:18:58.660 --> 00:19:01.570
choosing to use a different tool to make a project.
287
00:19:01.570 --> 00:19:01.730
Yeah.
288
00:19:01.730 --> 00:19:02.060
Mm hmm.
289
00:19:02.060 --> 00:19:11.220
to advance a project's goals, in which case, you know, the token spend is real, just like effectively the electricity spend that I, you know, you're saying it's free to like sit on my computer.
290
00:19:11.220 --> 00:19:11.400
Yeah.
291
00:19:11.400 --> 00:19:15.470
It isn't. It's my time away from things I could be doing.
292
00:19:15.470 --> 00:19:15.860
Right. Yeah.
293
00:19:15.860 --> 00:19:19.590
It is my electricity. It's my whatever. I know, but it's de minimis, really.
294
00:19:19.590 --> 00:19:19.960
Mm-hmm
295
00:19:19.960 --> 00:19:43.790
But... so those folks may get an inrush of good changes. I mean, that does not seem to be the case from what I read, but I mean, maybe there's a world in which that's okay, in which case, you know, now we've got the snowballing of this kind of... of this idea that the people are inclined to improve and make open software even more available and cool and feature-rich.
296
00:19:43.790 --> 00:19:44.360
Yeah.
297
00:19:44.360 --> 00:19:54.480
I mean... Did you see this? I mean, this is going to date when we're recording this, but a few days ago there was a... the release of essentially a Rust
298
00:19:54.480 --> 00:19:59.380
clean room AI re-implementation of... every
299
00:19:59.380 --> 00:20:01.080
I was made aware of that, yes.
300
00:20:01.080 --> 00:20:05.090
Adobe project. And I'm like, I don't know how I feel about this myself, right?
301
00:20:05.090 --> 00:20:05.240
Mm hmm.
302
00:20:05.240 --> 00:20:18.540
I mean, I certainly... have complicated feelings about Adobe over the years for all the things that they've done, but it's... you cannot say that they have not innovated in a space where...
303
00:20:18.540 --> 00:20:29.100
perhaps they needed to be a market leader. They're certainly a market leader for a couple of the big projects. Also, I was speaking to someone from Adobe, and they were telling me about a lot of other stuff that they do that perhaps you don't know that they do.
304
00:20:29.100 --> 00:20:37.990
But, like, you know, you think of Photoshop, obviously. You think of, like, PDF, Acrobat. You think of the video editing equivalent that I can't remember the name of right now.
305
00:20:37.990 --> 00:20:38.360
Okay.
306
00:20:38.360 --> 00:20:42.190
And those kinds of things, After Effects, whatever. That's cool. All those are great.
307
00:20:42.190 --> 00:20:42.600
Mm-hmm.
308
00:20:42.600 --> 00:20:53.770
They've gone to... you know, they used to be expensive to buy. They, I mean, it used to be a significant cost at, like, the game studios I used to work at. And now they've gone to a subscription model.
309
00:20:53.770 --> 00:20:54.340
Yeah, yeah
310
00:20:54.340 --> 00:21:03.190
And the first thing when I saw this and I was speaking with a friend at work, the first thing the friend said was, wow, this is really going to put... pressure on the pricing model of Adobe.
311
00:21:03.190 --> 00:21:03.660
Mm hmm.
312
00:21:03.660 --> 00:21:13.840
Because if you can... just recreate it with some amount of... tokens that have admittedly probably incredibly heavily subsidized right now, and unsustainably so.
313
00:21:13.840 --> 00:21:14.780
Right.
314
00:21:14.780 --> 00:21:18.000
So we don't know what the true cost of recreating Photoshop in Rust in
315
00:21:18.000 --> 00:21:23.360
Yeah, but certainly less than all of the future discounted revenues of Adobe. Right.
316
00:21:23.360 --> 00:21:23.760
Correct. Correct.
317
00:21:23.760 --> 00:21:25.920
by a few orders of magnitude.
318
00:21:25.920 --> 00:21:27.180
By, yeah, a number.
319
00:21:27.180 --> 00:21:27.440
Yeah.
320
00:21:27.440 --> 00:21:33.160
But, like, how is this changing the world? And is this... I'm not going to say is it better or worse. Is it... how is it changing? Yeah.
321
00:21:33.160 --> 00:21:34.380
Yeah.
322
00:21:34.380 --> 00:21:35.040
Yeah.
323
00:21:35.040 --> 00:21:47.840
Yeah. I mean, that's another interesting dimension to it of, like, you know, getting back to Bill and his letter and the sort of premise here, it's like open source eating Adobe's business.
