WEBVTT

00:00:00.150 --> 00:00:02.310
Software is dead.

00:00:02.650 --> 00:00:04.570
That's a statement I've seen a couple

00:00:04.570 --> 00:00:05.770
of times over the last couple of days.

00:00:05.770 --> 00:00:08.170
It's a quite heated debate, obviously,

00:00:08.170 --> 00:00:10.200
because the underlying question, I think,

00:00:10.219 --> 00:00:10.699
of course,

00:00:11.050 --> 00:00:14.090
is whether software development is dead.

00:00:14.417 --> 00:00:16.897
Or if paid software is dead,

00:00:16.933 --> 00:00:18.373
and there are a lot of strong

00:00:18.373 --> 00:00:20.133
opinions there, and I have some too,

00:00:20.293 --> 00:00:22.053
and some nuances too. So let's dive in.

00:00:22.053 --> 00:00:24.933
This, of course, was a post by levelsio,

00:00:24.933 --> 00:00:26.773
Pieter Levels. You may know him from X.

00:00:26.773 --> 00:00:28.213
He got some strong opinions,

00:00:28.773 --> 00:00:30.613
and he's stating that it's now

00:00:30.613 --> 00:00:32.053
the era of personal software,

00:00:32.133 --> 00:00:33.893
and everyone just builds what they need

00:00:33.973 --> 00:00:36.613
whenever they need it. I think. Now,

00:00:36.950 --> 00:00:39.110
I definitely disagree with this statement,

00:00:39.350 --> 00:00:41.990
and he also kind of clarified it

00:00:41.990 --> 00:00:43.510
thereafter. I'll get back to that.

00:00:43.850 --> 00:00:44.250
Now,

00:00:44.583 --> 00:00:46.983
a couple of answers there, mostly relate

00:00:46.983 --> 00:00:48.823
to him being a hypocrite or something.

00:00:49.223 --> 00:00:51.943
But then there is interesting

00:00:52.023 --> 00:00:53.143
responses like this one.

00:00:53.483 --> 00:00:54.523
Don't you think most people are

00:00:54.523 --> 00:00:56.363
just lazy and not eager to change?

00:00:56.603 --> 00:00:58.363
I get the premise of micro apps, but I

00:00:58.363 --> 00:01:00.043
feel like people won't have the patience

00:01:00.123 --> 00:01:01.563
to deal with going back and forth with

00:01:01.563 --> 00:01:03.867
AI to solve bugs or tune to their liking

00:01:03.987 --> 00:01:05.027
if their current apps do

00:01:05.027 --> 00:01:06.227
the job without an issue.

00:01:06.787 --> 00:01:10.307
And I strongly agree here. I do not think

00:01:10.650 --> 00:01:12.730
that software is dead because normal

00:01:12.730 --> 00:01:14.650
people will just building their own

00:01:14.650 --> 00:01:17.370
Photoshop clones or anything like that.

00:01:17.610 --> 00:01:20.250
Totally not the future I see happening.

00:01:20.330 --> 00:01:22.890
But one thing I do see happening,

00:01:22.890 --> 00:01:24.170
which is what Pieter Levels

00:01:24.170 --> 00:01:25.610
then also stated thereafter,

00:01:26.090 --> 00:01:28.410
is that they're not doing it in Claude

00:01:28.410 --> 00:01:30.490
Code in a terminal like tech people.

00:01:30.996 --> 00:01:33.796
But their AI chat app is already doing

00:01:33.796 --> 00:01:36.036
it for them without them even realizing.

00:01:36.036 --> 00:01:37.556
They're building personal software,

00:01:37.556 --> 00:01:39.716
like generating PDF reports, full web app

00:01:39.716 --> 00:01:42.036
dashboards, and even entire web apps.

00:01:42.436 --> 00:01:43.796
And that's the part I see.

