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

The gap between open weights LLMs and closed source LLMs

blog.doubleword.ai

Post reply on HN
Print view View original

201–210 of 264 posts

Re: The gap between open weights LLMs and closed source LLMs

#201

Post by roenxi »

port11 wrote: post #190
taffydavid wrote: post #181

> Now is probably a good time to liquidate your pension, fly to a remote island somewhere, and live out the remaining 6 months or so of civilization in peace. > So maybe the open source apocalypse won’t happen yet. Sorry I wasn't at the last doomer meeting, when did we decide good open source models are a harbinger for the apocalypse?

Cute. Climate change’s apocalyptical impact on food crops and cancer rates (post-ozone collapse) never convinced people to enact change. But hey, it’s open-model LLMs, the boogeyman! Can’t have that, it must be OpenAI or Anthropic safely controlling the market and calling all the shots.

Are you sure you haven't gotten your catastrophes crossed? Ozone depletion was a different crisis and people did enact change, the ozone hole has been closing fairly steadily. Wikipedia [0] thinks the prospects for the ozone layer are pretty good.

[0] https://en.wikipedia.org/wiki/Ozone_depletion#Prospects_of_o...

Re: The gap between open weights LLMs and closed source LLMs

#202

Post by roenxi »

echoangle wrote: post #155
Zetaphor wrote: post #121

Earlier quoted context omitted.

Because that has worked so well for: * Drugs * Media piracy * Alcohol * Sex work * Unlicensed gambling The government is not an all powerful entity with absolute control over its people. Even in countries under past and present dictatorship there are examples of people getting access to what the government deemed as illegal.

I was thinking of this one: https://en.wikipedia.org/wiki/Executive_Order_6102 Of course you’ll always be able to get access but the risk can be made so high that most people won’t try it. There are countries that have death penalty on dealing with drugs and really severe prison terms just for having a small amount of drugs. There are still people that do it, but most people are effectively deterred because it’s just…

> but the risk can be made so high that most people won’t try it.

Possibly I've been mislead about how easy it is to access illegal drugs. I get the impression most people have actually tried them; or at least have easy access if they feel like it. Although I haven't bothered to look up the stats so maybe that is mistaken.

Hypothetically... say I rent out some server space in Russia, host GLM 5.2 and charge/pay for everything with crypto (one of the privacy coins). How exactly would the US government shut that little operation down? Or make it more risky for any participant than marijuana or torrenting? Even detecting it is an interesting technical challenge. It seems like it'd be low-skill and low-risk, and take insane resources on the part of the US to stamp out for something so harmless. The hydra would be growing heads faster than they could cut them off.

This isn't bars of gold, they can search my house all they like and there isn't going to be a lot in it. They would probably struggle to figure out who I am to do a targeted raid, let alone all the other small fry who could pull of a similar scheme.

Re: The gap between open weights LLMs and closed source LLMs

#203

Post by BryanBigs »

gspr wrote: post #192
felooboolooomba wrote: post #18

Earlier quoted context omitted.

> they can never be taken away Your right to 3d print whatever you want is about to be taken away (in California). What software you can run on your computer can already be restricted. Absolutely everything can be taken away. The simplest way to remove open models is probably to declare them a tool that terrorists could use. Crazy? Yes, the world is totally crazy these days.

Anything can be taken away, yes. But in a free and democratic society, there's an enormous difference between "the democratically chosen state powers may take something from me" and "a private entity takes away something from me on an inscrutable whim with no recourse". Neither is good if you don't want the thing taken away. But removing the second mechanism is still a laudable goal.

I think one is worse than two. Only governments have the power of violence.

Re: The gap between open weights LLMs and closed source LLMs

#204

Post by felooboolooomba »

woctordho wrote: post #160
echoangle wrote: post #82

Earlier quoted context omitted.

Believe me, if the government wants to stop you from having access to something like that, they could do it. Just give people some incentive to report you and make really harsh punishments and everyone will be thinking really hard about how bad they want have access.

Fun fact: Hacker News is canonically banned in China, but I'm still talking here. There are plenty of techs to work around region block. The incentive to report somebody is comically called '50w' (500k CNY) and no one gives a shit about it in real life.

They know. They just don't want a international incident to deal with. But if the shit hits the fan, they know.

Re: The gap between open weights LLMs and closed source LLMs

#205

Post by felooboolooomba »

BryanBigs wrote: post #203
gspr wrote: post #192

Earlier quoted context omitted.

