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

DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via RL

arxiv.org

Post reply on HN
Print view View original

351–360 of 1001 posts

Re: DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via RL

#351

Post by LunaSea »

spaceman_2020 wrote: post #236
sdesol wrote: post #112

Earlier quoted context omitted.

Agreed. I am no fan of the CCP but I have no issue with using DeepSeek since I only need to use it for coding which it does quite well. I still believe Sonnet is better. DeepSeek also struggles when the context window gets big. This might be hardware though. Having said that, DeepSeek is 10 times cheaper than Sonnet and better than GPT-4o for my use cases. Models are a commodity product and it is easy enough to add a…

Curious why you have to qualify this with a “no fan of the CCP” prefix. From the outset, this is just a private organization and its links to CCP aren’t any different than, say, Foxconn’s or DJI’s or any of the countless Chinese manufacturers and businesses You don’t invoke “I’m no fan of the CCP” before opening TikTok or buying a DJI drone or a BYD car. Then why this, because I’ve seen the same line repeated everywh…

Any Chinese company above 500 employees requires a CCP representative on the board.

Re: DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via RL

#352

Post by brookst »

slt2021 wrote: post #323
brookst wrote: post #237

Earlier quoted context omitted.

Great as long as you’re not interested in Tiananmen Square or the Uighurs.

try asking US models about the influence of Israeli diaspora on funding genocide in Gaza then come back

Which American models? Are you suggesting the US government exercises control over US LLM models the way the CCP controls DeepSeek outputs?

Re: DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via RL

#353

Post by brookst »

thomas34298 wrote: post #316
brookst wrote: post #237

Earlier quoted context omitted.

Great as long as you’re not interested in Tiananmen Square or the Uighurs.

Have you even tried it out locally and asked about those things?

https://sherwood.news/tech/a-free-powerful-chinese-ai-model-...

Re: DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via RL

#354

Post by thrance »

mrtksn wrote: post #182
HarHarVeryFunny wrote: post #152

DeepSeek-R1 has apparently caused quite a shock wave in SV ... https://venturebeat.com/ai/why-everyone-in-ai-is-freaking-ou...

Correct me if I'm wrong but if Chinese can produce the same quality at %99 discount, then the supposed $500B investment is actually worth $5B. Isn't that the kind wrong investment that can break nations? Edit: Just to clarify, I don't imply that this is public money to be spent. It will commission $500B worth of human and material resources for 5 years that can be much more productive if used for something else - i.e…

And with the $495B left you could probably end world hunger and cure cancer. But like the rest of the economy it's going straight to fueling tech bubbles so the ultra-wealthy can get wealthier.

Re: DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via RL

#355

Post by raincole »

mrtksn wrote: post #182
HarHarVeryFunny wrote: post #152

DeepSeek-R1 has apparently caused quite a shock wave in SV ... https://venturebeat.com/ai/why-everyone-in-ai-is-freaking-ou...

Correct me if I'm wrong but if Chinese can produce the same quality at %99 discount, then the supposed $500B investment is actually worth $5B. Isn't that the kind wrong investment that can break nations? Edit: Just to clarify, I don't imply that this is public money to be spent. It will commission $500B worth of human and material resources for 5 years that can be much more productive if used for something else - i.e…

> Isn't that the kind wrong investment that can break nations?

It's such a weird question. You made it sound like 1) the $500B is already spent and wasted. 2) infrastructure can't be repurposed.

Re: DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via RL

#356

Post by BriggyDwiggs42 »

baq wrote: post #311
dtquad wrote: post #146

Larry Ellison is 80. Masayoshi Son is 67. Both have said that anti-aging and eternal life is one of their main goals with investing toward ASI. For them it's worth it to use their own wealth and rally the industry to invest $500 billion in GPUs if that means they will get to ASI 5 years faster and ask the ASI to give them eternal life.

Side note: I’ve read enough sci-fi to know that letting rich people live much longer than not rich is a recipe for a dystopian disaster. The world needs incompetent heirs to waste most of their inheritance, otherwise the civilization collapses to some kind of feudal nightmare.

I’m cautiously optimistic that if that tech came about it would quickly become cheap enough to access for normal people.

Re: DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via RL

#357

Post by sdesol »

Alifatisk wrote: post #295

DeepSeek V3 came in the perfect time, precisely when Claude Sonnet turned into crap and barely allows me to complete something without me hitting some unexpected constraints. Idk, what their plans is and if their strategy is to undercut the competitors but for me, this is a huge benefit. I received 10$ free credits and have been using Deepseeks api a lot, yet, I have barely burned a single dollar, their pricing are t…

Prices will increase by five times in February, but it will still be extremely cheap compared to Sonnet. $15/million vs $1.10/million for output is a world of difference. There is no reason to stop using Sonnet, but I will probably only use it when DeepSeek goes into a tailspin or I need extra confidence in the responses.

Re: DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via RL

#360

Post by thrance »

I tried the 1.5B parameters version of deepseek-r1 (same size as GPT2 xl!) on my work computer (GPU-less). I asked it find the primitive of f(x)=sqrt(1+ln(x))/x, which it did after trying several startegies. I was blown away by how "human" it's reasoning felt, it could have been me as an undergrad during an exam.
Post reply on HN