Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

IMO Grand Challenge

imo-grand-challenge.github.io

Post reply on HN
Print view View original

11–20 of 75 posts

Re: IMO Grand Challenge

#11

Post by ai_ia »

throwaway-ai-ai wrote: post #10

I'm curious: what are people's predictions for when it will be possible to run some program XYZ on a $10/hr EC2 machine and have it beat humans on the IOI/ICPC ? Furthermore, suppose such an open source program existed, how long would it take for it to start replacing remote contractors and then in-office programmers? ["start replacing" as in, say, 10% of human programmers]

> What are people's predictions for when it will be possible to run some program XYZ on a $10/hr EC2 machine and have it beat humans on the IOI/ICPC ?

A chess program can be trained on a EC2 instance that can possibly beat most grandmasters. My best guess, people will get over it and move on to more difficult challenges. The field of AI has a notorious record of shifting goalposts whenever something previously thought unattainable is achieved.

> how long would it take for it to start replacing remote contractors and then in-office programmers?

Although I don't think it's possible to replace "programmers", if such scenario does evolve, we would have come up with newer jobs that didn't existed before. Podcasting, Youtuber, UX Designers didn't exist 20 years back. Hard to guess what will come forth.

Re: IMO Grand Challenge

#12

Post by eternalny1 »

throwaway-ai-ai wrote: post #10

I'm curious: what are people's predictions for when it will be possible to run some program XYZ on a $10/hr EC2 machine and have it beat humans on the IOI/ICPC ? Furthermore, suppose such an open source program existed, how long would it take for it to start replacing remote contractors and then in-office programmers? ["start replacing" as in, say, 10% of human programmers]

> Furthermore, suppose such an open source program existed, how long would it take for it to start replacing remote contractors and then in-office programmers?

Until AI can create an Angular application with a .Net Core Azure back-end that meets ever-changing customer requirements ("can we remove the need for Bootstrap 4? Can we make the integration with Active Directory seamless?") on short notice, not very soon at all!

I'm throwing that out there because that's just one of my current projects, but a lot of programmers do things daily that AI simply isn't intended for.

Or college courses, for that matter. Although since you mentioned 10% you probably mean more in the area of research science, which may be a different story.

Re: IMO Grand Challenge

#13

Post by mrmonkeyman »

throwaway-ai-ai wrote: post #10

I'm curious: what are people's predictions for when it will be possible to run some program XYZ on a $10/hr EC2 machine and have it beat humans on the IOI/ICPC ? Furthermore, suppose such an open source program existed, how long would it take for it to start replacing remote contractors and then in-office programmers? ["start replacing" as in, say, 10% of human programmers]

In-house are the first to go. They are the expensive ones. Contractors go last.

Re: IMO Grand Challenge

#14

Post by saagarjha »

throwaway-ai-ai wrote: post #10

I'm curious: what are people's predictions for when it will be possible to run some program XYZ on a $10/hr EC2 machine and have it beat humans on the IOI/ICPC ? Furthermore, suppose such an open source program existed, how long would it take for it to start replacing remote contractors and then in-office programmers? ["start replacing" as in, say, 10% of human programmers]

I was nowhere near anywhere good enough to participate in an IMO officially, but judging by the types of problems they ask I'm not super optimistic. This might come back to bite me, but I'd predict that EC2 won't be around when this is solved.

Re: IMO Grand Challenge

#15

Post by saagarjha »

prvc wrote: post #6

I see no reason why the difference between "textbook" and "contest" problems would be anywhere near as significant to computers as it is to humans.

Depends on the textbook, but most ones used in schools are quite formulaic while IMO problems generally require significant insight to solve.

