oofbey

3 ms·

1.
by oofbey 2mo ago
“Offered to load conversations”. Let me translate that. “Can we please upload your data to our servers showing how you work with other AI agents?”
2.
by oofbey 2mo ago
If not a system prompt telling it not to read those files unless explicitly told to.
3.
by oofbey 2mo ago
Prompt results!
4.
by oofbey 2mo ago
I’ve tried reading this many times. But it’s sooooo long. Is it worth it?
5.
by oofbey 2mo ago
If only LLMs had been invented at the beginning of your career, you would have been right all along!
6.
by oofbey 2mo ago
Exactly. We as users have zero way to confirm they are honoring even the letter of these agreements, much less the intent. And it's super easy for them to weasel around and find a way to cheat while still having a legal claim to hono
7.
by oofbey 2mo ago
Yeah. Super icky. But this might be the first time in Zuck’s life he’s being honest about the business model.
8.
by oofbey 2mo ago
Wow. Ballmer fans. Hot take - I like the novelty! And yeah, I never really thought about how Bing, Azure, and Office 365 are all actually stellar products. Somehow I missed that. Azure was particularly good in the early days when you could
9.
by oofbey 2mo ago
Was Microsoft dying under Balmer? I think we’d say yes. Even if the SEC filings indicated otherwise.
10.
by oofbey 2mo ago
That's their motivation, for sure. But it's also unambiguously making their product worse and harder to use legitimately. Which pushes customers further towards use of open weight models which don't have these restrictions.
11.
by oofbey 2mo ago
Vector search is a great application for this. You can run the HNSW algorithm on the riscV cores local to the memory. Which is even better than GPU since it’s actually MIMD with tons of cores so each can follow its own branching logic. And
12.
by oofbey 2mo ago
This is such a novel and strange device. I can imagine it might be useful for a lot of things, but what? The non-Von Neumann architecture (putting compute next to memory) has been tried many times and AFAIK never taken hold. What would you
13.
by oofbey 3mo ago
That’s really the worst feature of it. When you first set it up you’re asked at least a dozen questions that you probably have no idea what they mean. But you have to pick something. And whatever you pick on that first day setup you are stu
14.
by oofbey 3mo ago
This is a massive gift to FB. $17B paid is a few weeks of revenue and a single quarter’s worth of profit. But even worse, this is paid out over 10 years! A billion dollars per year to eliminate this major risk is an absolute gift from the A
15.
by oofbey 3mo ago
They will definitely be selecting the lowest bidder for this. Or perhaps a more expensive bidder if they can find one whose proprietary scrubbing technology is “a half dozen regexes our intern thought up”.
16.
by oofbey 3mo ago
Former Googler here. E2EE is easy. Nobody gets promoted at Google for solving easy problems. In fact if you set out to solve an easy problem, it looks bad at performance review time.
17.
by oofbey 3mo ago
All true. HE will never be used for anything real because it’s way to slow and inefficient, meaning you can only run the stupidest models on it. And there’s no commercial incentive to make it work because collecting data is too valuable. Bu
18.
by oofbey 3mo ago
Yeah that was a clever bit of foresight there.
19.
by oofbey 3mo ago
Really curious to see how ByteDance’s 10T model works out.
20.
by oofbey 3mo ago
I think the Top500 benchmark is biased here. It’s a supercomputing benchmark which means it’s going to heavily weight SIMD and float64 performance, which are things that rarely matter for everyday computing. SIMD has uses in the real world
21.
by oofbey 3mo ago
The core idea of the RLM paper is to make a regular LLM act more like a coding agent - offload context to something external that needs to be explicitly queried instead of filling up valuable context. The "recursion" part of the
22.
by oofbey 4mo ago
As agentic coding matures, this kind of project is the right direction. The mechanics of writing the code become less important. But having a language framework that naturally resists mistakes will become increasingly useful and important.
23.
by oofbey 4mo ago
That’s for the consumer app / chatbot. For the api the terms are different: https://platform.kimi.ai/docs/agreement/modeluse
24.
by oofbey 4mo ago
That depends. There are two wry different processes that both get called distillation. One is where you have a fully trained large model and you are converting it to a smaller model. That kind if vastly cheaper than training a full model.
25.
by oofbey 4mo ago
These things take months to train. No chance this is a reaction to what just happened.
26.
by oofbey 4mo ago
Correct: can't opt out of training. This is well documented. "Can't use for commercial purposes" - incorrect AFAICT. In what sense do you mean this? The open weight MIT version obviously allows for commercial use, but
27.
by oofbey 4mo ago
Agreed on the likely mechanism. I'm not sure "overfitting" is even the right description. These things are of course absurdly complicated, and evaluating their quality down to a single number involves a lot of judgement and
28.
by oofbey 4mo ago
It’s not that. The tumors they create in mice are just really fragile compared to natural tumors. Natural tumors that actually cause problems have probably grown for years and learned to avoid immune responses. They’re not densely packed an
29.
by oofbey 4mo ago
The blog is highly suspect, but the study is real. That said it’s not a big deal. Curing cancer in a mouse model is not at all uncommon in new therapies. Mouse models like this are vastly easier to treat than real world cancer for a bunch o
30.
by oofbey 5mo ago
Another year, and OpenAI comes up with yet another naming scheme for their models. First it was integers (GPT2, GPT3). Then they added friendly names (remember Ada, Babbage, Curie, Davinci?), but decided against it. Instead we got dot in

More ›