324
00:21:47.840 --> 00:21:52.900
But it's a very particular kind of open source that has been accelerated by LLMs.
325
00:21:52.900 --> 00:21:55.600
Right. I mean, like, the GIMP has been around forever.
326
00:21:55.600 --> 00:21:58.140
It is all... And, you know, Audacity has been around forever.
327
00:21:58.140 --> 00:21:58.380
Right.
328
00:21:58.380 --> 00:21:58.620
Yep.
329
00:21:58.620 --> 00:22:02.490
And various... Kdenlive and other things have been around for a while.
330
00:22:02.490 --> 00:22:02.920
Yep.
331
00:22:02.920 --> 00:22:12.240
There's just something about them that screams, I don't know, I'm going to say this is going to make everyone hate me, you know, it screams, this is open source, right?
332
00:22:12.240 --> 00:22:13.190
Which means that, like, it's...
333
00:22:13.190 --> 00:22:13.310
Right.
334
00:22:13.310 --> 00:22:13.440
Right.
335
00:22:13.440 --> 00:22:17.260
There's a particular feel of the software. It's like they're without a strong...
336
00:22:17.260 --> 00:22:18.260
Yeah, yeah, yeah
337
00:22:18.260 --> 00:22:28.540
product lead. And, I mean, that's not to say, like, I know that, like... Oh, I can't even... forget the bloke's name, but he's just left there. But, like, I know there was a product lead for Audacity, who worked really hard
338
00:22:28.540 --> 00:22:28.540
Mm hmm.
339
00:22:28.540 --> 00:22:32.600
to try and, like, steer it in the right direction and all that kind of stuff. But, you know, effectively when you're dealing with...
340
00:22:32.600 --> 00:22:40.580
Yeah. The resources a person like that is going to have is going to be de minimis in comparison to Adobe and the stupid amount of money that they have to throw at, like, all kinds of problems around product development.
341
00:22:40.580 --> 00:22:43.650
Right, right. Well, I mean, also, it's not the cool stuff, right?
342
00:22:43.650 --> 00:22:43.980
Yeah.
343
00:22:43.980 --> 00:22:58.670
If you're a programmer doing it, you know, I don't know. Maybe there's just something about the programming mind that means that open source sounds cool and groovy and you just want to do it. And you're like, hey, I'm sharing this stuff. It's like, nobody goes, my hobby is product managing or managing small teams of developers, right?
344
00:22:58.670 --> 00:22:58.780
Right.
345
00:22:58.780 --> 00:23:04.570
That's my hobby. I just love doing it. On my weekends, I just love to tell people what to do. I mean, maybe there are people like this, but certainly I don't.
346
00:23:04.570 --> 00:23:04.700
Right.
347
00:23:04.700 --> 00:23:08.910
I don't know any, and I probably wouldn't want to hang out with somebody who thinks it's fun to tell me what to do.
348
00:23:08.910 --> 00:23:08.960
Yeah.
349
00:23:08.960 --> 00:23:18.270
Well, and that is actually kind of the point that I was making earlier, is that, like, where is the future where it's like, my hobby is being a meat proxy for, you know...
350
00:23:18.270 --> 00:23:18.880
Okay, okay.
351
00:23:18.880 --> 00:23:19.500
Right.
352
00:23:19.500 --> 00:23:32.510
LLMs to write the next generation of cool open source software. Now, this is a great counterexample to that, and I almost wonder if this is almost, you know, as much of a... let's just see if we can do this to figure out if it's possible.
353
00:23:32.510 --> 00:23:32.860
Mm-hmm
354
00:23:32.860 --> 00:23:40.680
Not necessarily, I want to spend my time on an open source clone of Photoshop. Right? Maybe you do. Maybe that's what you want.
355
00:23:40.680 --> 00:23:43.000
But... kind of the rabbit's... the rabbit's out of the bag. No, the... the... but
356
00:23:43.000 --> 00:23:44.940
The hat's out of the rabbit.
357
00:23:44.940 --> 00:23:46.440
Something like that, yeah.
358
00:23:46.440 --> 00:23:48.990
The squirrel is, I don't know. I've...
359
00:23:48.990 --> 00:23:52.050
What is the thing? Oh God, I'm having a moment. The cat's out of the bag.
360
00:23:52.050 --> 00:23:52.240
The cat's out of the bag, yeah.
361
00:23:52.240 --> 00:23:56.410
Oh my gosh. You know, like you can't put the genie back in a bottle here. Like once someone goes, oh, that's cool.
362
00:23:56.410 --> 00:23:56.680
Yeah, yeah, yeah.
