News for hackers. Stuff that matters.

starship731

Re:Open AI Announces $1.5M Bonus for Every Employee on Thursday August 07, 2025 @11:46AM (#44826030)
Yeah, Yahoo & AOL still exist today as independent companies don't they?
Re:Open AI Announces $1.5M Bonus for Every Employee on Thursday August 07, 2025 @11:35AM (#44825860)
Same thing was sad of the Internet in 1999 when Yahoo had a market cap of $125 billion, and AOL $200 billion. Where are both companies today?

It's déjà vu all over again it seems.

Re:Open AI Announces $1.5M Bonus for Every Employee on Thursday August 07, 2025 @11:33AM (#44825829)
The vast majority of countries in SS Africa, including ones with quite good QoL (Zambia, Botswana, Cote d'Ivoire, etc).
Re:DeepSeek-R1-0528 Did Not Have a Moment on Saturday June 07, 2025 @03:21AM (#44208036)
Deepseek R1, the Reasoning Model, never had a 'moment' when it was released.

It was DeepSeek, the Mobile App that consumers (including stock traders, analysts, & investors who don't follow the day-to-day developments within the LLM field) could download from App Stores that had the 'moment'.

And it was the AI chat app that led the professional talking heads to learn about the Language Model (DeepSeek R1) powering the Chat App, the Scientific Paper explaining the techniques used to build the core model (Deepseek V3), and the said paper's claims of the model being developed for a far cheaper amount than Western models.

Re:DeepSeek-R1-0528 Did Not Have a Moment on Saturday June 07, 2025 @03:12AM (#44208008)
What utter and complete rubbish. Deepseek is available on all the major LLM API hosts, from TogetherAI, to Replicate, to Fireworks, to OpenRouter, to NVIDIA NIM, to Groq, to Hugging Face, to GitHub, to Cloudflare Workers AI. Hell, even Scaleway, a FRENCH provider, offers it.

Pls do a bit of investigation before posting nonsense. Low information commenters like you detract from the conversation.

Best selling EV manufacturer in Brazil, Thailand, Malaysia, and the Entire Southeast Asia, and we're supposed to believe the zero-source provided claims of some random HN poster. European car manufacturers should really invest more in their "Trash Chinese Competitors' Online Reputation" department.
Re:DeepSeek-V3 Technical Report on Thursday March 27, 2025 @02:37AM (#43490931)
Deepseek v3-0324 (new checkpoint) beats ALL but 1 proprietary AND non-thinking LLMs by a significant margin. Check livebench.ai & Artificial Analysis benchmark for details.

The only non-thinking LLM the new V3 doesn't decisively thrash is GPT 4.5 which is more than 100 times more expensive than V3 and yet is only a few (essentially negligible) percentage points better than it.