Hacker News Reader

Comment by Ar-Curunir | Hacker News Reader

Thankfully research is not guided by notions like "what makes money in the current economy"; otherwise we'd be stuck with faster calculators instead of, well, all of modern computer science.

View on HN

  > Thankfully research is not guided by notions like "what makes money in the current economy"

Unfortunately I don't believe this is true for AI research... I think you'll find a strong correlation with each year's most cited papers and what's currently popular in industry. There's always exceptions, but we moved to a world where we're highly benchmark oriented, and that means we're highly reliant on having large compute infrastructures, and that means that the access and/or funding comes from industry. Who is obviously going to pressure research directions towards certain things.

I did not say anything about making money in the current economy. Research requires funding and only way to have academic freedom is to have a steady source of recurrent income or funding. One way to get that is to please those who have money but give up part of your control on what you study. Other way is to monetize your knowledge and acquire a ton of money. You seem to misunderstand what I said or have very little clue on how the academia works.

>Thankfully research is not guided by notions like "what makes money in the current economy"

That's right, it's guided by "what gets grant money in the current paradigm"

Most research I know about in academia is absolutely about what makes money.

Care to give some examples? It is largely untrue in CS academia, and definitely untrue in every other non-engineering field.

Reply on news.ycombinator.com