>> The dream of running my own company got me to learn programming. 20 years later I'm an employee at a company still dreaming of running my own company. But now I can realize that reality and dreams are not the same and that's ok. As in, I probably really don't want to run my own company. We'll see =P

To me having one own's company was just a means to the end: making enough money to live comfortably without the need to get a job ever again for the rest of my life. I too learned programming as a means to achieve that end but eventually realized that I don't need a company if I can short-circuit the path to money. By switching to the right domain - finance, where what I learn might be eventually put to use directly by investing capital into profitable trading strategies.

Back to OP's article, if there's a domain where money as a moat is not a problem, that's definitely finance: https://www.visualcapitalist.com/all-of-the-worlds-money-and...

I work in this domain since almost 20 years and can tell you, noone's gonna risk a billion dollars on crap vibe coded by AI. I wrote before, I don't know what crack these AI people are smoking but when there's real stakes at play, they don't play around with toys. And AI in programming is a toy. The unlikely triumph of "Can I haz teh codez?" CTRL+C / CTRL+V "prompt experts" (mocking it, lol) strategy on Stack Overflow, along with the people who employ it.

I'm not worried about MY particular future in this industry. I'm not worried that AI is gonna replace me, us, or write anything significant here at all in the foreseeable future until it fucking evolves into AGI which is somewhere 5000 years from now, optimistically.

The party's gotta come to an end really soon along with the figures on how much money AI makes versus it's real utility - which is, simply stated, "a toy".

Not necessary but here's my mood while writing this comment: - listening to this song: https://www.youtube.com/watch?v=6EWqTym2cQU&list=RD6EWqTym2c...

Also, as someone who's main source of income is a YouTube channel: there is a type of threshold effect, where your videos are not good enough to watch until one day they suddenly are.

This means that until you reach that threshold, it feels like you're not making progress, cause every video just gets the same result (no views). Even if below the surface, you're slowly inching closer to that moment where your videos will actually be watched.

To the extent that’s true it has much, much more to do with AWS, open source libraries, and collective knowledge, than it does LLMs.

But I honestly can’t think of anything you could do in a week that a company in 2015 would have paid billions for unless it’s something like tweaking an LLM. But in that case it’s the original model, not the 1 week or work you put in.

OK, say you build a Whatsapp clone in a week. How many Whatsapp users will switch to your app?

> things that FAANGs would have paid billions for 10 years ago

such as?

Creating marketing material has certainly gotten easier as well, it used to require a lot of work to create these spam pamphlets and company documents but today its trivial. Of course those are worthless to society so didn't help GDP but it filled our society with advertisements and spam and filled companies with worthless documents since now nobody thinks before making one.

That isn't what it's saying and I don't think the idea that "execution is what matters" is even true, other than to point out that ideas aren't valuable by themselves.

This is about marketing, about getting people to know and care that the thing you built exists. You can execute perfectly (in terms of making a great product) and not get a single eyeball.