shay_ker

4 ms·

31.
Apple sues OpenAI/io for stealing trade secrets(drive.google.com)
6 points by shay_ker 3mo ago|1 comments
32.
by shay_ker 3mo ago
Didn't we all know from the start that all of SWE-Bench was flawed? Even the authors concede the limitations and have long since moved on.
33.
by shay_ker 3mo ago
Great work! My question is about what you mentioned at the end - how well do world models operate when out of distribution? In some sense we hope these models learn something "deeper" about how the world works and can apply that k
34.
by shay_ker 3mo ago
I saw the integration with Hono - hadn't heard of it before, do many people use it?
35.
by shay_ker 3mo ago
I'm curious if this approach can be generalized beyond "doom loops". For instance, another way of thinking about a "doom loop" is wasted tokens, which happens all the time with larger models that are inefficient at
36.
by shay_ker 3mo ago
cost per benchmark task is definitely interesting! i've always wanted cost per prompt, but even that has too much variation.
37.
by shay_ker 4mo ago
random thought, but i wonder if html-in-canvas can really improve development for visualization tools specifically
38.
by shay_ker 4mo ago
I have a naive question: why did this take 15 years? I understand that good APIs need time and thoughtful design, but I struggle to understand why we couldn’t get to the same (or better) solution faster.
39.
by shay_ker 4mo ago
I do wonder - why not add this to jsx?
40.
by shay_ker 4mo ago
Seems like Angular has gotten better since v2 (my last experience). Has anyone done a modern Angular vs. React comparison that's not an AI slop article? I'm also curious if it's "simple made easy" for performant app
41.
by shay_ker 4mo ago
It's fascinating that cloud providers like AWS/GCP/Azure are now immovable "enterprise" technologies, in the way that IBM, Oracle, SAP, etc. were 15 years ago (and still are!). Fond memories when only startups used
42.
by shay_ker 5mo ago
hmm it does seem like a good point that browsers should be able to do this natively. the DOM already holds the content, and browsers manage what's rendered in the viewport.
43.
by shay_ker 5mo ago
hot take: token spend can be used a honey pot, especially when compared to what you deliver. spend accordingly!
44.
by shay_ker 5mo ago
In the three options OP presents, I wonder if there's a fourth: BYO model Customers give vendors metered access to their model. They can budget tokens per vendor. Vendors selling "AI products" can have a cleaner story and win
45.
by shay_ker 5mo ago
> "requirements were always the bottleneck" > "faster typing won't make you faster"..... I understand a Deloitte consultant has specific incentives. But let's first try to answer a baseline question: why
46.
by shay_ker 5mo ago
no tool calls!
47.
by shay_ker 5mo ago
right exactly, but clearly it's possible to elicit the behavior we want in the model, which means the capabilities are there!
48.
by shay_ker 5mo ago
Hmmm... I'd like someone to double check my thinking here. I posted this exact prompt for gpt 5.5 xhigh: ``` does this look right to you? don't do any searches or check memory, just think through first principles static int vpu_mm
49.
by shay_ker 5mo ago
hm yeah. it could be nice for apple to have a list of shortcuts that'd actually be useful based on real activity. but getting all the info needed is hard.
50.
by shay_ker 5mo ago
It's really fascinating to read this, since I've encountered similar memory issues in other languages (ruby, go, etc.). Debugging these issues is a pain. Is there a way to make all this much easier to debug and to prevent memory i
51.
by shay_ker 5mo ago
How does it compare to popular local inference engines, e.g. ollama, lm studio, or handrolled llama.cpp? I saw a brief benchmark in the readme but wasn't sure if there was more.
52.
by shay_ker 5mo ago
curious that they are doing speculative decoding and not baking MTP into the model, like Nemotron https://docs.nvidia.com/megatron-core/developer-guide/0.15.0...
53.
by shay_ker 5mo ago
antirez: i'm curious, with the final code, have you experimented with effectively one-shotting the final result? i wonder if we can get there with GEPA, and maybe there's something we can learn in how to elicit/prompt these m
54.
Claude Design Is Real Design(diverging.run)
2 points by shay_ker 6mo ago|0 comments
55.
by shay_ker 6mo ago
It's a good question - is blackbox hacking as effective as whitebox hacking, for AI agents? I've gotta assume someone at Anthropic is putting together an eval as we speak.
56.
by shay_ker 6mo ago
with a huge multi-tenant database, how do you deal with noisy neighbors? indexes are surely necessary, which impose a non-trivial cost at scale.
57.
by shay_ker 6mo ago
I'd love to see an example PR on lubeno vs github. That might show the comparison a bit better.
58.
by shay_ker 7mo ago
A general question - how do frontier AI companies handle scenarios like this in their training data? If they train their models naively, then training data injection seems very possible and could make models silently pwn people. Do the labs
59.
RocksDB unit test finds a CPU bug(rocksdb.org)
3 points by shay_ker 7mo ago|0 comments
60.
by shay_ker 7mo ago
I remember this post. But I'm not sure what the future really entails and I suspect it'll be very company/culture dependent. In some companies, the engineers are very savvy and understand the business well. In others, it&#x27

More ›