Software is DEAD. SaaS is DEAD. AI has solved EVERYTHING! — Transcript

Full transcript

1. 0:00Okay, so I just got to say that I've
2. 0:03been seeing a lot of stuff on Twitter
3. 0:06and on YouTube, a lot of videos and
4. 0:09people making posts about how software
5. 0:12is dead, SAS is dead, everything's dead,
6. 0:15right? Like coding is dead. The problem
7. 0:17has been solved. It's cheap. It's no
8. 0:20longer a bottleneck. It's no longer a
9. 0:22moat. It is something that we don't have
10. 0:26to worry about. And the industry is
11. 0:28changing in that sense. And I wanted to
12. 0:30just give my opinion on it and talk a
13. 0:32little bit about that because I've got
14. 0:35mixed feelings about that. And I'm a
15. 0:37little bit frustrated that I keep
16. 0:39hearing so many people say this. It's
17. 0:41like an echo chamber. I think I'm on
18. 0:43tech Twitter too much. So the big claim
19. 0:45is that software is dead and SAS is
20. 0:48dead. I think this first came out with
21. 0:50the tweet from Peter Levelvels where he
22. 0:53said software is dead. It's now the era
23. 0:57of personal software and everyone just
24. 0:59builds what they need whenever they need
25. 1:01it. I think and he was actually quoting
26. 1:03a tweet by S for Satoshi where he said
27. 1:07it is the beginnings of the end of
28. 1:09software. AI is starting to become good
29. 1:12enough at reverse engineering and
30. 1:14improving almost anything. The latest
31. 1:16models were a step change in
32. 1:18capabilities that flew under the radar.
33. 1:21I got to say like I'm thinking like man
34. 1:23yeah maybe we have a lot of people who
35. 1:25are non-devs that are vibe coding
36. 1:27working apps right and we have people
37. 1:29building personal tools instead of
38. 1:31buying them like that makes sense right
39. 1:33but that's a small group of people
40. 1:35that's the bubble on Twitter that's the
41. 1:38software developer bubble and it's not
42. 1:40even all software developers it's still
43. 1:42like a small percentage of them that are
44. 1:44doing this and that are saying this so
45. 1:46does this mean that if software is
46. 1:48easier to build that it's dying I I
47. 1:50don't think so. But of course, Peter
48. 1:53Levelvels has a huge following and this
49. 1:56kind of, you know, caused a lot of
50. 1:57ripples and had a lot of people talking
51. 1:59when he posted this last month. And I
52. 2:01mean, I don't think that Levelvels is
53. 2:03wrong. I think that he makes a valid
54. 2:05point. And I think that there are
55. 2:07people, especially capable developers
56. 2:09and even not so capable vibe coders,
57. 2:12that are building their own products and
58. 2:14their own solutions to little problems
59. 2:17that they have. And I think that's a
60. 2:18good thing. I don't think that's a bad
61. 2:20thing. I think that's awesome. It's
62. 2:22innovation. People are building more.
63. 2:24That means more software, which kind of
64. 2:26doesn't mean that software is dead. It's
65. 2:27kind of the opposite. But then you got
66. 2:29someone like Nikita Beer and he tweets,
67. 2:31"Technical Moes are mostly over at this
68. 2:34point. If you're building something that
69. 2:36is meant to be durable, the only moes
70. 2:39now are proprietary data, network
71. 2:42effects, and a physical component. But I
72. 2:46also don't think that the technical
73. 2:48modes are officially over at this point.
74. 2:52I think that people who are well-versed
75. 2:55in the technical aspects and understand
76. 2:58how to implement things and how to use
77. 3:01these tools can cross those emotes
78. 3:03easier than most people. And that's
79. 3:06still a small number of the population.
80. 3:08And that still mostly means software
81. 3:11developers. So, I think that those
82. 3:13technical moes technically are still
83. 3:15there. Sometimes these are things that
84. 3:17people are posting while they're sitting
85. 3:18on their toilet. So, you got to take
86. 3:19anything you see on Twitter with a grain
87. 3:20of salt, but they do get a lot of views
88. 3:23and they cause a lot of commotion. And
89. 3:25so, there is some truth to some of the
90. 3:26things that they're saying. Another
91. 3:28thing that I saw was the REA repo that
92. 3:31everybody was talking about recently,
93. 3:32which is the reverse engineer anything.
