Discussion
Loading...
Youse younglings seem to forget that for 20 years teh interwebs was not making any moni either...
The 'running out of cash' does not BTW mean all #AI will poof in a cloud of smoke. Its a stressor point only. #OpenAI has been successful at getting suckers^H^H^H^H^H^H^Hinvestors and the Inference slowly becomes amortized (cached, distilled, edge-run) and revenue may come from other streams (e.g. Militray).
...OOPS sorry I mis-spoke, let me join the cohort:
"WAAAH #AIBUBBLE STOCHIASTIC PARROT #AISlop " 😁 🤡 💀
Youse younglings seem to forget that for 20 years teh interwebs was not making any moni either
I guess that depends on what you mean by ‘the interwebs’. The ARPANet was a research project. The early Internet did not allow corporate or individual connections, it was limited to university and government connections. It did not make a profit then.
But before it allowed public and commercial connections, online connections were making money. CompuServe and AOL were both profitable and a bunch of companies were making money offering services over these in the early ‘90s. In France, the MiniTel network was doing online commerce and was a common way of buying and selling things. Again, people were making money.
When the Internet was open to the public, commercial ISPs started to appear (and CompuServe and AOL and a few smaller players like them started to offer bridges to the Internet). Telecoms companies were making money. As e-commerce started to take off, a lot of other companies started making money.
The dot com crash didn’t happen because the Internet was not profitable, it happened because a lot of people saw profitable companies on the Internet and said ‘I bet I can do that!’ And created or invested in a load of startups with no business plan.
Amazon was technically not profitable for a long time but that was because they were raising more funding and investing all of that and their profits in infrastructure to grow the company. They were shifting basically all of their operational revenue to capex. OpenAI is not doing this, they’re burning investor capital on opex.
So, two things: ‘the interwebs’ were profitable for a lot of the people in the ecosystem from very early. This is not the case for AI, where the only people making money are NVIDIA and a few companies in their supply chain. Everyone else is burning through capital at a ludicrous rate hoping that some magic will allow them to generate the trillions of dollars in revenue that they need to break even.
Second, ‘A was not profitable early on but then was later, B is not profitable early on and will make a load of money later’ is a logical fallacy. It ignores the thousands of technologies that were not profitable, never became profitable, and withered. Even in recent hype trains, look at NFTs and the Metaverse. These had billions of investment and the same kind of hype as ‘AI’, and are basically gone now. Things like the personal computer, Internet, and smartphones and incredibly rare. Silicon Valley is desperate to create the next one of these, but they don’t happen because of top-down push, they happen because the technology has reached a point where it can affordably solve a pre-existing set of problems.