[venantius]

It is possible for all of the following to be true: 1. This study is accurate 2. We are early in a major technological shift 3. Companies have allocated massive amounts of capital to this shift that may not represent a good investment 4. Assuming that the above three will remain true going forward is a bad idea

The .com boom and bust is an apt reference point. The technological shift WAS real, and the value to be delivered ultimately WAS delivered…but not in 1999/2000.

It may be we see a massive crash in valuations but AI still ends up the dominant driver of software value over the next 5-10 years.

[hangonhn]

That's a repeating pattern with technologies. Most of the early investments don't pay off and the transformation does happen but also quite a bit later than people predicted. This was true of the steam engine, the telegraph, electricity, and the railroad. It actually tends to be the later stage investors who reap most of the reward because by then the lessons have been learned and solutions developed.

[asadotzler]

The dot com boom gave us $1T in physical broadband, fiber, and cellular networking that's added many many trillions to the economy since. What's LLM-based AI gonna leave us when its bubble pops? Will that AI infrastructure be outliving its creators and generating trillions for the economy when all the AI companies collapse and are sold off for parts and scrap?

[n_ary]

It is hard to tell at this point about the AI/LLM. But the powerful hardware backing it has numerous potential in other research and innovations which are not yet clear to us. Just like the fiber and cellular network made Meta or Airbnb obscenely rich, the hardware may as well facilitate other forms of tech, may be quantum computing or heck even blockchain may be back(surprisingly these are suddenly very legal and less regulated this year!) or who knows what might be new.

While LLM trend is already going to the dumpster(notice how we no longer receive 1b/4b/8b models from Meta since LLama4 and similar competitors but only from china?), I firmly believe that commodity hardware will improve in this race to allow running LLMs(or their next iteration) on regular devices and become as ubiquitous as Siri/Cortana.

[venantius]

Among other things the big tech companies are literally planning to build nuclear power plants off this so I think the infrastructure investments will likely be pretty good.

[n_ary]

True. But like the dot come bust, ultimate winner may not be LLM or even AI but something else entirely using the left over hardware and infrastructure which was intended for AI.

[Sammi]

People overestimate what can be done in the short term and underestimate what can be done in the long term.