00:01:44.117 --> 00:01:45.957
Now, still with a lot of nuance,

00:01:45.957 --> 00:01:48.757
though, but I do see that, of course,

00:01:49.157 --> 00:01:51.477
normies—and I hate that term; it's so

00:01:51.477 --> 00:01:56.437
disrespectful—but normies will ask ChatGPT

00:01:56.757 --> 00:01:58.757
to remove the background of an image,

00:01:58.917 --> 00:02:01.237
and behind the scenes, ChatGPT may write

00:02:01.237 --> 00:02:03.797
some Python code or whatever and do it.

00:02:04.117 --> 00:02:04.837
Stuff like that.

00:02:04.997 --> 00:02:06.917
The same for a PDF report. Behind the

00:02:06.917 --> 00:02:10.037
scenes, ChatGPT writes HTML code and then

00:02:10.037 --> 00:02:12.837
writes more code that converts that to

00:02:12.837 --> 00:02:15.477
a PDF or prints that as a PDF document.

00:02:15.477 --> 00:02:17.157
That's already possible today.

00:02:17.157 --> 00:02:19.797
Obviously, now it's worth noting though

00:02:19.877 --> 00:02:22.037
that as Andrej Karpathy also mentioned.

00:02:22.482 --> 00:02:24.802
The vast majority of people are not

00:02:24.802 --> 00:02:28.882
really using AI. Only a very small group

00:02:28.882 --> 00:02:31.568
is actually using it, and even there,

00:02:31.808 --> 00:02:34.048
not a lot of people are paying.

00:02:34.128 --> 00:02:36.368
You may know that chart that only like

00:02:36.368 --> 00:02:38.448
two percent of U.S. households or

00:02:38.448 --> 00:02:40.288
something like this are paying for AI.

00:02:40.528 --> 00:02:43.008
So adoption is super slow,

00:02:43.168 --> 00:02:45.568
but that is the direction I see. That

00:02:45.568 --> 00:02:48.128
people will ask for problems to be solved,

00:02:48.128 --> 00:02:50.288
and those AI assistants will be able to

00:02:50.288 --> 00:02:53.071
solve a lot of problems. Now. Again,

00:02:53.207 --> 00:02:54.887
and I know people on the internet hate

00:02:54.967 --> 00:02:57.047
that, but you have to differentiate.

00:02:57.383 --> 00:02:59.483
There will be normal people who

00:02:59.483 --> 00:03:01.063
just want to remove the background,

00:03:01.063 --> 00:03:02.583
just generate a PDF, and yeah,

00:03:02.663 --> 00:03:04.503
AI assistants will do that for them now.

00:03:04.927 --> 00:03:07.647
Can you quickly just tell me how much did

00:03:07.647 --> 00:03:09.887
these people pay for software before?

00:03:09.887 --> 00:03:12.687
Oh, right, it was nothing. You did not pay

00:03:12.687 --> 00:03:16.127
for a PDF converter before. Maybe you did.

00:03:16.127 --> 00:03:18.296
Most people didn't. They were searching

00:03:18.296 --> 00:03:20.376
the web until they found some scammy

00:03:20.376 --> 00:03:22.696
website that would do it for them.

00:03:22.936 --> 00:03:25.256
They would upload their bank statements

00:03:25.336 --> 00:03:28.296
to some weird website where whatever

00:03:28.296 --> 00:03:31.496
happens with that data, and they would get

00:03:31.576 --> 00:03:34.216
the combined PDF or whatever out of it.

00:03:34.376 --> 00:03:36.216
And nowadays they do it with AI,

00:03:36.216 --> 00:03:38.456
where whatever may happen with that data,

00:03:38.456 --> 00:03:40.696
I guess. So not much changes there.

00:03:40.850 --> 00:03:43.170
Now you know who was or is paying

00:03:43.170 --> 00:03:46.050
for software, professionals, and

00:03:46.383 --> 00:03:48.943
if it's your hobby. Now, games, by the

00:03:48.943 --> 00:03:50.783
way, are a different kind of software.

00:03:50.863 --> 00:03:53.343
I don't see people vibing their own

00:03:53.343 --> 00:03:55.583
games because part of a nice game is

00:03:55.583 --> 00:03:57.503
not knowing what's going to happen when

00:03:57.503 --> 00:03:59.263
and how to perfectly beat the game.