94. 3:34I mean, it's blown up. I think it's got
95. 3:36over 400,000 stars on GitHub and it's
96. 3:38been forked who knows how many times
97. 3:40people have forked it by now. The idea
98. 3:42behind it is actually really cool,
99. 3:43right? You can clone any software that
100. 3:45you want. But what happens after you've
101. 3:48built a clone? There's a lot that goes
102. 3:49into software where just cloning a piece
103. 3:52of technology isn't the end goal. That's
104. 3:55just where you get started, right? You
105. 3:56know, I just get frustrated because we
106. 3:58we live in this bubble, this echo
107. 4:00chamber of tech Twitter and tech YouTube
108. 4:02and a lot of people talk about this
109. 4:04stuff and it gets clicks and you know,
110. 4:05it's part of the reason why I'm making
111. 4:07this video. But at the same time, I want
112. 4:08to, you know, tell you how I feel about
113. 4:10it because I feel that there's more
114. 4:12people that feel like me about this than
115. 4:15people who probably would agree with
116. 4:18like levels. Not to knock him, he's
117. 4:20obviously done very well for himself and
118. 4:22he's built software that a lot of people
119. 4:24use, but at the end of it, it's still a
120. 4:27lot of his stuff is like micro SASS or
121. 4:29like little features, right, that that
122. 4:31he sells and he makes good money off of,
123. 4:32but it's not true enterprisegrade
124. 4:35software that he's building. So, I feel
125. 4:37that it's hard for him to talk about
126. 4:39that stuff because I don't believe that
127. 4:41he's ever worked on enterprise level
128. 4:44software and has built big software for
129. 4:46big companies. I have experience doing
130. 4:48that and I think that a lot of people
131. 4:50that have that experience would agree.
132. 4:52I'm not saying that either of their
133. 4:53points are invalid. Right? The cost of
134. 4:56building software has dropped a lot.
135. 4:58Right? Small teams can now build like
136. 5:01big teams. What used to take a month can
137. 5:03now take a weekend for some people. And
138. 5:06personal software is real. We are at a
139. 5:08point now where if you've got, you know,
140. 5:11pain points as as a developer or even as
141. 5:13a regular person, you can turn to some
142. 5:15of these coding tools and take care of
143. 5:17those problems with software. Whereas
144. 5:19before, maybe you couldn't find specific
145. 5:21software to handle that exact problem.
146. 5:24And now you can build software to
147. 5:26literally take care of any technical
148. 5:29digital issue that you may have. The
149. 5:32bigger the issue, the harder it is to
150. 5:33build something that fixes that. But if
151. 5:35there is like one a tool that does one
152. 5:37thing that a person wants, they can just
153. 5:40build it themselves and they don't have
154. 5:41to go buy a piece of enterprisegrade
155. 5:44software that does a whole bunch of
156. 5:46things when they only want one feature
157. 5:48of that. They can build a small tool for
158. 5:50that one feature. And I again I think
159. 5:51that's awesome. And this is where it's
160. 5:53like micro SASS and thin SAS are more
161. 5:56exposed to this software is dead. It's
162. 5:59real easy to copy features, but it's
163. 6:01also like harder to charge to fix these
164. 6:06small little features. And that's why
165. 6:07people still turn to bigger solutions
166. 6:10that are more robust and offer all the
167. 6:13things that you need when you're trying
168. 6:14to solve a set of problems. That ends up
169. 6:17being more like a pricing problem. And
170. 6:19that's where, you know, developers who
171. 6:21just want one thing from something that
172. 6:23costs $100 a month can spend a weekend
173. 6:26building and then not have to pay that
174. 6:28hundred bucks. And it's great. You know,
175. 6:29that's that's that's awesome. I've got a
176. 6:31few tools like that myself. And yes,
177. 6:34code is a lot easier now. But anyone
178. 6:36who's worked in software knows that even
179. 6:38as a developer, code is kind of like
180. 6:42just a portion of our job. It's not the
181. 6:45whole job. And the longer you've done
182. 6:47this and the more experience you have as
183. 6:49a developer, you you'll know that that's
184. 6:51even more true. Maybe like a junior dev.