Re: IMO Grand Challenge

#16

Post by throwaway-ai-ai »

eternalny1 wrote: post #12
throwaway-ai-ai wrote: post #10

I'm curious: what are people's predictions for when it will be possible to run some program XYZ on a $10/hr EC2 machine and have it beat humans on the IOI/ICPC ? Furthermore, suppose such an open source program existed, how long would it take for it to start replacing remote contractors and then in-office programmers? ["start replacing" as in, say, 10% of human programmers]

> Furthermore, suppose such an open source program existed, how long would it take for it to start replacing remote contractors and then in-office programmers? Until AI can create an Angular application with a .Net Core Azure back-end that meets ever-changing customer requirements ("can we remove the need for Bootstrap 4? Can we make the integration with Active Directory seamless?") on short notice, not very soon at…

> Until AI can create an Angular application with a .Net Core Azure back-end that meets ever-changing customer requirements ("can we remove the need for Bootstrap 4? Can we make the integration with Active Directory seamless?") on short notice, not very soon at all!

I think you are arguing for "Not all programming tasks that can be automated away by a bot that can win IOI / ICPC."

However, what I'm asking is: "Does there exist 10% of programming that can be automated away by a bot that can win IOI / ICPC" ?

Furthermore, I think you are also assuming that human programmers remain static, while the AI has to be a 'drop in replacement for humans.' On the other hand, if we look at AWS -- AWS didn't build some AI that is a drop in replacement for sys admin work. AWS built their own API, humans adapted to it.

It seems to me that in a world where such an IOI/ICPC winning bot existed, programmers would adapt to it (just as programmers have adapted to AWS), by figuring out "how can I leverage this API so that I don't have to hire someone to do FOOBAR"

Re: IMO Grand Challenge

#17

Post by saagarjha »

OscarCunningham wrote: post #4

A lot of IMO geometry problems are easy for computers because they can simply use coordinates to convert the problem to algebra (this approach is computation heavy which is why humans don't use it even though it's guaranteed to work given enough time). So in a good year the computer might get two out of the six problems without having to do anything clever.

IMO problems are generally proof-based and don't always lend themselves well to brute force solutions using analytic geometry.

Re: IMO Grand Challenge

#18

Post by lordnacho »

Why would this be hard for the machines? A lot of math contest problems are about trying some rearrangements or substitutions, using a few principles to guide you towards the solution.

If you give the machine some equations, wouldn't it find a path to the solution pretty fast? Aren't there solvers that aren't considered AI that do that sort of thing?

I did some math contests in school, and a lot of the problems needed some brute force element along with a bit of overview so you didn't waste too much time. Often there'd be a problem that you'd look at and think "hmm this will solve via integration by parts" but the thing was complex enough that you'd potentially screw something up.

I looked in the Gitlab, it seems the problems are already encoded in Lean. Haven't you solved a major part of the problem already if you can formally encode it like that? Half the fun of math contests is getting over the WTF feeling.

Re: IMO Grand Challenge

#19

Post by throwaway-ai-ai »

saagarjha wrote: post #14
throwaway-ai-ai wrote: post #10

I'm curious: what are people's predictions for when it will be possible to run some program XYZ on a $10/hr EC2 machine and have it beat humans on the IOI/ICPC ? Furthermore, suppose such an open source program existed, how long would it take for it to start replacing remote contractors and then in-office programmers? ["start replacing" as in, say, 10% of human programmers]

I was nowhere near anywhere good enough to participate in an IMO officially, but judging by the types of problems they ask I'm not super optimistic. This might come back to bite me, but I'd predict that EC2 won't be around when this is solved.

I would guess that a computer beats humans on the IMO before a computer beats humans on the IOI/ICPC -- but also think that the latter would have far greater impact.

Re: IMO Grand Challenge

#20

Post by saagarjha »

throwaway-ai-ai wrote: post #19
saagarjha wrote: post #14

Earlier quoted context omitted.

I was nowhere near anywhere good enough to participate in an IMO officially, but judging by the types of problems they ask I'm not super optimistic. This might come back to bite me, but I'd predict that EC2 won't be around when this is solved.

I would guess that a computer beats humans on the IMO before a computer beats humans on the IOI/ICPC -- but also think that the latter would have far greater impact.

I'm not sure I'd agree; they both seem like they'd be of similar difficulty to me. Why do you think that IMO would be easier for computers?
Post reply on HN