00:03:59.263 --> 00:04:00.303
But that's a different story.

00:04:00.623 --> 00:04:02.463
But if we think about other hobbies.

00:04:02.782 --> 00:04:03.100
Let's say

00:04:03.100 --> 00:04:04.780
you're a hobby photographer,

00:04:04.860 --> 00:04:08.220
or you're a hobby video creator.

00:04:08.550 --> 00:04:11.430
Maybe you paid for some Adobe products.

00:04:11.430 --> 00:04:14.390
You paid for the Photoshop.

00:04:14.390 --> 00:04:16.710
You paid for Premiere Pro.

00:04:16.710 --> 00:04:17.750
Something like that.

00:04:18.083 --> 00:04:20.163
I don't see that those people will

00:04:20.163 --> 00:04:22.083
necessarily vibe code their own

00:04:22.083 --> 00:04:23.763
alternatives because guess what?

00:04:23.763 --> 00:04:27.043
If you do that, you have to maintain it.

00:04:27.123 --> 00:04:29.283
You have to fix all those annoying

00:04:29.283 --> 00:04:31.203
bugs you'll encounter over time.

00:04:31.283 --> 00:04:33.443
That may be fine for certain tasks,

00:04:33.603 --> 00:04:34.563
but if it's your hobby,

00:04:34.563 --> 00:04:36.483
if you're working a lot with it, if you're

00:04:36.483 --> 00:04:38.259
interested in the more advanced features.

00:04:38.579 --> 00:04:40.339
If you're maybe also interested in

00:04:40.579 --> 00:04:42.739
getting new features you didn't expect

00:04:42.739 --> 00:04:45.139
in the future, then you really want

00:04:45.139 --> 00:04:47.619
someone else developing that for you.

00:04:47.950 --> 00:04:50.990
Now prices may go down though because

00:04:50.990 --> 00:04:53.710
now it's easier than ever to launch your

00:04:53.710 --> 00:04:55.710
Photoshop alternative, and indeed I'll

00:04:55.710 --> 00:04:57.710
get back to that thing and this entire

00:04:57.710 --> 00:04:59.470
reverse engineering trend which we have

00:04:59.470 --> 00:05:01.070
right now and the downsides of that.

00:05:01.070 --> 00:05:03.150
But yeah, prices may go down,

00:05:03.150 --> 00:05:05.630
but I definitely see those people not

00:05:05.710 --> 00:05:07.710
vibe coding their own alternatives now.

00:05:08.050 --> 00:05:09.650
There will be some middle ground here.

00:05:09.970 --> 00:05:11.650
It's absolutely possible that

00:05:11.650 --> 00:05:14.210
you had certain things you did,

00:05:14.520 --> 00:05:16.120
which were very specific,

00:05:16.200 --> 00:05:19.160
and now you just ask ChatGPT or Claude to

00:05:19.160 --> 00:05:21.560
build you a tool that does that one thing,

00:05:21.560 --> 00:05:23.400
which you do over and over again.

00:05:23.640 --> 00:05:25.000
On its own, so that it builds you

00:05:25.000 --> 00:05:26.520
a tool for one specific task.

00:05:26.600 --> 00:05:28.120
I definitely see that this could

00:05:28.120 --> 00:05:32.600
happen right now already, but soon I

00:05:32.600 --> 00:05:35.080
don't see those hobbyists and so on

00:05:35.417 --> 00:05:37.817
switch away from paid software necessarily

00:05:37.897 --> 00:05:40.217
now. Long term on the horizon,

00:05:40.217 --> 00:05:41.737
nobody knows what happens. Right,

00:05:41.977 --> 00:05:45.017
with AI, that is impossible to predict.

00:05:45.017 --> 00:05:46.697
There's too much going on, so

00:05:46.867 --> 00:05:48.947
it's impossible to say what will be the

00:05:48.947 --> 00:05:51.267
case or what will be true in ten years

00:05:51.267 --> 00:05:53.827
or even five years. But short term,

00:05:53.827 --> 00:05:55.667
yeah, I don't see that now.

