nl

13 ms·

181.
by nl 3mo ago
> the training cost of the model(s) that Kimi distilled? The distillation process involves getting conversation traces from the model you are distilling from, and then training your model against them. You still have to do the training
182.
by nl 3mo ago
OpenAI's financials leaked and showed this pretty convincingly. Anthropic was probably profitable last quarter, without training costs: https://www.wsj.com/tech/ai/mind-blowing-growth-is-about-to-...
183.
by nl 3mo ago
I think it's unclear the the DS hosted prices are profitable. AFAIK that haven't claimed that. OTOH, the multiple providers who have settled around the same price point ($3.48/M output tokens for multiple providers with good
184.
by nl 3mo ago
This is untrue. Read the "Chipping Away at the CUDA Moat: Using Agents to Enable Day 0 Support" section of [1] where they list a number of agent-created PRs that have been integrated into production vLLM/SGLang code and as we
185.
by nl 3mo ago
Did anyone think RAM prices were coming down soon?
186.
by nl 3mo ago
Freiburg University is pretty well known isn't it? Hiring in university towns is a pretty standard practice for startups outside SF.
187.
by nl 3mo ago
> I can not imagine some shelling out $200/month and then using that product lightly. The people paying for the plan are not the same people using it. Of 6 people I have data on the $200/plan only 2 regularly use more than $400
188.
by nl 3mo ago
There's nothing stopping you using OpenAI models for scam call centers now. OpenAI themselves reported on similar use in February: https://www.reuters.com/world/asia-pacific/dating-scams-fake...
189.
by nl 3mo ago
The two aren't comparable. You can't reconstruct the same LLM weights even with the same training data. You can fine tune an open weights model. Much of the value in modern LLMs is in the RL envrionments, not the pre-training data
190.
by nl 3mo ago
> Sony Walkmans are this The cheapest one I'm seeing is the Sony NW-A306. That's around $600 - I think it fits into the OP's category of "some sort of enthusiast audiophile flac player thing that costs way too much&qu
191.
by nl 3mo ago
You are on a Pro plans and standard seats on Team plans: Fable 5 isn't included in your plan's usage limits. For Premium seats it is. https://support.claude.com/en/articles/15424964-claude-fable...
192.
by nl 3mo ago
> I'm not convinced that the 200 dollar plans are unprofitable. Especially considering not everyone is tokenmaxxing, and in most parts of the world people take leave and companies do not cut their subscriptions. I suspect they are p
193.
by nl 3mo ago
You can bet that most people on those plans do not tokenmax.
194.
by nl 3mo ago
> If I want to buy a new MP3 player, I have to buy something from a company I've never heard of with a terrible screen, use an old phone, or buy some sort of enthusiast audiophile flac player thing that costs way too much. Is this s
195.
by nl 3mo ago
It's wrong - most of this work was published before 2024. But Tsimerman in particular is very AI pilled and has talked about how he thinks LLMs will be doing better work that most mathematicians in 2 years: https://x.com&#x2
196.
by nl 3mo ago
Here are a couple of Fields medalists' work that had direct practical applications less than 5 years after publication: Terence Tao's work (with Emmanuel Candès and Justin Romberg) on compressed sensing. Published in 2004-05. By 2
197.
by nl 3mo ago
Tsimerman's work is directly applicable in two fields of computer science: O-minimality can be used to simplify formal verification.
198.
by nl 3mo ago
That's interesting. I only did a couple of attempts with Fable and it seemed fine. Mostly I have been using GPT 5.5 and now 5.6 That is notable because I do almost exclusively use Claude for coding.
199.
by nl 3mo ago
AI consumed around 0.5% of the world’s electricity in 2025[1] Where I'm from data centers help the renewable mix by subsidizing transmission from other geographic zones. I'm actually improving the environment by using it. [1] http
200.
by nl 3mo ago
I think it's important to note the context though: There had been a RAM glut that meant Micron lost money every quarter in 2023, with revenue almost halving vs 2022. This was post-pandemic, and demand had dropped a huge amount. The RAM
201.
by nl 3mo ago
On RAM supplies, the best estimates I've seen are https://www.tomshardware.com/pc-components/dram/cxmt-close-t... They are estimating a 25% shortfall in supply remaining in 2030, even with Chinese companies r
202.
by nl 3mo ago
ChatGPT 5.5 and Sol 5.6, Fable are good. I haven't really tested Opus 4.8, but 4.7 wasn't nearly as good as ChatGPT 5.5.
203.
by nl 3mo ago
> I’m sure they will end up with a fun little toy from the whole endeavor they can play with for 2 weeks before abandoning. That's exactly the idea - except it's more like 2 hours before I prototype the next version. > Maybe
204.
by nl 3mo ago
Because it reads the code and generates this in a single pass in 5 minutes. I haven't read the code. Why would I do it in a slower, more difficult way for something that's going to be outdated in 2 hours?
205.
by nl 3mo ago
To add to to this looking at the #baaw (bikes against a wall) tag on insta It was over 200 pics before I found a pic of a bike that wasn't facing left-to-right https://www.instagram.com/explore/search/keyword&
206.
by nl 3mo ago
For example I do SVG diagrams to explain architecture and specific data flows in a multi-tier app. LLMs do a great job because they understand both code and SVGs well. Edit: An example for a synth I'm buulding: https://imgur
207.
by nl 3mo ago
I use LLMs for 3D CAD design in OpenSCAD. There seems to be a very strong correlation between models that are good at SVG and models that are good at 3D CAD. Anecdote I know, but there does seem to be generalization going on here.
208.
by nl 3mo ago
The hyperscalers signing these contracts have decent legal departments. Think about Oracle for example - I'm pretty sure they know every trick there is about beneficial contract drafting. I don't think they need some special pro
209.
by nl 3mo ago
As noted in my other comment you can get obsolete GPUs (P100s, BC250s) with lots of RAM on AliExpress now. It hasn't proven revolutionary.
210.
by nl 3mo ago
There's nothing stopping you doing this now. You can get used 16GB P100s on AliExpress for ~$100 if you want obsolete GPUs. Allegedly new AMD BC 250s are only slightly more. I've looked at this some but I already have a GTX1070 wh

More ›