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

Pause Giant AI Experiments: An Open Letter

futureoflife.org

Post reply on HN
Print view View original

661–670 of 1001 posts

Re: Pause Giant AI Experiments: An Open Letter

#661

Post by ntr-- »

lannisterstark wrote: post #405
reso wrote: post #380

To those saying "this is impossible in our society", there is a long history of scientific fields mutually agreeing to pause certain work until safety and regulations could be put in place. The Asilomar Conference is an example. [1] The idea that "you can't put the genie back in the bottle" is very popular in tech culture, but many have suggested that simply extending elements of copyright or tort law to AI would be…

It's not about it being impossible. I guess the major issue right now is that people are focusing on the wrong aspect of it. They should be focusing on how most LLMs are currently only controlled by bigger coroporations. Democratizing the LLMs and Giving access to people would resolve the x-poly we have currently. This also solves the "safety and regulations" issue to an extent - Do you really trust that Microsoft, G…

> you cant take all the money!

A cursory reading of the list of signatories shows mostly academics, so your summary of their motivations appears to be incorrect.

> not like that, not by democratizing it, but under OUR CONTROL

I also don't know how you came to that conclusion, from TFA:

> Such decisions must not be delegated to unelected tech leaders

Re: Pause Giant AI Experiments: An Open Letter

#662

Post by m_ke »

tome wrote: post #397
m_ke wrote: post #41

GPT-4 might be close to the best we'll get on the general LLM model front for a while since they trained on a huge chunk of web text. Next real advances will probably be in tuning them for specific applications in medicine, law, accounting, marketing, coding and etc. As someone running a one man company I can't wait for the cost of accounting, legal and copywriting to approach 0. Cost of shipping products will also g…

> Cost of shipping products will also go down 10-20x. How can a large language model achieve that?

Ask chatgpt to implement some of the things you worked on the last few months. I was very skeptical too until I tried this.

Here are some sample prompts that I tried and got full working code for:

- "write pytorch code to train a transformer model on common crawl data and an inference service using fastapi"

- "write react native code for a camera screen that can read barcodes and look them up using an API and then display info for matched results in a widget under the camera view"

- "write react code for a wedding website"

- "write code to deploy a django website on GCP using terraform and kubernetes"

- "how do I dockerize the app, it uses pytorch and faiss, also push it to a container registry"

- "implement a GPT style transformer model in pytorch", "write a training loop for it with distributed support and fp16"

- "how would you implement reinforcement learning with human feedback (RLHF)", "can you implement it in pytorch"

- "write code to compress a model trained in pytorch and export for inference on iOS"

- "how would you distill a large vision model to a small one"

- "what are the best CV architectures for mobile inference?"

For all of these it gave me code that was 95% usable, all in under 15 minutes, and which would have taken me a week or two to do on my own.

Re: Pause Giant AI Experiments: An Open Letter

#663

Post by drdeca »

int_19h wrote: post #608
biscottigelato wrote: post #476

Earlier quoted context omitted.

What we need is more funding into AI Alignment research. An X-Prize equivalent. All current big AI companies should fund it, to the tune of 20%+ of revenue from AI businesses. And I am okay to have government enforce this if needed even tho I'm mostly a libertarian. Government should only be there to remedy tragedy of the commons and nothing else. This is exactly the biggest tragedy of the commons.

How exactly do you envision the enforcement? Only sell GPUs to people with a "certified ethical AI researcher" license? Purge torch, transformers etc from PyPI and only provide controlled access to them? Remove all the books and papers on the building blocks from public libraries and online?

I think you are talking about enforcing different things? The thing to enforce, which the parent commenter described, seemed to be a 20% tax on AI profits of “big AI companies”, to be paid to AI safety research groups.

That doesn’t sound unenforceable to me. That’s just like, a tax on a kind of business.

Re: Pause Giant AI Experiments: An Open Letter

#664

Post by nwatn »

This would only work if every country stopped together. China will not stop, so this is not a sensible thing to do.

The US has the most to lose from not getting AGI first, because it is the world superpower. Any major disruption to Pax Americana would not be pretty.

Re: Pause Giant AI Experiments: An Open Letter

#665

Post by m_ke »

yreg wrote: post #446
m_ke wrote: post #41

GPT-4 might be close to the best we'll get on the general LLM model front for a while since they trained on a huge chunk of web text. Next real advances will probably be in tuning them for specific applications in medicine, law, accounting, marketing, coding and etc. As someone running a one man company I can't wait for the cost of accounting, legal and copywriting to approach 0. Cost of shipping products will also g…

>since they trained on a huge chunk of web text What did they trained it on? Why is it unprobable to train on a better/bigger dataset any time soon?

OpenAI is working with Microsoft so they definitely had access to the full Bing index and data from their other platforms like Github and Linkedin. They also paid for private datasets, from what I heard they might have gotten a copy of Quora and I'm sure they got a dump of all digitized books from someone.

