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

AI's top startups are barely publishing their research

science.org

Post reply on HN
Print view View original

261–270 of 341 posts

Re: AI's top startups are barely publishing their research

#261

Post by fluoridation »

1718627440 wrote: post #254
fluoridation wrote: post #151

Earlier quoted context omitted.

What's your point?

That the point he is making does not apply to open source, but rather to freeware (which is open source too). Open source can be completely B2B, without releasing it to third parties. Also open source is not free software, so you can also forbid your costumers from sharing the source further. EDIT: That's crap. I only thought about the aspect of continuing sharing, but if you also don't want your direct customers to…

The point is "most companies don't publish research and most software is proprietary [because that entails giving up the competitive edge]". I don't see anything in there about price. Also, I think most people would say that if you have to pay an exorbitant license fee to see the source code then the software is still proprietary. I don't see how it's not true that most companies keep their software proprietary to retain a competitive advantage.

Re: AI's top startups are barely publishing their research

#262

Post by nlitened »

witx wrote: post #251
nlitened wrote: post #243

Earlier quoted context omitted.

> The least these companies can do is to publish everything they own Well, being able to talk to superintelligent all-knowing savant 24/7 on your phone for $20 a month is also something

You people really have rotten morals

Well if it's moral for gatekeepers of talent to hoard immense intellectual rights to be protected by the violence of the state, then I'm immoral in that sense. Information wants to be free

Re: AI's top startups are barely publishing their research

#264

Post by SturgeonsLaw »

sillysaurusx wrote: post #115
ninjahawk1 wrote: post #90

I can’t speak for other startups, but I applied to the most recent YC batch with my idea for making AI proactive instead of reactive, and pre-being selected I’ve published a paper on recursive self-improvement mapped to the Epoch AI data. I contacted a professor from a university in the UK and he responded since he was working on similar work, then asked me if I wanted to meet with him. We talked for about an hour si…

That's also a reason the big labs stopped. Publishing is most valuable to people who have no other way to get the attention of smart strangers. Once you can hire nearly anyone and everyone already returns your calls, the main remaining effect of publishing is to tell your competitors which things worked. This is what happens to every field as it turns from a science into an industry. Chemists published freely until d…

The profit motive is the reason that most of the things that are shit, are shit.

Re: AI's top startups are barely publishing their research

#265

Post by adastra22 »

sixtyj wrote: post #246
teleforce wrote: post #123

Earlier quoted context omitted.

Which is a very ironic and selfish situation when your business model dependent mostly on model training based on available published data, academic and non-academic.

I had some meeting with non-creative business person, just chatting over coffee. I wondered how much he used from our meeting for his enrichment. Public papers with described research give advance to similar people. They jump over you and in many cases they give nothing back. (Intellectual) greed is everywhere and is cross-border. In case of public papers there is one extra vulnerability - competitor(s) can build any…

Academics are just as quick to steal ideas. It cuts both ways.

Re: AI's top startups are barely publishing their research

#266

Post by _0ffh »

antonvs wrote: post #63
HDBaseT wrote: post #60

Earlier quoted context omitted.

When the second biggest player is named 'OpenAI', I would expect some papers to be published.

OpenAI and Anthropic are identified in the article and study as top publishers of AI research papers.

I don't see that, where do you read that?

Also, what kind of papers do they publish at all?

When I look at the recently (say, last 1-2 years) published practical improvements in the LM space, they're not even close to what the Chinese are publishing.

Re: AI's top startups are barely publishing their research

#267

Post by jurgenaut23 »

sillysaurusx wrote: post #115
ninjahawk1 wrote: post #90

I can’t speak for other startups, but I applied to the most recent YC batch with my idea for making AI proactive instead of reactive, and pre-being selected I’ve published a paper on recursive self-improvement mapped to the Epoch AI data. I contacted a professor from a university in the UK and he responded since he was working on similar work, then asked me if I wanted to meet with him. We talked for about an hour si…

That's also a reason the big labs stopped. Publishing is most valuable to people who have no other way to get the attention of smart strangers. Once you can hire nearly anyone and everyone already returns your calls, the main remaining effect of publishing is to tell your competitors which things worked. This is what happens to every field as it turns from a science into an industry. Chemists published freely until d…

Such a biased and utilitarian view. The goal of publishing is to communicate your findings to the scientific community for the sake of advancing our knowledge as a society.

Of course you might want to keep _some_ of this knowledge as trade secret, but then don’t claim you are doing research.

You are advancing your shareholder’s interests, not that of the broader society.

Re: AI's top startups are barely publishing their research

#268

Post by petesergeant »

WithinReason wrote: post #256
petesergeant wrote: post #249

Earlier quoted context omitted.

Glory and altruism

People only care about those because it brings them more resources. People seek glory for power. Altruism is not a stable strategy so will be exploited.

Trivially disproved by self-sacrifice in the persual of both

Re: AI's top startups are barely publishing their research

#269

Post by fakedang »

sillysaurusx wrote: post #115
ninjahawk1 wrote: post #90

I can’t speak for other startups, but I applied to the most recent YC batch with my idea for making AI proactive instead of reactive, and pre-being selected I’ve published a paper on recursive self-improvement mapped to the Epoch AI data. I contacted a professor from a university in the UK and he responded since he was working on similar work, then asked me if I wanted to meet with him. We talked for about an hour si…

That's also a reason the big labs stopped. Publishing is most valuable to people who have no other way to get the attention of smart strangers. Once you can hire nearly anyone and everyone already returns your calls, the main remaining effect of publishing is to tell your competitors which things worked. This is what happens to every field as it turns from a science into an industry. Chemists published freely until d…

On the contrary, I've found that stuff in the mining industry is actually well published, to the point of replicability (one of our projects is essentially copying the patented design of a well known mining technology company, albeit with some modifications, as the patent will expire next month).

Perhaps because the real moat in the industry isn't the technology per se, but the concessions.

Re: AI's top startups are barely publishing their research

#270

Post by MassiveOwl »

ninjahawk1 wrote: post #90

I can’t speak for other startups, but I applied to the most recent YC batch with my idea for making AI proactive instead of reactive, and pre-being selected I’ve published a paper on recursive self-improvement mapped to the Epoch AI data. I contacted a professor from a university in the UK and he responded since he was working on similar work, then asked me if I wanted to meet with him. We talked for about an hour si…

What exactly makes it proactive? I created a similar system that ran on a ticker, but then could also set itself to run in n seconds in the future. RSI came from the actions available to the system and granting the ability to modify itself. The prompt was a string of messages made from static and dynamically generated sections (like memories or plans of tasks or outputs from actions taken on the previous turn). It worked really well
Post reply on HN