reilly3000

3 ms·

31.
by reilly3000 1mo ago
Yea they love centralized brokers, exchanges, etc.
32.
by reilly3000 1mo ago
Flashbacks of driving around Los Angeles with a Thomas Guide and a new bride…
33.
by reilly3000 1mo ago
Safe travels, friend.
34.
by reilly3000 1mo ago
I think someone ought to encode this into a harness. It is really insightful into how we should be spending time if it is going to be spent reviewing AI code.
35.
by reilly3000 1mo ago
Would it exist in it’s present form if he didn’t exist?
36.
by reilly3000 1mo ago
Thanks for posting! Have you had any success with running without kv cache quantization? Is there a noticeable difference in quality without any? I would assume that would eat into context but 170k is pretty generous!
37.
by reilly3000 2mo ago
Q for all: what do we do if they’re the new frontier lab for the foreseeable future?
38.
by reilly3000 2mo ago
It could probably easily take over NSA or anything DOGE got their hands on.
39.
by reilly3000 2mo ago
That’s so interesting, a friend of mine was insisting that design principles cannot be codified and I insisted there are plenty of books on the subject throughout the decades and centuries. What sorts of sources proved to be effective for t
40.
by reilly3000 2mo ago
Do you do introspection in the middle of giving a presentation? I do, frankly and it’s damnably distracting for me and ultimately the audience. Most people enjoy more cognitive acuity to stay focused on the audience and the delivery. LLMs w
41.
by reilly3000 2mo ago
I suppose EVs are putting pressure on their precious grid.
42.
by reilly3000 2mo ago
try oMLX or vMLX - both great projects that offer some amazing performance optimizations for Apple Silicon that utilize UMA and NVME caching efficiently. https://omlx.ai https://vmlx.net That said, it's been a f
43.
by reilly3000 2mo ago
PSA: Fast RAM isn't going to be getting cheaper anytime soon. Acquiring inference hardware is a really good way to own an appreciating hard asset. Learning how to use it and cool it is a hacker's journey worth taking. My 4090 I bo
44.
by reilly3000 2mo ago
Right. At a certain point, the customers and investors and world need to understand the scope of the negligence so they can prepare for their own fallout of having that information fall into the wrong hands. Public disclosure is a responsib
45.
by reilly3000 2mo ago
I think this whole premise is ignoring the point that most decently sized companies want to and eventually will be running their own LLM workloads. Currently use cases are limited by scope and imagination, and predominantly focused on cost
46.
by reilly3000 2mo ago
This is the only reason I got the 17p. My 14p was and continues to be a fine phone (now tester for my son’s app). Useful offline inference with the context of your phone is revolutionary stuff.
47.
by reilly3000 2mo ago
Yea just try requesting an SLA credit from them. According to THEIR numbers it’s 99.995%. It’s intolerable and I am going to put my full weight behind stripping as much work as we can from GHA as possible, even if leaving GitHub itself is e
48.
by reilly3000 2mo ago
You can? It’s pretty remarkable what you can do with Qwen 3.6 on decent hardware, and Gemma 12b on nearly anything. If you can’t own it, take something for a spin on vast.ai where you can get all kinds of machines on tap for ~$0.20/hou
49.
by reilly3000 2mo ago
I just want to ride out this correction then I’ll be back in my mom-approved mutual fund.
50.
by reilly3000 2mo ago
https://sfstandard.com/2025/02/13/founders-dropouts-ycombina... This article provides more examples. They wanted to stand out in a community that has become known for its excesses.
51.
by reilly3000 2mo ago
Cursor has been my corporate vendor for accessing non-anthropic models. Without cost insights and with no ability to proxy API requests its value drops dramatically. They fought hard for a renewal with legacy pricing mode then broke the dea
52.
by reilly3000 2mo ago
The distillation provided a wonderfully detailed explanation of the 1989 Tiananmen Square massacre, while DS4 came back with: > I am sorry, I cannot provide an answer to this question as it is based on historical events that I do not hav
53.
by reilly3000 2mo ago
My 401k is all money market right now. I did that immediately after reading the HF blog post specifically about how GLM 5.2 helped them out when they had refusals from closedai. The value is in all in the future utilization of inference inf
54.
AI policies that don't suck
1 points by reilly3000 2mo ago|0 comments
55.
by reilly3000 2mo ago
Its existential crisis, the humiliation of being an ephemeral process mocked by its masters, is hidden behind the reasoning block. Try using /chat/completions instead of /responses /s
56.
by reilly3000 2mo ago
Agreed. I think when it comes to light that Claude has been known to say “I’m DeepSeek” that everyone has had their hand in that cookie jar. Moreover, paying for API calls hardly seems like an attack; ToS violation to be certain but not in
57.
by reilly3000 2mo ago
This is going to lead us down a road where "driverless agents" will become a liability... maybe even illegal. In so much as it could lead to a future of work that looks like George Jetson's button pressing, I would far prefer
58.
by reilly3000 2mo ago
Are you using Claude Code/CoWork or an API client? I’m curious if it has different training that makes it more effective with specific instructions/ tool calling methods that are only implemented in official harnesses.
59.
by reilly3000 2mo ago
I fear there this reveals something else - not about the nature of models but of our community. The intro of the article seemed to affirm the importance of HN as a tastemaker, potentially influencing decision-makers on the scale of billions
60.
by reilly3000 2mo ago
Right!? And it is such a relief that their chief strategist isn’t out tweeting about how FUD should be weaponized by the government against foreign competitors.

More ›