185. 6:53Yeah, they're the code monkey. They're
186. 6:55the script kitty. They're the ones that
187. 6:57write all the code. And you know,
188. 6:58they're also the ones struggling the
189. 7:00most now with this shift that we're
190. 7:01having. But senior level developers,
191. 7:04we're the problem solvers. We're the
192. 7:06ones that have to, you know, talk to
193. 7:08business and figure out what they want.
194. 7:10We're the ones that have to think about
195. 7:11the technical implementations. We got to
196. 7:13think about the architecture. We got to
197. 7:15think about the actual solutions. We got
198. 7:16to think about planning and reviewing
199. 7:19and maintaining code. It's not just
200. 7:23writing code. And the fact that a lot of
201. 7:25these tools has made it a lot easier for
202. 7:27us to write code. It's eliminated that
203. 7:29problem. But all of those other things
204. 7:31that we had to focus on are still there.
205. 7:33And writing code was never what building
206. 7:36software products was about. Some of the
207. 7:38loudest voices are these niche tool
208. 7:41builders, these micro SAS builders. A
209. 7:44lot of it comes from tech Twitter and
210. 7:46people in that bubble and the people
211. 7:48that are close to them in that space,
212. 7:50right? Like it's not a true
213. 7:52representation of businesses that pay
214. 7:55for software, large businesses and small
215. 7:57businesses alike. Many of the times
216. 8:00those are nonsoftware related businesses
217. 8:03and they don't want to spend money and
218. 8:06time and resources on building their own
219. 8:08software. Many of the times they just
220. 8:10want to pay for a solution to whatever
221. 8:13they need handled. And many of the times
222. 8:14if they do need software, they're not
223. 8:16going to bring in people on their own
224. 8:18unless they already kind of have a team
225. 8:20around that. Many of the times they'll
226. 8:22hire agencies, which is where, you know,
227. 8:24the company that I work for comes in. We
228. 8:26go in and we build software for
229. 8:27companies that don't want to build
230. 8:29software that don't want to have to deal
231. 8:31with that aspect because they have more
232. 8:34important things to focus on where that
233. 8:37generates their revenue and brings in
234. 8:38their money. And software is just
235. 8:40something that they need in order for
236. 8:41them to keep making money in the area
237. 8:43that they make money in. And that's
238. 8:45where a lot of real software is built.
239. 8:47And that's where a lot of tools are
240. 8:48sold. And that's where a lot of these
241. 8:50enterprise solutions get sold to. And
242. 8:54things that are being built for
243. 8:56companies around software aren't being
244. 8:58built by these non-technical companies.
245. 9:01They're still selling stuff. And people
246. 9:02are still buying it. And businesses will
247. 9:05always need software. And most of the
248. 9:07time, they're not going to want to build
249. 9:08it themselves. They're not going to want
250. 9:09to maintain it. they'll either pay
251. 9:11somebody to do it or they'll just pay
252. 9:12for the product that they need to solve
253. 9:15their problems. Now, we're seeing these
254. 9:17things like the REA repo and I think
255. 9:20there was somebody who went viral
256. 9:21recently. I don't know if he used that
257. 9:23repo, but I know he went and he
258. 9:25basically open sourced his own uh
259. 9:28variations of Adobe software. And if
260. 9:31anybody wants to see Adobe crash
261. 9:33[snorts] and burn, it would be me. I'm
262. 9:34not a big fan of Adobe. I've actually
263. 9:36worked for Fortune 500 and Fortune 100
264. 9:39companies that used Adobe products.
265. 9:42Adobe Experience Manager is their
266. 9:44content management system. I have a few
267. 9:46years of experience working in that. I
268. 9:48absolutely hate it. It's a terrible
269. 9:50piece of software, but I can tell you
270. 9:53that all of those Fortune 500 companies
271. 9:56aren't going to drop Adobe to go pick up
272. 9:59these clones or these open-source
273. 10:02products because it's too ingrained in
274. 10:04their system. And that's the truth for a
275. 10:06lot of things, right? So cloning isn't
276. 10:09owning. These clones skip the fact that
277. 10:12there's maintenance, there's overhead,
278. 10:14there's engineering needed to keep
279. 10:16software current that, you know, when
280. 10:18you clone something, you don't have a
281. 10:19roadmap. You don't have the security
282. 10:21patches that come, you don't have the
283. 10:22integration and the features that they
284. 10:24build for it. You don't have the support
285. 10:25or the compliance that's built into
286. 10:27them. And these are all things that is
287. 10:29very important to a lot of businesses.
