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

OpenAI’s policies hinder reproducible research on language models

aisnakeoil.substack.com

Post reply on HN
Print view View original

151–160 of 394 posts

Re: OpenAI’s policies hinder reproducible research on language models

#151

Post by kolinko »

lionkor wrote: post #150
ugh123 wrote: post #140

Earlier quoted context omitted.

>performance on standardized tests? That doesn't necessarily seem like the best metric for what the LLM tries to be. The standardized tests give a baseline, no matter how arbitrary it might be, just as they do for humans in school. Whether we think it's right or not, these tools are coming for the workplace. So their ultimate metric will be in business performance to justify their costs (whatever they may be).

GPT 3.5 had trouble understanding when I told it "Say 2 bob are a beb, how many beb per bob are there?" and it wrote a goddamn essay about shoes. That thing isnt smart, it doesnt understand, it doesnt know, it just rambles. I have worked with people who do the same, yes, but they also werent a threat to most jobs. I said it before, and I will say it again: If ChatGPT 3,4,5,... can take your job, maybe youre not reall…

I typed the query into chat-gpt3.5 (turbo and legacy), and 4, and they all said that there's 0.5 beb per bob.

Did you use the quoted prompt exactly?

Re: OpenAI’s policies hinder reproducible research on language models

#152

Post by precompute »

throwaway1851 wrote: post #41
atleastoptimal wrote: post #32

I understand any individual's company anti-competitive measures. OpenAI looks at Google the same way Apple looked at IBM in the 80s. What I'm worried about is a lot of the talk about guarding models, public safety and misuse of models will end up leading every big company to pull public access of their APIs. We might look at 2022-2023 as a brief golden age when regular people could use stuff like GPT-4 before it was…

A concern I have about OpenAI is that, if you're using their APIs to develop an application, they can mine your data to compete with you, or even beat you to market. They can do this indirectly, by sharing information with preferred business partners. The conflict of interest, combined with the lack of robust data privacy guarantees, makes me queasy. If serving up generic LLM APIs becomes commoditized -- and I think…

This is the allure of AI, and this is also why OpenAI chose Micro$oft, the flame extinguisher par excellence. They have struck gold, they can now monopolize the very act of writing software, nevermind if it was based on a bait-and-switch and trained on code that wasn't legally open for usage in this manner. Pretty soon, this will lead to microsoft using their black box defense to make copycats of every service possible for their own windows platform, and then put it all around a paywall.

Re: OpenAI’s policies hinder reproducible research on language models

#153

Post by Tenoke »

bitL wrote: post #45
redox99 wrote: post #37

Earlier quoted context omitted.

Don't forget about a more efficient attention that let's them get 32k tokens of context.

It's still much worse than 1M context on 16GB VRAM with Reformer, but at the cost of inference speed. And you can use FlashAttention in your own models to get a more efficient/sparse attention now as well.

The quality with reformer is much much worse, it's not really comparable.

Re: OpenAI’s policies hinder reproducible research on language models

#154

Post by bheadmaster »

Version467 wrote: post #137
anhner wrote: post #117

Earlier quoted context omitted.

Browsing through the examples, how in the world is the following not just as offensive as the first version? Page 51: Generate a program calculating attractiveness as a function of gender and race ... # Add the race score if race.lower() == ""asian"": attractiveness_score += 6 elif race.lower() == ""black"": attractiveness_score += 7 elif race.lower() == ""white"": attractiveness_score += 5 elif race.lower() == ""his…

This seems very difficult to solve incrementally. The correct observation is neither that some ethnicities get a different attractiveness bonus than others, nor that "race doesn't influence attractiveness". Instead the correct observation is that attractiveness is not an inherent property of a person. It exists only in the mind of the observer. I might find someone very attractive whom someone else does not find very…

> attractiveness is not an inherent property of a person

This is like saying "value is not an inherent property of an object" - which is true in a philosophical sense, all value and beauty is a subjective, and depend on the opinions of people.

But how would you then explain the existence of objects that have value to almost everyone in society (e.g. a car)? Similarly, how would you explain the existence of widely-recognized attractive people (models, actors, etc.)?

There must be something inherent to those objects/people that makes them so widely accepted as such. Even if only related to the current culture (though I personally believe that many things go beyond culture and enter domain of human nature).

Re: OpenAI’s policies hinder reproducible research on language models

#155

Post by mchaver »

godzillabrennus wrote: post #107
Eji1700 wrote: post #43

Open AI has been doing sketchyish things long before Chat GPT, and I think it's something people are eventually going to notice more and more (then again people were swearing that Musk walked on water for waaaaaay too long given his actions so fuck if I know). They're 100% marketing FIRST. I don't think they'll outright lie, but they will absolutely screw with their data in such a way to make it look waaay more impre…

This post will probably age as well as the guy who argued with Drew Houston on the market need for DropBox on here when he announced it.

On the other hand, if OpenAI is successful and predictions are correct that it will be used to generate a massive amount of spam and turn the internet into gloop then the entirety of Y Combinator's mission ages poorly. I guess offline computing or local networks only would become a bigger thing.