363
00:23:56.680 --> 00:24:01.270
Even if Adobe shuts it down,
someone else will bring it up. Or, you know, even if it's just there, it can be forked a million times.
364
00:24:01.270 --> 00:24:01.650
Right. Yeah.
365
00:24:01.650 --> 00:24:02.420
Yeah.
366
00:24:02.420 --> 00:24:06.880
Right. Like once, once it's proven that this is possible, it's going to be incredibly difficult to stop.
367
00:24:06.880 --> 00:24:06.940
So.
368
00:24:06.940 --> 00:24:12.520
And there might even be a little bit of Streisand effect here where it's like the harder Adobe tries to stop it, the more people are going to.
369
00:24:12.520 --> 00:24:12.520
the more...
370
00:24:12.520 --> 00:24:13.590
fight against it. Right.
371
00:24:13.590 --> 00:24:13.690
Yeah.
372
00:24:13.690 --> 00:24:14.260
Just to be like, you know.
373
00:24:14.260 --> 00:24:15.600
Very true.
374
00:24:15.600 --> 00:24:18.470
You don't get to have the choice here, Adobe.
375
00:24:18.470 --> 00:24:18.940
Yeah.
376
00:24:18.940 --> 00:24:20.340
But, but I
377
00:24:20.340 --> 00:24:22.490
I mean, did you see their stock price took a, like a...
378
00:24:22.490 --> 00:24:22.680
Donk.
379
00:24:22.680 --> 00:24:23.840
I did.
380
00:24:23.840 --> 00:24:24.510
I saw the graph.
381
00:24:24.510 --> 00:24:24.600
Yeah.
382
00:24:24.600 --> 00:24:25.900
That was, yeah.
383
00:24:25.900 --> 00:24:28.850
And I mean... maybe rightly so. Maybe rightly so.
384
00:24:28.850 --> 00:24:29.270
Yeah.
385
00:24:29.270 --> 00:24:29.680
It's...
386
00:24:29.680 --> 00:24:31.140
It's interesting.
387
00:24:31.140 --> 00:24:33.380
So, I don't know. I just... this sort of struck me.
388
00:24:33.380 --> 00:24:33.720
Yeah.
389
00:24:33.720 --> 00:24:35.500
I was... you know, randomly something.
390
00:24:35.500 --> 00:24:39.860
So your thesis really is, like, have we gone full circle? Are we now paying for software but in a different way?
391
00:24:39.860 --> 00:24:49.260
Yeah. Yes, we're paying in the form of tokens, and that is going to create... situations where the hobbyists now, in order to...
392
00:24:49.260 --> 00:25:03.840
do the same thing that they have done in the last 30 years, which is create these open source projects that are so impactful, that are so foundational, to repeat that again over the next 30 years.
393
00:25:03.840 --> 00:25:14.590
They're... they're kind of not going to be able to. Because in order to do it, they're going to have to pay for tokens that they didn't have to really pay for before. And your, you know, your point about, like, okay, there's electricity and there's other things.
394
00:25:14.590 --> 00:25:14.800
Right.
395
00:25:14.800 --> 00:25:18.600
But in comparison, it's just... it's going to be many orders of magnitude more expensive.
396
00:25:18.600 --> 00:25:20.940
It's not going to be at a hobby level at the moment.
397
00:25:20.940 --> 00:25:21.320
At the moment. At the moment, right?
398
00:25:21.320 --> 00:25:21.540
At the moment.
399
00:25:21.540 --> 00:25:27.290
Because, you know, like in the 1960s, prior to even, you know, this letter being written, there were people who would...
400
00:25:27.290 --> 00:25:27.340
Yes.
401
00:25:27.340 --> 00:25:43.180
tinkering with this kind of stuff, but they were beg... borrowing and stealing access to Crays, or, you know, getting... you know, going to universities and finding their way into the places to be able to, like, work on stuff that they thought was interesting, and, you know, punching cards and stuff, and then getting, you know...
402
00:25:43.180 --> 00:25:43.340
Right. Right.
403
00:25:43.340 --> 00:25:43.940
Yeah. Yeah, yeah, yeah.
404
00:25:43.940 --> 00:25:46.150
So, you know, maybe it'll change again.
405
00:25:46.150 --> 00:25:46.280
Maybe there's hope.
406
00:25:46.280 --> 00:25:55.650
And then, you know, like, in this, I mean, it's honestly... isn't it staggering that in the space of our lifetimes we've gone, literally, from this letter exhorting people with their expensive hobbyist machines.
