Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

They Don't Have the Money: OpenAI Edition

platformonomics.com

Post reply on HN
Print view View original

51–60 of 63 posts

Re: They Don't Have the Money: OpenAI Edition

#51

Post by danpalmer »

a_vanderbilt wrote: post #23
danpalmer wrote: post #7

Earlier quoted context omitted.

Have they? It seems to be getting clearer every month that they are not going to reach AGI, that the products are (good, but) not living up to the promises that were made. I suspect you're right that Tesla is in a different league here, but I don't think OpenAI are in a good spot.

The definition of AGI is diffuse enough to make it an argued point - until we can mostly agree it's already happened. For now, the stats are improving well enough across the industry to maintain investor attention. Will it all come crashing down a-la the .com bubble? It's seeming more likely by the quarter. Like the digital economy post .com burst, I think AI will survive and grow far beyond its current market of cha…

> The definition of AGI is diffuse enough to make it an argued point

This hits the nail on the head. 2-3 years ago when the current round of AGI hype started everyone came up with their own definitions of what it meant. Sam Altman et al made it clear that it meant people not needing to work anymore, and spun it in as positive a way as they could.

Now we're all realising that everyone has a different definition, and the Sam Altmans of the world are nit picking over exactly what they mean now so that they can claim success while not actually delivering what everyone expected. No one actually believes that AGI means beating humans on some specific maths olympiad, but that's what we'll likely get. At least this round.

LLMs will become normalised, everyone will see them for the 2x-3x improvement they are (once all externalities are accounted) for, rather than the 10x-100x we were promised, just like every round of disruption beforehand, and we'll wait another 10-20 years to get the next big AI leap.

Re: They Don't Have the Money: OpenAI Edition

#52

Post by thisisit »

pants2 wrote: post #2

Yes, their capex spending is mind-blowing. Astronomical. But OpenAI has also pretty much lived up to its promises so far. I don't see them straight up lying to investors the same way that Tesla does. Not only has OpenAI launched multiple viral products a year multiple years in a row, but their mission is to create God, so I think the TAM is pretty large.

If you go back to the starting years of Tesla you could make the same case. They weren’t straight up lying to investors.

Not only they launched viral products and Tesla Autopilot but their mission was to produce a full self driving car and capture a huge chunk of the auto market so their TAM was pretty large. You can read some early HN posts to see the amount of hype.

A casual observer was still pumped about Tesla lowering costs and delivering full automated driving in couple of years. One could say they were being overly optimistic. It is only when years passed by and tech didn’t materialise that now we say that they lied to investors.

I think same case can be made for OpenAI. They might hit a plateau on their advancement but continue to make overly optimistic projections.

Re: They Don't Have the Money: OpenAI Edition

#53

Post by Incipient »

kibwen wrote: post #46
typpilol wrote: post #28

Earlier quoted context omitted.

Their moat is 700m active users. What's facebooks moat? There's tons of social media sites. Facebooks moat is the 3b users. This comment is so idiotic it's starting to annoy me. "WHaTs tHe MoAt" for a company with almost 1b active users

Please read and understand https://en.wikipedia.org/wiki/Network_effect before commenting further.

But chatbot don't really have any network effect, compared to Facebook etc. There isn't any added value two people using ChatGPT together. The new sora network, will have said network effect.

Re: They Don't Have the Money: OpenAI Edition

#54

Post by pas »

pants2 wrote: post #37
typpilol wrote: post #28

Earlier quoted context omitted.

Their moat is 700m active users. What's facebooks moat? There's tons of social media sites. Facebooks moat is the 3b users. This comment is so idiotic it's starting to annoy me. "WHaTs tHe MoAt" for a company with almost 1b active users

I think it really is different though because there's no network effect required for ChatGPT. Unless you're talking about the training data they get from those users which probably is invaluable.

... is it?

Anyone can start making sponsorship deals and putting their AI into some service. And if that's really the secret sauce then AI firms will have to pay for it instead of people paying them to ask questions.