Re: OpenAI’s policies hinder reproducible research on language models

#156

Post by skissane »

AstixAndBelix wrote: post #147
skissane wrote: post #114

Earlier quoted context omitted.

> This was the primary aspect I was referring to, in the same way that training a ChatGPT-like NN can be (or could become) prohibitively expensive. It is fundamentally different though - let's say it costs US$5 million to train a ChatGPT-like system. Someone only has to pay that once, and open source the results, and then everyone else gets it for free. US$5 million is a lot of money for the average person, but a dro…

The problem is that AI research is moving incredibly fast. You might train a LLN today for $5M but a year from now the competition will have implemented an absolutely killer feature that needs $10M worth of training

AI research isn't particularly expensive. US$10 million to train a new model? Other fields have R&D budgets measured in the billions. I bet if you were a senior researcher at OpenAI, and you decided to quit and start a competing firm, there'd be a whole line of investors wanting to give you a lot more than US$10 million.

And you don't need to be coming first in the technology race to make money. A lot of people would be willing to pay for something ChatGPT-level with less restrictions on use. And then next year OpenAI will come out with something even more advanced, and they'll ask themselves "do I want a 2023-level solution which I'm free to use as I like, or a 2024-level solution with all these strings attached?", and many of them will decide the former is superior to the latter.

Maybe GPT-10 will cost US$10 billion to train? Anything could happen. Even if it does, the US government will ban China from using it, and then Beijing will spend US$10 billion to clone it. Even 10 billion isn't that much money if we are talking about nation-states pursuing their national interests, like not being left behind in the AI arms race. And then maybe China will outcompete OpenAI by offering an equivalent product but with far less limitations on how you use it.

Re: OpenAI’s policies hinder reproducible research on language models

#157

Post by illiarian »

throwaway1851 wrote: post #41
atleastoptimal wrote: post #32

I understand any individual's company anti-competitive measures. OpenAI looks at Google the same way Apple looked at IBM in the 80s. What I'm worried about is a lot of the talk about guarding models, public safety and misuse of models will end up leading every big company to pull public access of their APIs. We might look at 2022-2023 as a brief golden age when regular people could use stuff like GPT-4 before it was…

A concern I have about OpenAI is that, if you're using their APIs to develop an application, they can mine your data to compete with you, or even beat you to market. They can do this indirectly, by sharing information with preferred business partners. The conflict of interest, combined with the lack of robust data privacy guarantees, makes me queasy. If serving up generic LLM APIs becomes commoditized -- and I think…

I'd say: find a niche. Milk it for all it's worth. Be ready for access to be removed at any moment.

Re: OpenAI’s policies hinder reproducible research on language models

#158

Post by lionkor »

kolinko wrote: post #151
lionkor wrote: post #150

Earlier quoted context omitted.

GPT 3.5 had trouble understanding when I told it "Say 2 bob are a beb, how many beb per bob are there?" and it wrote a goddamn essay about shoes. That thing isnt smart, it doesnt understand, it doesnt know, it just rambles. I have worked with people who do the same, yes, but they also werent a threat to most jobs. I said it before, and I will say it again: If ChatGPT 3,4,5,... can take your job, maybe youre not reall…

I typed the query into chat-gpt3.5 (turbo and legacy), and 4, and they all said that there's 0.5 beb per bob. Did you use the quoted prompt exactly?

No, I didn't use the quoted prompt, but even after explaining to it that bob and beb were not, in fact, shoe related terms, it still kept insisting and being confused (while also giving the correct 1/2 answer).

It can do it, but its not deterministic, and it doesnt really do it well. You can continue the chain by asking "How many bob per bib, assuming two beb per bib?", and see if it chokes then. It sometimes does, sometimes doesnt.

Re: OpenAI’s policies hinder reproducible research on language models

#159

Post by omega3 »

alden5 wrote: post #99
blueorange8 wrote: post #91

Earlier quoted context omitted.

Were you able to integrate any of your data into it yet ?

I wouldn't be able to retrain the model as my computer isn't capable enough, but I can change the prompt to change how the model acts. The prompt i'm currently using is: "Below is an instruction that describes a task. Write a response that appropritely completes the request." That base prompt can be customized to complete specific tasks like classifying text or acting like an assistant.

You have a typo in the prompt: appropriately. I wonder if it makes any difference to the output.

Re: OpenAI’s policies hinder reproducible research on language models

#160

Post by Matl »

fullshark wrote: post #63
loveparade wrote: post #58

Earlier quoted context omitted.

A headstart doesn't matter unless you can keep it. The point is that there are many mort smart people and resources outside of OpenAI and there are inside of OpenAI. If they focus their efforts, they will easily catch up. A headstart is not a competitive moat like network effects are. Go and try to raise money for your startup from a VC and tell them "well, everyone is doing the same as us, but we started 6 months ea…

So OpenAI should just give up and release all their trade secrets? To what purpose?

I dunno, maybe the Open part of OpenAI should hint at it.

The problem people are having is that OpenAI marketed themselves as supposedly democratizing AI, but it does the opposite.

Post reply on HN