407
00:25:55.650 --> 00:26:13.130
where, you know, maybe it was... I can't even think where I was going with this, but, like, you know, like two, two... like, it's just democratized. Everyone's got, like, several computers on their body at any point in time, be it a watch and a phone and a laptop and a tablet and a, you know.
408
00:26:13.130 --> 00:26:18.340
Right. Yes. Yes.
409
00:26:18.340 --> 00:26:25.670
You know, and it's... it's... We don't even think about that now. The price of having a computer to work on is, like, small. And even if you can't afford a computer to work on, you can get something in the cloud relatively cheap or free, or you go to a website.
410
00:26:25.670 --> 00:26:26.040
Yep.
411
00:26:26.040 --> 00:26:27.750
As long as you've got access to a website, you can do some work.
412
00:26:27.750 --> 00:26:27.840
Yeah.
413
00:26:27.840 --> 00:26:32.580
So that level of democratization has changed in our... in this period of time.
414
00:26:32.580 --> 00:26:33.720
Yeah. Yeah.
415
00:26:33.720 --> 00:26:54.080
I wonder... you know, obviously, we don't know how... where this is going at all. But... open models, or models that have access, some amount of, you know, free weight, so I was going to say open is... I don't know that the open thing makes sense for models. It's like, you know... it's open, you know, it's more free as in beer rather than free as in software, because, like, I'm going to hand you...
416
00:26:54.080 --> 00:26:54.170
Yeah.
417
00:26:54.170 --> 00:26:54.460
Yeah. Yeah. Yeah.
418
00:26:54.460 --> 00:26:54.560
Yeah.
419
00:26:54.560 --> 00:27:03.110
14 terabytes of floating point numbers and say, good luck, there's your brain. You want to make a change to it?
420
00:27:03.110 --> 00:27:03.300
Right. Right. Right. Yes. If you want to make a change. Yeah, exactly.
421
00:27:03.300 --> 00:27:06.300
Good luck. You know, I'll see you in it.
422
00:27:06.300 --> 00:27:06.960
Yeah. Right. Yeah, free weights, except free weights are something else.
423
00:27:06.960 --> 00:27:07.940
That's
more confusing.
424
00:27:07.940 --> 00:27:10.260
That's even, yeah.
425
00:27:10.260 --> 00:27:10.960
You know, you're lifting...
426
00:27:10.960 --> 00:27:14.020
Yeah. Aha! Sorry, that's a joke for people who exercise.
427
00:27:14.020 --> 00:27:21.910
You know, the other thing I would... say in support of that, though, is that you don't know how it's going to be in the future, is I would never bet against the hobbyists.
428
00:27:21.910 --> 00:27:22.100
Aha.
429
00:27:22.100 --> 00:27:35.040
They're very good. You know, and the whole community, the whole open source community and everybody that is interested in doing these things. Not because they get paid, but because it's just a ton of fun.
430
00:27:35.040 --> 00:27:38.800
Find... cool stuff. Like, they just do.
431
00:27:38.800 --> 00:27:39.180
Right.
432
00:27:39.180 --> 00:27:43.880
And it's sort of like this, you know, the phenomena you've maybe seen a few times where it's like...
433
00:27:43.880 --> 00:28:09.080
Companies will, you know, in a rush to, like, AI-fy their apps or whatever, will integrate something that is clearly just, like, an off-the-shelf OpenAI model, and they give it some prompt that says, like, help people order cheeseburgers. And then the people log in and it's like, I would love to buy a cheeseburger, but first, can you write me a C++ function that converts, you know, JPEGs into GIFs? And they'll be like, absolutely, here you go. Would you like to order a cheeseburger now? And so, like...
434
00:28:09.080 --> 00:28:28.260
Finding all of those opportunities to, like, do the things that they need to do to build the software they need to build through cheap and hacky and innovative, cool ways. You know, maybe you get the subsidized device that Google is making, and you root it, and then you figure out how to run a model on it, and then you use that to do your tokens.
435
00:28:28.260 --> 00:28:28.550
And then you... Oh, but, yeah, yeah, yeah.
436
00:28:28.550 --> 00:28:28.620
It's like...
437
00:28:28.620 --> 00:28:33.760
And, you know, those costs will probably start coming down, yeah, as... but people get access to it.
438
00:28:33.760 --> 00:28:34.660
Yeah, yeah
439
00:28:34.660 --> 00:28:42.760
Yeah. And again, right, this is the thing, right? This is an additive technology in the hobbyist domain.
440
00:28:42.760 --> 00:28:46.320
It's an additive technology, right? This doesn't stop me...