288. 10:31And and that's why many of them will
289. 10:33prefer to keep their Adobe or whatever
290. 10:36license and have them deal with it than
291. 10:39to try to build their own in-house thing
292. 10:41or say, "Hey, you know what? We're going
293. 10:42to save some money here and we're going
294. 10:44to use this clone or this open- source
295. 10:46repo that could, you know, maybe become
296. 10:49a thing and people will start using or
297. 10:52will dwindle and die out and now we're
298. 10:54stuck having to maintain this ourselves
299. 10:56or hiring people that can fix and keep
300. 10:59things updated." Let's just say some big
301. 11:01business were to take a clone of
302. 11:04something that was built that replaces
303. 11:06something that they pay a ton of money
304. 11:08for. Well, now if they need to maintain
305. 11:12this, they need developers, right? They
306. 11:14can hire consultants. They can hire
307. 11:15freelancers. They can hire an agency.
308. 11:17They can hire their own devs, but they
309. 11:19still need devs even with AI. And you
310. 11:22still need people to maintain this
311. 11:23stuff. If you need developers, then
312. 11:25you're going to need project managers.
313. 11:26You're going to need product owners.
314. 11:28You're going to need scrum masters and
315. 11:30you're going to need all of the stuff
316. 11:31that comes with developers, right? I
317. 11:34mean, sure, you can have a small team
318. 11:36now with augmenting with AI, but then
319. 11:38you still need to pay for those tools,
320. 11:39right? And I will say that I think AI
321. 11:42tools aren't going to get any cheaper.
322. 11:44We've seen a lot recently with tooling
323. 11:46going up in prices and you know usages
324. 11:49being cut and new plans being promoted
325. 11:52and you know even Elon's coming out
326. 11:54saying that you know this is not
327. 11:56sustainable with with the Grockbot and
328. 11:58the usages and how much infrastructure
329. 12:00costs for that stuff. I think Replet's
330. 12:02uh CEO just came out saying something
331. 12:04about that as well. as someone who pays
332. 12:06hundreds of dollars for multiple tools.
333. 12:08When you do that at a, you know,
334. 12:10enterprise level or a large corporation,
335. 12:12now you're paying for tooling for
336. 12:14multiple devs. So, not only are you
337. 12:15paying developer salaries, you're paying
338. 12:16for all the tooling on top of that. So,
339. 12:19yeah, AI makes things easier, but it
340. 12:21still costs money to run and you're
341. 12:23still going to need devs. And when it
342. 12:24comes down to the bottom line, these
343. 12:26companies are going to figure out what's
344. 12:28cheaper. Can we just pay for the tool
345. 12:29that we already had or are we going to
346. 12:31build a whole team around maintaining a
347. 12:33tool that does the same thing that we
348. 12:35were already doing for a fraction of the
349. 12:37cost? And I want to just give a personal
350. 12:39experience, right? I've I've worked in
351. 12:40enterprise software for a while. I've
352. 12:42worked for big companies. I've worked
353. 12:45for smaller companies. I've worked for
354. 12:48the state of Washington and help build
355. 12:50out a massive application used by
356. 12:53millions of people in the state of
357. 12:54Washington. And I'm experienced with
358. 12:58large software, software that needs to
359. 13:02be used by a lot of people and software
360. 13:05that is critical for businesses to make
361. 13:06money. And right now I work for a web
362. 13:08agency. For the last 2 years, I've been
363. 13:11working on a pharmaceutical management
364. 13:13system that handles dispensing and
365. 13:16compounding. This was an evergreen
366. 13:17project where we were taking a legacy
367. 13:20piece of software and basically
368. 13:23refreshing it and making a new version
369. 13:26of it for a company that is paying us to
370. 13:29do so. And I don't know exactly what
371. 13:32we're charging, but I know it can't be
372. 13:33cheap. We have a large team of
373. 13:35developers. Just on the agency side, I
374. 13:37would say there's about 10 of us. And
375. 13:40you know, that's not a massive team.
