Investors Have Sky-High Hopes for AI. Can the Tech Deliver?

by virtualwhyson 2/6/24, 12:07 AMwith 3 comments
by moose_manon 2/6/24, 1:08 AM

I doubt it in the short term (like the next five years). All the use cases with low enough stakes for it to be adopted quickly it is too expensive to be practical, all the use cases where it is critical enough to warrant the costs, it is too dangerous / unreliable. The technology holds revolutionary promise but the costs for most businesses to implement it will be higher than is practical or efficient. We're not there yet. There are also fundamental issues like copyright, hallucinations, the ability to move toward generalizations, LLM poisoning that need to be solved. People talk about having automated fast food order systems but is it really cheaper than someone being paid minimum wage? The costs need to be scalable to be practical. There are venture back AI startups that are throwing in the towel over costs, what do you think most public companies that operate on quarterly earnings will do? There are some cool use cases and proof of concept products but to be as truly revolutionary in the next five years as the AI hype machine makes it out to be, costs will need to come down dramatically like tomorrow and it will have to solve all the hard open questions in the next year.

by bearjawson 2/6/24, 12:14 AM

VC's have sky high hopes of AI.

Why? Because they have absolutely lost it in the Crypto / Web3 / NFT / Micro mobility / Office as a Service sectors.

If they want their 100x growth, they are pretty much all in on AI, because they are down 50%+ on everything else.