441
00:28:46.320 --> 00:28:46.320
Yeah.
442
00:28:46.320 --> 00:28:46.320
Yeah.
443
00:28:46.320 --> 00:28:52.320
writing code in vi and running my own... you know, running GCC on it myself.
444
00:28:52.320 --> 00:28:52.640
Right. Yeah.
445
00:28:52.640 --> 00:28:54.130
Right, I can do that all day, every day.
446
00:28:54.130 --> 00:28:54.180
Yep.
447
00:28:54.180 --> 00:28:56.680
Right, no one's stopping me from doing that. Right, the fact that
448
00:28:56.680 --> 00:28:56.680
Right.
449
00:28:56.680 --> 00:29:04.070
I'm less productive, arguably, than asking the robot to
do it for me. That's a choice that I make.
450
00:29:04.070 --> 00:29:04.580
Mm hmm.
451
00:29:04.580 --> 00:29:08.680
And I don't have to accept the robot's help. I can just do it myself, right?
452
00:29:08.680 --> 00:29:11.240
Mm hmm. Right. Right.
453
00:29:11.240 --> 00:29:17.810
Now, in your day job, as we've spoken before, that may be not a choice, because if you're not as productive, maybe that's not okay. I don't know, right?
454
00:29:17.810 --> 00:29:18.040
Yeah, yeah
455
00:29:18.040 --> 00:29:23.390
Maybe it depends on the level of finesse and what your company wants. Obviously, everyone's going to be different on that point.
456
00:29:23.390 --> 00:29:23.600
Yeah.
457
00:29:23.600 --> 00:29:29.320
If I'm choosing to do in my hobby time what I like, no one's going to stop me from...
458
00:29:29.320 --> 00:29:29.320
Right.
459
00:29:29.320 --> 00:29:38.000
carving and crafting, which... I think we said this. Yeah. So, dear listener, we went to the pub last night and we had a drink or two.
460
00:29:38.000 --> 00:29:38.620
Yeah we did.
461
00:29:38.620 --> 00:29:50.160
So I'm slightly feeling a little worse for wear. But we did, in fact, I did say that a friend of mine had posted something that has said something along the lines of, is programming now a craft?
462
00:29:50.160 --> 00:29:52.730
And it was a very ambiguous statement.
463
00:29:52.730 --> 00:29:52.880
Right.
464
00:29:52.880 --> 00:29:53.040
Right.
465
00:29:53.040 --> 00:30:05.660
And then the folks who interpreted it in different ways were interesting in the replies, to behold. And I think that is an interesting way of thinking about it, right? It's a craft, absolutely. You know, it always has been a craft, but, like, the craft comes with this sort of implicit feeling of, like, hobbyist, hobbying and crafting.
466
00:30:05.660 --> 00:30:05.660
It's like nobody, again, we talked about this, like knitting, right?
467
00:30:05.660 --> 00:30:05.660
Yeah, yeah, yeah.
468
00:30:05.660 --> 00:30:15.230
There are people, there are important people in your life who enjoy knitting and knit. Right. But, like...
469
00:30:15.230 --> 00:30:15.480
Yeah.
470
00:30:15.480 --> 00:30:18.760
It's... No one's going to stop them from knitting.
471
00:30:18.760 --> 00:30:19.400
Right.
472
00:30:19.400 --> 00:30:23.620
But I'm going to go to the store and I'm going to buy one myself, if it's okay with you.
473
00:30:23.620 --> 00:30:23.820
Right.
474
00:30:23.820 --> 00:30:24.020
Heh.
475
00:30:24.020 --> 00:30:42.520
Right. Well, and I think one of the concerns that you could have about this, though, because, your point about, like, I can always just bust out vi and do this myself. Like, that's the beauty of open source, is that at this point, it's never going away. It almost can't go away. It will always be there. And that's fantastic. And it's like a wonderful thing that humanity has created, honestly.
476
00:30:42.520 --> 00:30:52.740
but there is a concern that one could have that the craft and the profession are going to diverge in ways that they have not. In fact, they've converged
477
00:30:52.740 --> 00:31:01.310
over the last 30 years, and it's like, what you would do as a software engineer in a real job and what you would do as a hobbyist working on open source software are essentially identical
478
00:31:01.310 --> 00:31:02.200
Oh man.
479
00:31:02.200 --> 00:31:04.270
but maybe they won't be 10 years from now.
480
00:31:04.270 --> 00:31:04.560
Yeah.
481
00:31:04.560 --> 00:31:15.740
And so, like, that ability to apply the craft, because, you know, the practices and you have the tools, and it's like, literally, I do this all day long, and then I come home and I do... I do it again. It's like, no, what you're doing in your day job is now...