00:05:56.017 --> 00:05:59.137
Normal people, hobbyists are, of course,

00:05:59.483 --> 00:06:01.723
there. They're not the most interesting

00:06:01.723 --> 00:06:03.323
market for the reasons I stated.

00:06:03.643 --> 00:06:04.843
You know what is interesting?

00:06:05.208 --> 00:06:06.808
Companies, businesses,

00:06:07.208 --> 00:06:08.728
and here again we have to differentiate.

00:06:09.128 --> 00:06:11.928
Large corporations—they will be slow.

00:06:12.168 --> 00:06:15.928
There is a lot of inertia here, so they

00:06:15.928 --> 00:06:20.168
will be slow in adopting AI solutions.

00:06:20.168 --> 00:06:22.728
They will likely not vibe code all their

00:06:22.728 --> 00:06:24.728
code or all their software anytime

00:06:24.728 --> 00:06:26.968
soon for a broad variety of reasons:

00:06:26.968 --> 00:06:29.928
compliance, aligning different teams,

00:06:30.088 --> 00:06:31.448
the maintenance burden,

00:06:31.768 --> 00:06:34.088
all that fun stuff, and even more.

00:06:34.417 --> 00:06:36.577
Small business owners, different thing.

00:06:36.897 --> 00:06:38.577
They will definitely also pay for some

00:06:38.577 --> 00:06:40.897
software for mission critical parts

00:06:40.897 --> 00:06:42.977
where they really don't want that bug to

00:06:42.977 --> 00:06:45.617
surface right when they need the software.

00:06:45.950 --> 00:06:48.910
But for those annoying side tracks,

00:06:49.150 --> 00:06:51.311
I definitely see that. Or for other

00:06:51.311 --> 00:06:53.550
things like your website, and I mean,

00:06:53.790 --> 00:06:56.590
small business owners before were likely

00:06:56.590 --> 00:06:58.990
using something like Wix or Squarespace.

00:06:58.990 --> 00:07:01.550
Now they can use ChatGPT with ChatGPT

00:07:01.550 --> 00:07:03.070
sites, or they can use Lovable.

00:07:03.390 --> 00:07:05.390
So they're just switching.

00:07:05.470 --> 00:07:07.710
They weren't primarily interested in

00:07:08.110 --> 00:07:09.950
software built by other developers

00:07:09.950 --> 00:07:12.048
there before. Well, I guess Wix,

00:07:12.048 --> 00:07:14.128
but Lovable is also built by developers.

00:07:14.128 --> 00:07:16.848
If you look at it like this, so that is

00:07:16.848 --> 00:07:19.408
some switch I see here, and I do also see,

00:07:19.408 --> 00:07:21.248
by the way, a nice opportunity here.

00:07:21.328 --> 00:07:23.328
If you're a developer, a freelancer,

00:07:23.328 --> 00:07:26.319
let's say. Offering to your customers that

00:07:26.559 --> 00:07:29.679
you are building software, whatever it is,

00:07:29.679 --> 00:07:32.319
with AI, but also with you,

00:07:32.399 --> 00:07:34.319
the human in the loop, that can be

00:07:34.319 --> 00:07:37.039
a nice business use case right now.

00:07:37.039 --> 00:07:39.039
Because I think there will be lots

00:07:39.039 --> 00:07:40.799
of business owners and generally

00:07:40.879 --> 00:07:44.399
people that are looking to lower

00:07:44.399 --> 00:07:46.879
their costs or get a result quicker,

00:07:46.973 --> 00:07:48.173
but that don't want to go

00:07:48.173 --> 00:07:49.773
all in on AI necessarily.

00:07:49.773 --> 00:07:51.533
They may not want to use Lovable.

00:07:51.693 --> 00:07:53.933
They may want a developer who uses

00:07:53.933 --> 00:07:56.493
Lovable or whatever Claude Code instead.

00:07:56.493 --> 00:07:57.533
So I think that, by the way,

00:07:57.533 --> 00:07:58.973
is a nice business opportunity

00:07:59.453 --> 00:08:01.933
right now and in the near future.

00:08:01.933 --> 00:08:03.933
Again, long term, nobody knows.

