> switched to actual DeepSeek provider
I meant actual DeepSeek API.
11,355 karma · joined August 11, 2008
I meant actual DeepSeek API.
Then I read somewhere (I think X) that OpenRouter adds stuff and breaks caching (telemetry? headers? can't remember). So I stopped the job, switched to actual DeepSeek provider, and voilá, caching 3x more tokens per request (on average).
I hate the fake European foreign-backed right-wing parties but they didn't cause the current situation.
But I'm afraid it might be too late as the cancer spread and did too much damage. Insane regulations, no energy, looming demographic/pension crisis, tax hell, and collapsing industries.
I have a solution for that. Let's use AI to replace all these corporations who just lost their big moats. Conveniently, they just laid off a bunch of people with all the critical know-how and I bet they are very willing to just give it up out of spite.
I don't like Altman and I am still upset about his memory deal last year but he prepared for the current shortages months before anybody else. Meanwhile, Anthropic seems to lack any plans besides third party contracting. IMHO they got very lucky with xAI and Google having spare capacity and willing to rent it. But what about next year?
I'm personally on the other end: local small models working well for specific tasks (i.e. "agentic")
https://nousresearch.com/ They do much more than Hermes
https://pluralis.ai/ has distributed models the network owns
And many others filling very different niches.
I would also count DeepSeek because they publish and share so much. Their new caching prices are very good for not that bad models. Last week I needed to generate content via API and I spent literally $3.40 where before I would've paid easily $200. But you have to connect directly as it seems OpenRouter messes up caching somehow.
Please join/help open source groups doing small + local or distributed models. There's a lot to do. Support truly open source companies.
Let's walk the walk.
With what money? They have to spend the money they get on hardware ASAP else they are left behind.
"At 32 years old, Bonin was the least experienced pilot on the flight crew. When the aircraft's pitot tubes froze and iced over in a storm, the automated systems temporarily failed and disabled the autopilot. This forced the crew into manual flight. Because the Airbus A330 features independent, non-linked sidesticks, the other co-pilot in the cockpit, David Robert, could not physically feel that Bonin was holding his stick back. The aircraft's computer simply averaged their opposing inputs."
The experience of pilots has been dropping like a stone. This is hidden due to new technology, but when unusual situations arise many current pilots have no situational awareness.
--
"A vote to reduce the 1,500-hour rule for pilot training will mean blood on your hands when the inevitable accident occurs as a result of an inadequately trained flight crew."
-- Captain Chesley "Sully" Sullenberger
Do you know anything about the state of bond markets?
This negative cycle is even more pathological in co-ops than in a normal corporation/for profit.
I hope some day someone figures a better way.
Even distillation would be obvious.
And I wouldn't be surprised Anthropic is running the smaller inference models, keeping the base large model in machines they fully control/own.
I tried registering a couple of years ago, but it required a female co-founder, N employees in the EU, and a laundry list of ridiculous demands unrelated to technology or real startups.
And even the paperwork to start a company in the EU is a PITA. You have to incorporate in Bulgaria or something if you want anything in reasonable times. And good luck with Social Security and tax agencies.
Forget it.
Over-promising is a very stupid thing. Nobody will value the intermediate steps. Nobody will value all the effort because they will always compare us with frontier models made with billions and we will become a running joke. So please stop.
[0] US AI firms team up in bid to counter Chinese 'distillation' (Apr 7) https://finance.yahoo.com/sectors/technology/articles/us-ai-...
I wish this was true but it is not. And I am working on open source models so if anything, I would have a bias towards agreeing with you.
Frontier closed models (GPT/Claude) are gaining distance to everybody else. Even Google, once the king.
Your claim is a meme coming from benchmark results and sadly a lot of models are benchmaxxed. Llama 4, and most notably the Grok 3 drama with a lot of layoffs. And Chinese big tech... well they have some cultural issues.
"Qwen's base models live in a very exam-heavy basin - distinct from other base models like llama/gemma. Shown below are the embeddings from randomly sampled rollouts from ambiguous initial words like "The" and "A":"
https://xcancel.com/N8Programs/status/2044408755790508113
---
But thank god at least we have DeepSeek. They keep releasing good models in spite of being so seriously resource constrained. Punching well above their weight. But they are not just 6 months behind, either.
BTW thank you for raising this.
Macros, dude.
It's a bit annoying to become disciplined in text structure/regularity to have macros work flawlessly, but it's a bliss afterwards.
And then... recursive macros.
(I know Emacs has better macros but vim macros are way easier and fun for everyday use).
Weird. There's a recent NVIDIA MLIR that is quite good and fast. Or they could target the even easier and more recent/fashionable tile IR [1] used by CuTile [2] (a little bit higher level but significantly easier to target, only loses on epilogue fusion and similar).