376. 13:41That's not hundreds of developers, but
377. 13:43our company is very AI forward. We've
378. 13:45been using AI from the beginning and now
379. 13:48as we are transitioning from building
380. 13:51out the MVP and rolling it out to a
381. 13:53fewarmacies, the company that we work
382. 13:55for that has hired our dev agency that I
383. 13:58work for is starting to build their own
384. 14:00team. And I think that they have about
385. 14:02five or six developers. And this is just
386. 14:04developers, right? There's still a lot
387. 14:07of subject matter experts. There's still
388. 14:09a lot of product owners. There's still a
389. 14:11lot of project managers. There's a lot
390. 14:13of people involved in building this one
391. 14:17application that solves one of this
392. 14:21company's problems. This isn't their
393. 14:23only software. This is just a software
394. 14:26that was their biggest pain point that
395. 14:28they wanted to revamp and redo because
396. 14:31the old system just wasn't keeping up
397. 14:33with what they needed it for. There's
398. 14:35just so many moving parts and so many
399. 14:38people involved in the process of
400. 14:40building real enterprise software.
401. 14:42There's so many meetings to figure out
402. 14:45how things need to work. There's so much
403. 14:47going on and what I have noticed is that
404. 14:51we ship a lot faster. I think if there
405. 14:53wasn't AI the team would need to be
406. 14:55double but since we are AI forward and
407. 14:58we use a lot of these tools we ship
408. 15:00faster but now we have a lot of
409. 15:02different bottlenecks. code review. Even
410. 15:05though we use things like Code Rabbit
411. 15:07and C-Pilot to help us review code, like
412. 15:10there's still a a big human process
413. 15:12there because we have to make sure that
414. 15:14things are good, right? Like there's
415. 15:15there's security involved. there's
416. 15:17testing and there's there's QA so much
417. 15:21that is involved in building big
418. 15:24software that when I see people who
419. 15:28have only experience building small
420. 15:31software small software that is just
421. 15:35direct to customer you know it does one
422. 15:38thing and it has a stripe payment and
423. 15:40then that's it right like there's not
424. 15:42much involved in that it's hard to hear
425. 15:46a person with that kind of experience
426. 15:48say software is dead when I've been
427. 15:50working for a company that's building
428. 15:53big software that's completely AI
429. 15:56forward that that uses all the tools at
430. 15:59their disposal and still see like how
431. 16:02much time it takes to build software and
432. 16:03we're talking about years and there's
433. 16:06still more years down the pipeline for
434. 16:08this to get to the point where it
435. 16:10replaces that legacy piece of software.
436. 16:12This is just one application for one
437. 16:15smallish company. We're not contracted
438. 16:17to a Fortune 500 company. This is a
439. 16:19smaller pharmaceutical company. And it's
440. 16:22crazy to think when you hear people say
441. 16:25that software is dead and SAS is dying
442. 16:26and all this stuff when you see it
443. 16:28firsthand that there's no way. No way
444. 16:30that this is getting fully automated.
445. 16:32Maybe in the future it might happen, but
446. 16:35not now. I'm telling you right now,
447. 16:37there's no way that AI does it all.
448. 16:39There's too much that goes into software
449. 16:42for it to take care of everything.
450. 16:44That's just my little example, right?
451. 16:46Another note I have here is that AI
452. 16:48hasn't like eliminated the hard parts
453. 16:50and I kind of talked about that a little
454. 16:51bit, right? But like the output of code
455. 16:55has grown. The review is now like the
456. 16:58new bottleneck. I think we're far from
457. 17:01fully automating it. I I see the
458. 17:03feedback that I get from Code Rabbit. I
459. 17:05see the feedback that I get from C-Pilot
460. 17:08and it's helpful, right? But a lot of
461. 17:10times it flags stuff that that it
462. 17:13doesn't know what it's talking about.