Their best bet now is getting more supervised conversational data, which they should be getting a ton of from Bing and ChatGPT usage (they can use it as is with RLHF dataset which they had to pay people to generate by having fake conversations).

I wouldn't be surprised if they partner with Microsoft and hire a large team of doctors to tune it to handle specific medical conditions like diabetes.

Re: Pause Giant AI Experiments: An Open Letter

#666

Post by anileated »

reso wrote: post #380

To those saying "this is impossible in our society", there is a long history of scientific fields mutually agreeing to pause certain work until safety and regulations could be put in place. The Asilomar Conference is an example. [1] The idea that "you can't put the genie back in the bottle" is very popular in tech culture, but many have suggested that simply extending elements of copyright or tort law to AI would be…

Copyright law should recognize that training models is already a copyright violation. If we just declare “no using AI generated work” unscrupulous people will simply pass it for their own work.

Re: Pause Giant AI Experiments: An Open Letter

#667

Post by m_ke »

boringg wrote: post #333
m_ke wrote: post #41

GPT-4 might be close to the best we'll get on the general LLM model front for a while since they trained on a huge chunk of web text. Next real advances will probably be in tuning them for specific applications in medicine, law, accounting, marketing, coding and etc. As someone running a one man company I can't wait for the cost of accounting, legal and copywriting to approach 0. Cost of shipping products will also g…

I think if history has bearing on things I don't see the cost of accounting, legal or copywriting ever approaching 0. If anything you will see those paywalled behind a company who will extract that from you. It's wishful thinking that somehow that goes to 0.

ChatGPT is already better at copywriting than 90% of startup founders and marketing people at big cos. You'll soon be able to let it generate 1000s of different versions of marketing material to A/B test or personalize based on user info.

Soon you'll have multi modal transformers from dozens of companies and open source projects that will be able to parse and categorize all of your financial data and they'll have all of the incentives in the world to get it down to the cost of a quickbooks subscription.

Re: Pause Giant AI Experiments: An Open Letter

#668

Post by fasterik »

munificent wrote: post #617
fasterik wrote: post #587

Earlier quoted context omitted.

There seems to be a contradiction in the picture of the world you're drawing. Humans deeply need to be needed by others, yet when all our problems are solved we will sit around alone consuming media? If we were truly in that situation, why wouldn't our deep needs drive us to spend our days with friends and family, pursuing creative projects, and inventing new problems to solve?

For the same reason we: * Eat a bag of chips instead of cooking a meal. * Watch a travel show instead of leaving the house. * Binge watch a vlogger and become a superfan instead of nurturing a real symmetric friendship. * Read books about how to draw but never fill a sketchbook. * Buy a gym membership but never go. Because delaying gratification is hard and we're just poor apes doing the best we can.

Sounds like your issue isn't with automation, it's with the human condition. It reminds me of this passage from Schopenhauer:

If every desire were satisfied as soon as it arose how would men occupy their lives, how would they pass the time? Imagine this race transported to a Utopia where everything grows of its own accord and turkeys fly around ready-roasted, where lovers find one another without any delay and keep one another without any difficulty; in such a place some men would die of boredom or hang themselves, some would fight and kill one another, and thus they would create for themselves more suffering than nature inflicts on them as it is.

While I see some truth in this kind pessimism, I don't think it means that so-called utopia is necessarily worse than the world we live in today. I agree that human problems aren't going away, but I think we underestimate the potential value of removing the need to have a "job" to acquire the basic necessities for survival and the amount of suffering that causes today.

Re: Pause Giant AI Experiments: An Open Letter

#669

Post by endtime »

jongjong wrote: post #620
zamalek wrote: post #599

Earlier quoted context omitted.

The problem is that anyone with sufficient money could rent out the infrastructure required to train GPT5 or what-have-you. They need not share it publicly, they can use it for their nefarious purposes privately. The analogous scenario would be preventing your military from experimenting with TNT, when the recipe for TNT is public knowledge. You can't copyright away public knowledge.

Good point. As always, it's always the good guys who follow the rules and the bad guys who don't. This pause would just allow the bad guys to catch up. The solution is to allow as many different entities as possible to develop their own AIs so that they end up competing with each other and providing alternatives. What we don't want is a single AI which dominates our entire belief system.

Ah yes the old "give everyone in the world their own nuclear weapon" solution.

Re: Pause Giant AI Experiments: An Open Letter

#670

Post by CMay »

omnimus wrote: post #455
CMay wrote: post #454

The problem with this is that at present, it is largely good people who are ahead in AI development and that gap might widen over time if China has difficulty acquiring large numbers of high end chips. China and Russia feel that they need to cheat in order to compete in many ways. China does invest a lot into AI and some good work does come from there, but any advantage they can gain that their opponent will not take…

Ah the good people. Of course i trust those.

It's less about full trust and more about willingness to negotiate, act in good faith and take feedback from society if side effects surface. Ideally oriented towards empowering the people rather than the government.
Post reply on HN