Anything can be taken away, yes. But in a free and democratic society, there's an enormous difference between "the democratically chosen state powers may take something from me" and "a private entity takes away something from me on an inscrutable whim with no recourse". Neither is good if you don't want the thing taken away. But removing the second mechanism is still a laudable goal.

I think one is worse than two. Only governments have the power of violence.

Currently there's very much of a blur between "democratically chosen state" and "a private entity" (with a lot of money).

Re: The gap between open weights LLMs and closed source LLMs

#206

Post by gspr »

BryanBigs wrote: post #203
gspr wrote: post #192

Earlier quoted context omitted.

Anything can be taken away, yes. But in a free and democratic society, there's an enormous difference between "the democratically chosen state powers may take something from me" and "a private entity takes away something from me on an inscrutable whim with no recourse". Neither is good if you don't want the thing taken away. But removing the second mechanism is still a laudable goal.

I think one is worse than two. Only governments have the power of violence.

And we want them to. And to have the power to take things away. It's all a matter of degree and details.

Re: The gap between open weights LLMs and closed source LLMs

#207

Post by dboreham »

c0rruptbytes wrote: post #166
profsummergig wrote: post #5

IMHO, the biggest problem with the future of open weights models is that currently, open weights models are the result of philanthropy by some private org. (e.g. DeepSeek). The spigot can be turned off at any time. Until there's some sort of "community owned hardware", open weights models are always at risk of being discontinued.

Deepseek isn't philanthropy, it's a hedgefund trying to short the western AI market by saying "hey we can do 90% of they can (arguably better at a density metric) for a 1/10th of the cost" it's my theory at least, the Hindenburg Research of AI

Widely held belief in investor circles is that the Chinese government has a goal to deflate the US AI bubble and Deepseek is part of the plan to achieve that goal.

Re: The gap between open weights LLMs and closed source LLMs

#208

Post by taffydavid »

roenxi wrote: post #201
port11 wrote: post #190

Earlier quoted context omitted.

Cute. Climate change’s apocalyptical impact on food crops and cancer rates (post-ozone collapse) never convinced people to enact change. But hey, it’s open-model LLMs, the boogeyman! Can’t have that, it must be OpenAI or Anthropic safely controlling the market and calling all the shots.

Are you sure you haven't gotten your catastrophes crossed? Ozone depletion was a different crisis and people did enact change, the ozone hole has been closing fairly steadily. Wikipedia [0] thinks the prospects for the ozone layer are pretty good. [0] https://en.wikipedia.org/wiki/Ozone_depletion#Prospects_of_o...

If only climate change was as easy a solve

Re: The gap between open weights LLMs and closed source LLMs

#209

Post by _pdp_ »

mft_ wrote: post #195
_pdp_ wrote: post #175

Earlier quoted context omitted.

GLM 5.2 has replaced "normie" agentic workflows previously backed by Sonnet and Opus. So I don't know. From my end it seems to me they are perfectly capable of working agenticly.

Maybe we have different definitions of 'normie'. I'm talking about people who aren't in IT, and who are maybe just learning to use LLMs for aspects of their daily work. These people only know of the big three models, at best - they very rarely know of the open-weight models, and would even more rarely (given their model access is likely determined at a corporate level) be able to access them.

That's my point too.

If you take GLM and call it ChatGPT or Claude Opus is anyone going to notice? If you are not into agentic AI I would argue that the model type makes zero difference for day to day use because GLM 5.2 is hitting the benchmarks hard.

Now for a specialised use case (narrow fields), say cyber, Mythos is possibly better.

Re: The gap between open weights LLMs and closed source LLMs

#210

Post by Eridrus »

Onavo wrote: post #137
Eridrus wrote: post #136

Earlier quoted context omitted.

I think the bigger issue is the ever increasing capital requirements, which may cause even the closed weight companies to fall away from the frontier, e.g. Google & Meta are barely hanging on. For Google it feels a bit existential to remain at the frontier, but even then they're barely there. I hope that we find ways of continuing to improve these models besides continuing to exponentially increase capex spend until…

Google and Meta's failures are more due to mismanagement no?

I have no idea what's going wrong inside Google/Meta, they certainly have capital. But when you need this much cash, not many people are going to be able to have a shot on goal. It would not surprise me if Meta threw in the towel. Microsoft and Apple aren't even trying.
Post reply on HN