00:08:04.283 --> 00:08:06.763
Now, what about software getting cheaper?

00:08:06.763 --> 00:08:09.403
There being more alternatives, of course,

00:08:09.483 --> 00:08:11.723
competition is definitely ramping up,

00:08:11.723 --> 00:08:15.049
and we have projects like this,

00:08:15.449 --> 00:08:18.409
REA repository, reverse engineer anything

00:08:18.489 --> 00:08:20.169
which is in the end an MCP you can give

00:08:20.249 --> 00:08:23.609
to your AI so that it can easily inspect

00:08:23.609 --> 00:08:26.329
binaries and reverse engineer software.

00:08:26.649 --> 00:08:27.849
We had this

00:08:28.115 --> 00:08:31.155
ArtCraft thing where people rebuilt

00:08:31.155 --> 00:08:34.115
the entire Adobe suite. Broken.

00:08:34.595 --> 00:08:38.515
I only played around with it, but yeah,

00:08:38.595 --> 00:08:40.355
these are things that are happening.

00:08:40.355 --> 00:08:40.915
I think

00:08:41.250 --> 00:08:43.970
it's easier than ever to build software.

00:08:44.210 --> 00:08:46.050
Good software still

00:08:46.383 --> 00:08:49.830
tricky. With vibe coding alone may not

00:08:49.830 --> 00:08:52.230
be the path to get there, but again,

00:08:52.390 --> 00:08:54.410
who knows what better models in the

00:08:54.410 --> 00:08:56.090
near future, or definitely in a

00:08:56.090 --> 00:08:57.670
couple of years, will be able to do.

00:08:57.990 --> 00:09:00.070
But right now, not necessarily there yet.

00:09:00.390 --> 00:09:03.190
And if you're thinking about using a piece

00:09:03.190 --> 00:09:05.830
of software, you may not want to go for

00:09:05.830 --> 00:09:07.910
the reverse-engineered, vibe-coded ones

00:09:07.910 --> 00:09:10.150
because nobody may be maintaining them.

00:09:10.470 --> 00:09:13.190
You don't know exactly if you can trust

00:09:13.430 --> 00:09:16.567
these software products now. Of course,

00:09:16.887 --> 00:09:18.887
part of the truth also is there's a lot

00:09:18.887 --> 00:09:21.527
of paid shitty software out there too.

00:09:22.007 --> 00:09:23.207
I'm not looking at Adobe here,

00:09:23.207 --> 00:09:25.767
of course, but there is software out

00:09:25.767 --> 00:09:29.127
there with shitty business practices or

00:09:29.127 --> 00:09:31.207
where the software itself sucks, and

00:09:31.550 --> 00:09:34.350
I think it's a good thing that software

00:09:34.430 --> 00:09:37.470
is easier to create, especially by

00:09:37.470 --> 00:09:38.750
people who know what they're doing.

00:09:38.830 --> 00:09:40.990
Especially if you're not just vibe coding,

00:09:40.990 --> 00:09:42.190
but if you're leveraging your

00:09:42.190 --> 00:09:43.230
knowledge as a developer,

00:09:43.630 --> 00:09:46.510
and that can ultimately be a nice way of

00:09:46.850 --> 00:09:49.490
entering a niche or a market that

00:09:49.490 --> 00:09:51.170
was way harder to enter before,

00:09:51.170 --> 00:09:53.490
because you can probably build your

00:09:53.490 --> 00:09:57.170
Premiere Pro clone in a not shitty version

00:09:57.410 --> 00:09:59.330
and maybe gain some customers there.

00:09:59.330 --> 00:10:02.370
So I think right now, in short term,

00:10:02.610 --> 00:10:04.370
this is a very good thing.

00:10:04.717 --> 00:10:06.717
Now, long term on the horizon,

00:10:06.717 --> 00:10:09.757
as AI does everything, if that is where

00:10:09.757 --> 00:10:12.077
we're going, and nobody knows, of course,

00:10:12.557 --> 00:10:15.677
what's the remaining moat, which software

00:10:15.677 --> 00:10:18.533
will survive in the sense of people

00:10:18.533 --> 00:10:19.973
will still pay for it, and I guess

00:10:19.973 --> 00:10:21.413
it's not just software, by the way.