463. 17:14It's incorrect. It doesn't understand
464. 17:16the proper implementation. It doesn't
465. 17:18understand some of the reasoning behind
466. 17:20certain things. It doesn't understand
467. 17:22the business needs. It's kind of just
468. 17:24looking over the code that you push and
469. 17:26it's got like a lot of context that it
470. 17:28needs to go through and sometimes it
471. 17:30catches some really good stuff and then
472. 17:31sometimes it's just like nitpicky
473. 17:35dumb stuff that doesn't make any sense
474. 17:38because it's just a bot at the end of
475. 17:41the day. And I think there was a Harvard
476. 17:43study um that I'm like quoting here, but
477. 17:47it said that after adopting AI agents,
478. 17:50pull requests opened rose by 35%. And
479. 17:53then the average time for review
480. 17:55submission to approval rose by 56%. It's
481. 17:58very true that code review is now
482. 18:00becoming the bottleneck. And I feel like
483. 18:01that's one of the harder areas to solve
484. 18:03because if you just let the the bots run
485. 18:06a muck, eventually you're going to have
486. 18:07a mess. And at what point are you not
487. 18:10just vibe coding enterprise software? It
488. 18:12just means more code that you need to
489. 18:14understand, that you need to test, and
490. 18:15you need to be able to maintain
491. 18:17eventually. I guess if you're just like,
492. 18:19"All right, let AI take the wheel," then
493. 18:21those are risks that you have to be
494. 18:23willing to take because when things
495. 18:25break, the people that you're building
496. 18:27software for, the companies that are
497. 18:29paying you to do this stuff, are going
498. 18:30to have questions. And if if you don't
499. 18:34know what's going on, you're not going
500. 18:35to have the answer and you're going to
501. 18:36have to start prompting your bot to
502. 18:37explain stuff to you. And the truth is,
503. 18:40even with the the code review tools that
504. 18:42we use and some of the findings in this,
505. 18:45you know, study or or uh article that I
506. 18:48found says that even with code review
507. 18:51tools, it's still a bottleneck, right?
508. 18:53Because you get the code review
509. 18:54feedback. You have to go through that.
510. 18:57You have to determine what makes sense
511. 18:58to fix, what doesn't. And there's
512. 19:01there's still there's still a lot of
513. 19:03slowdown when it comes to that, right?
514. 19:05And you can't just skip it because if
515. 19:07you skip it, then then you're going to
516. 19:10have to be willing to take that
517. 19:11responsibility. In a micro SAS, a a bug
518. 19:15that you might find is probably like a
519. 19:17small annoyance. It might be a little UI
520. 19:19thing or it might just be something not
521. 19:21critical to anybody's, you know,
522. 19:23day-to-day life, but a bug that can
523. 19:26you know, some serious piece of
524. 19:28software like pharmaceutical dispensing
525. 19:31and compounding software, like that's a
526. 19:34big deal. Like there's so much legal
527. 19:35stuff involved and there's so many
528. 19:37things that we have to make sure are
529. 19:38done right that we're not willing to
530. 19:42take that risk. I don't think software
531. 19:44is dead. If anything, I think software
532. 19:46is booming, right? I saw a meme recently
533. 19:48where it kind of showed uh you know
534. 19:50before AI and after AI I'll try to share
535. 19:52it right and it had like builders versus
536. 19:55users and before it was like a smaller
537. 19:57amount of builders and a slightly larger
538. 20:00amount of users in comparison to the
539. 20:02amount of builders and now in the age of
540. 20:04AI it's like you have all of these
541. 20:06builders and the number of users are
542. 20:08actually getting smaller because some of
543. 20:11them are in fact building their own
544. 20:12stuff. So, you've got a lot of boom
545. 20:16coming out of, you know, software and
546. 20:18there's a lot of slop and there's like a
547. 20:20lot of copycats and clones. I saw
548. 20:22something on Twitter with Beehive that's
549. 20:24like an email service and and another uh
550. 20:28tech influencer uh indie hacker Mark
551. 20:31Lou. I'm sure if you're familiar with
552. 20:32the scene, you know who he is, but
553. 20:34apparently he tried to build his own
554. 20:35beehive. We're getting a lot of like new
555. 20:38software and software that's replacing
556. 20:40tools that already do the job better
557. 20:42because people are like, "Oh, I'll go
558. 20:43ahead and build my own, you know,
559. 20:46beehive or I'll build my own service. I
560. 20:48don't need you anymore." And I think