482
00:31:15.740 --> 00:31:19.710
very different from what you would do in sort of that more craft mentality, because that's what you can afford to do as your hobby.
483
00:31:19.710 --> 00:31:19.760
Yeah.
484
00:31:19.760 --> 00:31:28.000
I mean, that's another way of... that's another sort of... that's another relitigation of the, where are we going to find the software engineers of the future,
485
00:31:28.000 --> 00:31:29.540
Right.
486
00:31:29.540 --> 00:31:57.970
argument-ish. And, you know, and then it leads to the conclusion, potentially, or at least down that path of, like, well, we don't need them anymore. We just need... no one needs authors anymore, we only need critics, right? That's, you know, a very big difference between us. So, well, you know, the robot writes the book, and then I read it and criticize it, and that's a different skill set. It's overlapping in some ways, but it's different. And then at some point we go, but no one could do this anymore, and we go, well, that doesn't matter, because nobody's doing that. Why would we do that? You know, nobody does that. No one needs to do that, except, you know, like...
487
00:31:57.970 --> 00:31:58.160
Yeah.
488
00:31:58.160 --> 00:31:58.540
Hmm. Mm-hmm.
489
00:31:58.540 --> 00:32:00.000
Weird Tim over there.
490
00:32:00.000 --> 00:32:00.580
Ha ha. Yeah.
491
00:32:00.580 --> 00:32:03.980
Tim loves to write it himself, you know, doing it in his way.
492
00:32:03.980 --> 00:32:04.720
Yes.
493
00:32:04.720 --> 00:32:11.670
And, you know, and everyone's going, that's lovely. What a lovely pastime that is. And that becomes genuinely a hobby, and it's utterly divorced from profession.
494
00:32:11.670 --> 00:32:11.880
Right.
495
00:32:11.880 --> 00:32:12.250
I get it.
496
00:32:12.250 --> 00:32:12.330
Yeah.
497
00:32:12.330 --> 00:32:12.420
Yes.
498
00:32:12.420 --> 00:32:12.500
Yeah.
499
00:32:12.500 --> 00:32:16.020
I've just re-derived what you said earlier because I'm a slow person.
500
00:32:16.020 --> 00:32:19.720
Well, false, but okay.
501
00:32:19.720 --> 00:32:19.960
Wow.
502
00:32:19.960 --> 00:32:27.840
Yeah. I don't know how I feel about any of the things that we've spoken about today, which has been my,
honestly, it's been my lived experience for the last... six months is...
503
00:32:27.840 --> 00:32:28.070
Yeah.
504
00:32:28.070 --> 00:32:29.750
I know, this just sort of, like...
505
00:32:29.750 --> 00:32:29.980
is...
506
00:32:29.980 --> 00:32:31.420
you know, worry combined with, like...
507
00:32:31.420 --> 00:32:31.660
On...
508
00:32:31.660 --> 00:32:33.090
I don't know, wonder. Yeah. Yeah.
509
00:32:33.090 --> 00:32:38.100
So there's an element of delight, yeah, delight and wonder at some of the things that I am able to do.
510
00:32:38.100 --> 00:32:42.400
A little bit of hollowness afterwards, feeling, if I've done something cool, then go, have I done it?
511
00:32:42.400 --> 00:32:42.640
Yeah.
512
00:32:42.640 --> 00:32:46.460
I don't know. Me? No. Yes. But like, I have a thing now. It's really cool.
513
00:32:46.460 --> 00:32:46.700
I...
514
00:32:46.700 --> 00:32:49.800
And no one else has this thing. And that's great.
515
00:32:49.800 --> 00:32:56.600
I heard something... I heard something a couple of weeks ago that actually was a great summary of this whole situation, which is...
516
00:32:56.600 --> 00:33:08.880
LLMs are a cheat code. And just as with a game, if you use a cheat code to just warp to the last level and beat the boss immediately, it's like... cool. What of it?
517
00:33:08.880 --> 00:33:09.260
Yeah.
518
00:33:09.260 --> 00:33:10.060
Now what? What?
519
00:33:10.060 --> 00:33:11.140
It's hollow and yeah.
520
00:33:11.140 --> 00:33:11.560
It's like...
521
00:33:11.560 --> 00:33:15.140
Okay, I didn't really do anything. Right. And it's like...
522
00:33:15.140 --> 00:33:15.480
Yeah.
523
00:33:15.480 --> 00:33:20.720
It gives you this power, but it takes away all of the elements that made it engaging.
