4 ms·

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

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]

ai_ia 7y ago
> 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.
- NieDzejkob 7y ago
  I'm not sure if you're trying to imply the shifting of goalposts is bad, but it sure does sound like it. I would argue that all of research is like that - once something is achieved, people will want to do better things. After all, it's not an argument.
eternalny1 7y ago
> 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.
- throwaway-ai-ai 7y ago
  > 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"
mrmonkeyman 7y ago
In-house are the first to go. They are the expensive ones. Contractors go last.
saagarjha 7y ago
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.
- throwaway-ai-ai 7y ago
  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.
  - saagarjha 7y ago
    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?
  - th-th-throwaway 7y ago
    Have you seen this paper? https://openreview.net/pdf?id=ByldLrqlx https://openreview.net/pdf?id=ByldLrqlx You can see some of the simpler programs that can already be generated on page 12.
eru 7y ago
Well, we are constantly replacing programmers by computers. Think of eg compilers and interpreters. Thanks to them, perhaps you need only one guy to write software to solve a problem that used to take 10 people: 9 people replaced.
deepnotderp 7y ago
Program synthesis of large programs is an incredibly difficult problem. I could definitely be wrong but I don't think it's solvable with current AI techniques.

ai_ia 7y ago

> 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.

NieDzejkob 7y ago
I'm not sure if you're trying to imply the shifting of goalposts is bad, but it sure does sound like it. I would argue that all of research is like that - once something is achieved, people will want to do better things. After all, it's not an argument.

eternalny1 7y ago

> 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.

throwaway-ai-ai 7y ago
> 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"

mrmonkeyman 7y ago

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

saagarjha 7y ago

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.

throwaway-ai-ai 7y ago
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.
- saagarjha 7y ago
  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?
- th-th-throwaway 7y ago
  Have you seen this paper? https://openreview.net/pdf?id=ByldLrqlx https://openreview.net/pdf?id=ByldLrqlx You can see some of the simpler programs that can already be generated on page 12.

eru 7y ago

Well, we are constantly replacing programmers by computers. Think of eg compilers and interpreters. Thanks to them, perhaps you need only one guy to write software to solve a problem that used to take 10 people: 9 people replaced.

deepnotderp 7y ago

Program synthesis of large programs is an incredibly difficult problem. I could definitely be wrong but I don't think it's solvable with current AI techniques.