561. 20:49that what's happening is that there's a
562. 20:52boom and there's actually more software,
563. 20:53but now we're in the age of software
564. 20:55slop. It's it's not new, right? We've
565. 20:57got slop for video. We've got slop for
566. 21:01uhformational products and slop for blog
567. 21:04websites and you know bot responses and
568. 21:07we've got like a lot of slop and I think
569. 21:09that there's a boom in software and
570. 21:11there will continue to be a boom in
571. 21:14software because if more people are able
572. 21:16to build it, smaller teams are able to
573. 21:18build faster than larger teams or as as
574. 21:20fast as a larger team would have been
575. 21:22able to build doesn't always mean that
576. 21:24that's going to be better, right? And
577. 21:25and there's just there's so much slop
578. 21:27out there. I don't know how recent this
579. 21:29is, but I remember seeing it where like
580. 21:30the app stores were having this huge
581. 21:32influx of apps and there was like a huge
582. 21:35like slowdown on them approving apps
583. 21:39because so many people are building apps
584. 21:41now. There's more and more slop. There's
585. 21:43more and more people that are building
586. 21:44stuff that aren't capable of building
587. 21:47but are because these tools make them
588. 21:50capable and then they hit a point where
589. 21:53they have built something and it's it's
590. 21:55a mess and then they're going to need to
591. 21:57hire developers and I think that that's
592. 21:58going to be a thing. In the past, I
593. 22:01remember you would hire contractors and
594. 22:04you'd get cheap contractors
595. 22:06and then they would build something
596. 22:08super fast and then hand it over to the
597. 22:10team and you would end up with this mess
598. 22:13that needed to be fixed and then they
599. 22:15needed to hire more devs. actually
600. 22:17happened to me in various jobs. And I
601. 22:19think that now we're in a boom where
602. 22:22there's going to be so much more of
603. 22:23that. And I honestly think that
604. 22:25eventually we're going to have an influx
605. 22:27of companies hiring experienced
606. 22:30developers to clean up the mess of all
607. 22:33of this software that's going to be
608. 22:34coming and all of this slop that's
609. 22:36happening. And for people to be saying
610. 22:39that software is dead, I I think it's
611. 22:42it's peaking. I mean literally a lot of
612. 22:44the stuff that I talked about here shows
613. 22:46you that right you can clone any repo
614. 22:49you can you can reverse engineer
615. 22:50anything people are building open-
616. 22:52source products of applications that
617. 22:56they don't want to pay for for companies
618. 22:58they don't like working with and some of
619. 23:00those things are great but there's a lot
620. 23:02of mess and there's a lot of stuff that
621. 23:04isn't great and it's it's far from dead
622. 23:07right I do believe that code is cheap
623. 23:10now but I don't think that good software
624. 23:13is. I think that software is growing.
625. 23:18And even if these companies are laying
626. 23:20people off, I I think at the end of the
627. 23:22day, they're just trying to make more
628. 23:24money. They're not going to slow down,
629. 23:25and they're not going to decelerate.
630. 23:27They're going to accelerate in their
631. 23:28building. They're going to take on more
632. 23:31AI. They're going to build more
633. 23:32software. Smaller teams are going to
634. 23:33keep building. Individuals are going to
635. 23:35keep building. People can build their
636. 23:37own stuff. If you can build your own
637. 23:38stuff, if it makes it so much easier for
638. 23:40solo founders and and small teams to to
639. 23:44build software, how is it dead? Please
640. 23:46tell me. I really want to know. How is
641. 23:49software dying? Because I'm tired of
642. 23:50seeing the clickbait titles. I don't
643. 23:53think that software is dead. I think
644. 23:55that's far from the case. And also when
645. 23:58these tools get more expensive, it's
646. 23:59going to be interesting to see what
647. 24:00happens with a lot of these people that
648. 24:03now um have more overhead than um have
649. 24:07to deal with, you know, the the bed they
650. 24:09made and they got to sleep in it. All
651. 24:11right, I think that pretty much sums up
652. 24:13what I wanted to say here. I hope that
653. 24:16you enjoyed this video. I hope that you
654. 24:19found it entertaining or relatable or
655. 24:22helpful. I don't know if it was helpful,
656. 24:23but nonetheless, thanks for watching and
657. 24:26I'll see you next