00:10:21.573 --> 00:10:23.253
You can ask this question general:

00:10:23.253 --> 00:10:25.013
which businesses will survive in

00:10:25.013 --> 00:10:27.493
a future where AI does a lot?

00:10:27.998 --> 00:10:29.438
And one or a couple of things you

00:10:29.438 --> 00:10:31.358
hear frequently there are data.

00:10:31.438 --> 00:10:32.958
If your business, your software,

00:10:32.958 --> 00:10:35.039
whatever it is, depends on data,

00:10:35.118 --> 00:10:37.512
only you have. Community.

00:10:37.512 --> 00:10:41.272
If you can get people together, content.

00:10:41.272 --> 00:10:43.192
If you are creating content that is

00:10:43.192 --> 00:10:46.952
original and hard to copy or hard to

00:10:46.952 --> 00:10:49.432
recreate by AI in the way you do it.

00:10:49.750 --> 00:10:52.630
That all is something that could work,

00:10:52.710 --> 00:10:54.390
and there probably are more moats.

00:10:54.790 --> 00:10:57.510
But ultimately, I think we just have to

00:10:57.830 --> 00:11:00.070
accept and, as much as it's possible,

00:11:00.150 --> 00:11:03.430
embrace the fact that we don't know

00:11:03.590 --> 00:11:05.350
where we're heading. And this is hard.

00:11:05.590 --> 00:11:07.110
I totally get it. And by the way,

00:11:07.110 --> 00:11:08.870
we're talking about my business here too.

00:11:08.870 --> 00:11:10.150
We're talking about our jobs.

00:11:10.150 --> 00:11:11.270
We're talking about businesses.

00:11:11.270 --> 00:11:13.510
My business. This is hard. But the

00:11:13.510 --> 00:11:16.710
reality is that nobody knows what will be

00:11:17.050 --> 00:11:19.290
the case in two or three years and so on.

00:11:19.450 --> 00:11:21.610
We can only plan for right now. Right now,

00:11:21.610 --> 00:11:23.690
there are many opportunities here,

00:11:23.930 --> 00:11:25.610
and we don't know if the data

00:11:25.610 --> 00:11:27.050
part will matter in the future.

00:11:27.050 --> 00:11:28.730
We don't know if the community part will

00:11:28.730 --> 00:11:30.970
matter, or if we all get more lonely

00:11:30.970 --> 00:11:33.050
because of AI. Because AI does it all,

00:11:33.050 --> 00:11:34.570
and we rely on it so much that we

00:11:34.570 --> 00:11:36.890
don't wanna get into any community.

00:11:37.130 --> 00:11:38.970
We don't know if we get so used to

00:11:38.970 --> 00:11:41.130
AI-generated content that that's the

00:11:41.210 --> 00:11:42.810
only content we want to consume.

00:11:43.150 --> 00:11:43.870
We don't know it.

00:11:44.217 --> 00:11:46.457
So the best thing we can do is try to

00:11:46.457 --> 00:11:48.457
see opportunities we have today and

00:11:48.457 --> 00:11:50.297
plan for today and the near future.

00:11:50.377 --> 00:11:52.457
As unsatisfying as that is,

00:11:52.857 --> 00:11:55.017
but to get back to that initial statement,

00:11:55.337 --> 00:11:57.657
software is absolutely not dead.

00:11:57.657 --> 00:11:59.737
We'll have more code and more software

00:11:59.737 --> 00:12:01.737
than ever, in one form or the other.

00:12:02.137 --> 00:12:04.857
Software development also not dead right

00:12:04.857 --> 00:12:07.177
now. Also not dead in the near future.

00:12:07.497 --> 00:12:08.377
Long term.

00:12:08.717 --> 00:12:09.357
Nobody knows,

00:12:09.683 --> 00:12:12.643
but being able to adapt and use these

00:12:12.643 --> 00:12:14.563
tools is the best thing we can do.

00:12:14.803 --> 00:12:15.283
I guess.