524
00:33:20.720 --> 00:33:29.320
Right. There is a reason why I haven't Googled anything about, like, Skyward Sword, because I can enjoy it myself at the moment. I'm not cheating in any way.
525
00:33:29.320 --> 00:33:37.320
I'm just enjoying it for itself, and I will not... look up anything. I'm having a great time. But, yeah, like, I have also... but so, but to counter that...
526
00:33:37.320 --> 00:33:37.610
Yes.
527
00:33:37.610 --> 00:33:37.890
Yes.
528
00:33:37.890 --> 00:33:47.710
What did I play the other day? Oh, I played the original Legend of Zelda, which I'd never done before. You know, it's... it's perfectly tractable to beat it, but I was, like, I've got this morning.
529
00:33:47.710 --> 00:33:48.480
Mm-hmm. Mm-hmm.
530
00:33:48.480 --> 00:33:51.600
I've never beaten it before. I just want to see the game.
531
00:33:51.600 --> 00:33:51.800
Right.
532
00:33:51.800 --> 00:33:54.940
And so I got the guide and I enjoyed playing the thing from one end to the other.
533
00:33:54.940 --> 00:33:55.060
Right.
534
00:33:55.060 --> 00:34:02.420
And I was like, oh, that's why this is great. And of course it would have been a... I'm going to say better, different experience to have
535
00:34:02.420 --> 00:34:08.790
pound it through it, but I haven't got the time, and this was my way of enjoying the game at least at some level and getting to the end, of being like, oh, there we go. Right.
536
00:34:08.790 --> 00:34:09.140
Mm hmm.
537
00:34:09.140 --> 00:34:09.490
Mm hmm.
538
00:34:09.490 --> 00:34:09.660
Brilliant.
539
00:34:09.660 --> 00:34:10.360
Mm hmm.
540
00:34:10.360 --> 00:34:16.510
So, yeah, it's different. And, you know, there have been analogies about, you know, like... it's like sending the robot to the gym for you so that, you know...
541
00:34:16.510 --> 00:34:17.060
Yeah. It's different.
542
00:34:17.060 --> 00:34:20.050
But, like, that implies that you want to do that thing, right?
543
00:34:20.050 --> 00:34:20.820
Yes. Yeah.
544
00:34:20.820 --> 00:34:23.100
Or that it's valuable to you. And maybe it isn't. But...
545
00:34:23.100 --> 00:34:23.100
Yeah.
546
00:34:23.100 --> 00:34:35.720
I just had to find this on my phone here. This is from... Alice Averlong, I don't know how to pronounce her name, but, like, the AI industry is full of people who would automate away playing with a kitten.
547
00:34:35.720 --> 00:34:36.320
Hmm.
548
00:34:36.320 --> 00:34:45.870
And I'm like, yeah, that's when you've gone too far. There's a point at which you've gone too far. That's the thing. It's like, oh, yeah, yeah, hey, I can... you know, you go and do the fun thing. I mean, we talked about this last night, right?
549
00:34:45.870 --> 00:34:46.140
Yeah.
550
00:34:46.140 --> 00:34:55.350
I'm sat here, like, that the... the computer is doing something that I would like to do, and I'm doing my tax returns. I'm like, this is exactly the wrong way around, right?
551
00:34:55.350 --> 00:34:55.600
Right. Right. Yeah.
552
00:34:55.600 --> 00:35:08.180
This is not how I want to be spending my time. But... but here we are. Yeah. So... Taking it all the way back in to try and conclude this, as I'm looking at the time and thinking...
553
00:35:08.180 --> 00:35:11.560
We're not very good at ending podcasts.
554
00:35:11.560 --> 00:35:12.920
Yeah, well...
555
00:35:12.920 --> 00:35:26.770
I mean, I think it's still an open question, all these years later, about whether Bill was right and that, you know, basically software, in order to be successful, needs to be a commercial activity.
556
00:35:26.770 --> 00:35:30.730
It certainly was not in the last 30 years. Right.
557
00:35:30.730 --> 00:35:31.130
Yeah.
558
00:35:31.130 --> 00:35:35.090
Forty years, really, we... yeah, 40 years at this point.
559
00:35:35.090 --> 00:35:35.490
I'm...
560
00:35:35.490 --> 00:35:35.880
Wow.
561
00:35:35.880 --> 00:35:37.470
God, I'm old. But...
562
00:35:37.470 --> 00:35:43.810
I just turned 50. So, you know, like it's 50 years from before to this time.
563
00:35:43.810 --> 00:35:44.200
Yeah.