Re: They Don't Have the Money: OpenAI Edition

#55

Post by piva00 »

9cb14c1ec0 wrote: post #9

> I suspect we can design a very interesting new kind of financial instrument for finance and compute that the world has not yet figured it out Hmm, can't figure out why this statement makes me think of Enron. After all, OpenAI certainly isn't trying to do massive infrastructure build outs while struggling with a relatively limited cash flow, or anything like that.

I can't find the article right now because the keywords for searching just bring a lot of noise around OpenAI articles, pure slop most of the time, but Sam Altman said in an interview something around the lines "real innovation happens only through financial innovation" (or enabled by, can't remember the exact words), a statement that triggered a lot of red alarms in my head when I read it.

It's the same type of stuff Enron, the CDOs/CDSs from the 2008 crisis, and other financial frauds through history have thought. Let's repackage this unattractive financial product into layers of other stuff, hide the risks, rebrand it as something new and exciting, promising returns, and fuck everything up in the end.

“The four most dangerous words in investing are: ‘This time it’s different.’” - Sir John Templeton

Re: They Don't Have the Money: OpenAI Edition

#57

Post by pllbnk »

All these doomsday posts are just wishful thinking. On one hand, it would be bittersweet to see them fail. On the other, the cat's out of the bag and boring world's realities such as high energy consumption won't put it back. People see the usefulness it brings. Even if it doesn't translate to huge productivity gains just yet, I think even most critics understand that _quite likely_ it's just a matter of time, so the investors' money will be there as it's currently about the only thing in the popular technology space that attracts investment in the first place.

Re: They Don't Have the Money: OpenAI Edition

#59

Post by Yizahi »

teitoklien wrote: post #26
macintux wrote: post #10

Earlier quoted context omitted.

I’m on a long, meandering road trip and I decided to start using the conversational feature in ChatGPT. Truly amazing stuff. Discussed some sights I’d want to see along the Natchez Trace (although perhaps unsurprisingly it did not anticipate that this time of year was not ideal for hearing and seeing waterfalls), told me about Meriwether Lewis’s bizarre death. Basically like holding a conversation with Wikipedia, whi…

Same here, Youtube and Social Media algorithms driven by outrage to boost statistics, has made it unbearable for me to watch news or read it online, everything is framed and created with maximum outrage in mind. I added a DNS level blocker on all news app and restricted youtube itself to stop myself from watching news no matter what, and I only use chatgpt advanced voice mode now and sometimes perplexity pro to get m…

I read comments like these and feel like I'm taking crazy pills. Alice in Wonderland moment, truly. Like, we are all engineers here, even if just a little. Willingly and knowingly constraining your information intake to a random character generator is a bold move. Incomprehensible for me, but definitely bold :) .

Re: They Don't Have the Money: OpenAI Edition

#60

Post by Yizahi »

999900000999 wrote: post #45
ambicapter wrote: post #44

Earlier quoted context omitted.

As someone who uses ChatGPT around topics I'm relatively experienced about, it seems absolutely insane to me to put your entire worldview in its hands.

Chat GPT likes to make up things. It told me the other day that on a multiple hard drive/SSD system I could set secure boot on each drive independently. Of course this is nonsense, since secure boot is set in the BIOS. Whatever, Chat GPT got it's engagement metrics. I'm going to predict within the next few years someone's going to lose a billion dollars relying on Chat GPT or another LLM. It's still at the level of a…

> I'm going to predict within the next few years someone's going to lose a billion dollars relying on Chat GPT or another LLM.

I strongly suspect this has already happened, possibly even multiple times. Some inbred oil sheikh fatfingering some insane sum of money because "the computer told me, and computers are always correct" is completely believable scenario. Now that I think about it, doesn't Line city fits it perfectly? :)

What will happen in the next few years is probably not just a cash loss, but some large industrial accident with dead people, due to relying on the LLM bullshit. Now that would make headlines (at least until next Truth Social post).

Post reply on HN