4 ms·
The momentum is slowing on the basic research side, not as many outrageous improvements to benchmarks coming out these days. At the same time, a lot of non-triv
by svantana 8y ago
The momentum is slowing on the basic research side, not as many outrageous improvements to benchmarks coming out these days. At the same time, a lot of non-trivial tasks are close to being automated using ML methods, including high-paying tasks in medical, legal, security, etc. So if those revenue streams can be captured, that would account for these valuations and then some.
As for P/S ratios, a lot of tech startups (in various fields) have zero revenue, which makes the ratio infinite, so I'm not sure that says very much.
- new299 8y agoI’m curious to know which tasks specificly are close to being automated? Also, are there any AI applications which have already generated a large amount of revenue?
- snaky 8y agoThe tracking of everybody all the time is close to be automated.
- new299 8y agoWell... that’s great. But also doesn’t really require much in terms of advances in AI? Unless you’re saying this is coming from advances in face recognition?
- snaky 8y agoIt depends on how you define 'advances', https://news.ycombinator.com/item?id=18629063 https://news.ycombinator.com/item?id=18629063