564
00:35:44.200 --> 00:35:50.540
So, I mean, you're not old yet.
565
00:35:50.540 --> 00:35:54.220
Yeah. So it certainly has not been since, like, the sort of... you know, birth of the commercial
internet and the sort of booming of open source software and the communities around them.
566
00:35:54.220 --> 00:36:00.540
Like... it absolutely was not true. Will it be true in the next 40 years?
567
00:36:00.540 --> 00:36:01.500
Yeah.
568
00:36:01.500 --> 00:36:02.480
We're going to find out.
569
00:36:02.480 --> 00:36:04.580
We are going to find out one way or another.
570
00:36:04.580 --> 00:36:06.140
Yeah.
571
00:36:06.140 --> 00:36:07.680
Or... Roko's basilisk is going to torture us all to death.
572
00:36:07.680 --> 00:36:08.380
Right.
573
00:36:08.380 --> 00:36:19.940
in a very... Wow, that's a... No, we can't end there. We can't end there. I'm not even above... I know I'm too lazy to edit that out. So now we just have to turn it around and be like, well, that might not happen. It may be like, what was the opposite of it?
574
00:36:19.940 --> 00:36:22.960
I think Davison's weasel is the opposite of that.
575
00:36:22.960 --> 00:36:23.120
Mm hmm.
576
00:36:23.120 --> 00:36:24.880
And that would come and make everyone lovely and everything would be great.
577
00:36:24.880 --> 00:36:25.790
We're going to have a benevolent...
578
00:36:25.790 --> 00:36:26.280
Benevolent AI.
579
00:36:26.280 --> 00:36:28.720
AI that's going to... treat us all like pets.
580
00:36:28.720 --> 00:36:34.290
I'm... I want the Culture AI robot. That's what I'm hoping for. You know.
581
00:36:34.290 --> 00:36:35.260
Yeah.
582
00:36:35.260 --> 00:36:42.460
Iain Banks, no. Anyway. Well, friend, before I open a whole new worm can... Food for thought.
583
00:36:42.460 --> 00:36:42.940
Yeah.
584
00:36:42.940 --> 00:36:47.080
No conclusion at all, which is, you know, typical of this podcast.
585
00:36:47.080 --> 00:36:48.830
Yeah. Listeners discuss amongst yourselves.
586
00:36:48.830 --> 00:36:50.100
amongst your... yourself.
587
00:36:50.100 --> 00:36:51.060
Yeah, right.
588
00:36:51.060 --> 00:36:56.280
Singular. Oh, that's... now it's getting complicated. I've... I've met... now I've... that joke is... we're wearing thin with everybody I meet now. I... I...
589
00:36:56.280 --> 00:37:00.680
No, there's at least two, maybe even three listeners out there, at least.
590
00:37:00.680 --> 00:37:01.760
So... really?
591
00:37:01.760 --> 00:37:03.720
We should have a meetup.
592
00:37:03.720 --> 00:37:11.110
I mean, yeah, I suppose so. We could do. Yeah, that would be an interesting one. I mean, we've talked about doing a live version of this, or doing a recorded version of this, but that's far too much effort involving...
593
00:37:11.110 --> 00:37:11.520
Yes.
594
00:37:11.520 --> 00:37:13.700
I'd actually have to care about my appearance, which at the moment is...
595
00:37:13.700 --> 00:37:14.680
Oh my goodness.
596
00:37:14.680 --> 00:37:15.550
Yeah, no one wants to do that.
597
00:37:15.550 --> 00:37:16.040
Wow. That seems like a lot of work.
598
00:37:16.040 --> 00:37:17.490
But maybe another time. Anyway.
599
00:37:17.490 --> 00:37:17.640
Yeah.
600
00:37:17.640 --> 00:37:21.260
Let us know in the comments or whatever.
601
00:37:21.260 --> 00:37:23.760
Don't forget to like and subscribe. No, what am I supposed to say?
602
00:37:23.760 --> 00:37:24.260
And I... I don't know, man.
603
00:37:24.260 --> 00:37:24.380
I forget now. I don't know.
604
00:37:24.380 --> 00:37:25.260
I don't do this. This is not my jam.
605
00:37:25.260 --> 00:37:27.180
No. That's not your job now. This is not a job.
606
00:37:27.180 --> 00:37:28.320
No.
607
00:37:28.320 --> 00:37:30.900
Anyway, friend, this has been fun.
608
00:37:30.900 --> 00:37:31.660
Yes.
609
00:37:31.660 --> 00:37:33.920
I will speak to you another time.
610
00:37:33.920 --> 00:37:35.920
Next time.