Hacker News Reader

Comment by joenot443 | Hacker News Reader

I think Ed maybe buries the lede a little bit with all the ranting. About halfway through the piece he drops this chart[1] which personally left me shocked.

To my knowledge, a 10,000% growth in revenue over 7 years isn't a reality once you're at that level of volume. Asking 4o about this projection acknowledges the reality -

"OpenAI’s projection to grow from approximately $1 billion in revenue in 2023 to $125 billion by 2029 is extraordinarily ambitious, implying a compound annual growth rate (CAGR) exceeding 90%. Such rapid scaling is unprecedented, even among the fastest-growing tech companies."

Am I missing something? I like OAI and I use ChatGPT every day, but I remain unconvinced of those figures.

[1] https://lh7-rt.googleusercontent.com/docsz/AD_4nXcTvV_KScCMt...

View on HN

The fact that people are generally uncritical of this growth claim is just stunning. You have to believe that this is as big as smartphones while ignoring all of the differences with this cycle of hype.

Nobody intelligent is actually uncritical about these growth claims. What you are seeing is the other side of "the market can remain irrational longer than you can remain liquid". There's no financial benefit to calling bullshit, because timing when the market is suddenly going to correct is impossible. Rational investors are either in the market of boosting the stock to sell it in the short term (a bigger fools scam) or staying as far away from it as possible.

It doesn't have to hit the projections. It just has to keep looking like it could 2 months from now when you sell.

The market has never been determined by the actual value of the underlying tangible thing, but rather the market value of that tangible thing. Although they may seem similar they are not the same. As the market becomes increasingly detached from reality, you'll see more of this.

The vision they are selling is that AI will replace high value professionals in the economy. It's a vision that entails the collapse of the entire western societal model. If you believe in that investing in them becomes a existential matter, and the actual immediate value doesn't matter. You simply can't afford not to invest.

The numbers are meaningless. They are just meant to evoke "we will control the economy".

ChatGPT was released Nov 2022. Who would have predicted on launch day that they'd do $1B in 2023? Or $4B in 2024? Or $13B in 2025?

This chart actually shows a substantial slowing of the growth they've already achieved over the last 2.5 years. Why are you so convinced it'll slow even faster than that?

Because of the nature of growth when you can only acquire a user once. By definition, the first customers you get will be the most engaged (because they responded first). As you keep hitting people over and over again, conversion rates tend to decline.

Note that this is not always true in B2B, but even there the first few enterprises you sign will probably be better quality than the last few.

Because nowhere in recorded history, at least that I am aware of, has growth consistently followed tremendous gains found at optimistic starts.

If I have a successful small store, and next year I employ 2 helpers, that's 200% growth in size. If the year after that I open a new store, with 4 workers at each, and me just managing the whole thing, that's another year of 200% growth.

If I then predict that in ten years my company will be employing half a million people, I'll be laughed out of the bank.

But that's what you are expecting: linear growth FACTORS.

In the real world, there's actually a limit to how many small stores of my niche type are needed. When you hit the point where there's a Starbucks inside a Starbucks, you may have overreached...

Ed's style is very much to rant about this stuff lol. But it's really important to have critical voices right now since a miss with this stuff could have devastating consequences for the industry and the people who work in it.

Reply on news.ycombinator.com