NortySpock

6 ms·

1.
by NortySpock 16d ago
" diagnosis during childhood so if you're 30 or 40 you might have no chance " Huh? You don't have to be tested as-a-child, you don't have to call up your fourth grade teacher. You can self report your subjective exp
2.
by NortySpock 21d ago
Risk management. Nvidia has a market cap of $5T USD today, and a decent chunk of that is due to LLM speculation. Does Nvidia want their stock price to be at risk of being tanked by a download service being in the news? No, they want to make
3.
by NortySpock 22d ago
I love good programming languages, if that counts for anything, and I think there's a huge amount of room and value in "using an LLM to translate one programming language to another". I also think your instincts are on a usef
4.
by NortySpock 1mo ago
I've been jealous of the work antithesis is doing and support you in spirit; the world needs deterministic, tested software. If you want a challenging open source deterministic game to test, may I suggest Beyond All Reason, powered by
5.
by NortySpock 1mo ago
See this comment https://news.ycombinator.com/item?id=49208635
6.
by NortySpock 1mo ago
Yeah, been fighting variants of this problem as a data engineer where the upstream database would only let you see the current state of the table and did not provide you a changelog / cdc So you have to (a) repeatedly query incremental
7.
AMD buys Taalas, chip startup that hardwires AI models into its silicon(cnbc.com)
3 points by NortySpock 1mo ago|1 comments
8.
by NortySpock 1mo ago
My understanding is that Taalas burned a Large Language Model as analog silicon, leading to something that ran (roughly) 10x faster for 10x less power ...at the cost of it being burned in forever. Non-programmable.
9.
by NortySpock 1mo ago
Agree, copy-paste feels like it works 80% of the time, which is maddening when 1 in 5 copy paste attempts fail. (mostly from zed to something else) Why?? At least Excel has the courtesy to telling you "Hey the copy-paste buffer said it
10.
by NortySpock 2mo ago
> In a compute-constrained world where model demand is growing faster than capacity I don't buy it. There have been recent weeks where some of the mid-level models (Hy3, Laguna M.1) are free (true for parts of June and July, see H
11.
by NortySpock 2mo ago
Why settle for sampling when you can have the whole dataset? The spiral pattern is an excellent example. "Sure it looks like this when you zoom out, but when you zoom in, you can see the finer structure of the points..."
12.
by NortySpock 2mo ago
Gleam ? Elixir's exhaustive pattern matching and gradual type checking is also on my radar. I guess I don't know what you mean by effects typing.
13.
by NortySpock 2mo ago
Signatories appear to group into (a) sellers of hardware (Nvidia, IBM) and those who rent hardware out to others (Amazon, Microsoft) They want open models because they don't have to license it to run it on their hardware, so they can o
14.
by NortySpock 2mo ago
https://openrouter.ai/deepseek/deepseek-v4-pro#providers Look through the provider list for a company you are willing to do business with?
15.
by NortySpock 2mo ago
Data is heavy. I would say "it's risky and requires a lot of labor to migrate without corruption, loss of data" and also minimizing downtime. Sure anyone can run pg_backup, but can you do it across 90 databases? Can you do it
16.
by NortySpock 2mo ago
Dunno, I have not used either of those. (Had been using zed and ollama, and ollama had plenty of odd defaults that needed fixing) Glancing through the docs, I would be digging down in the config of both Android studio and lm studio for eith
17.
by NortySpock 2mo ago
If the local model is served via ollama, there's a default timeout of 10 minutes , which can be adjusted either per-call , or (as I did) in the systemd service environment variables https://docs.ollama.com/faq#how-do-i-
18.
by NortySpock 2mo ago
It's difficult to make money if you have to spend all the money on defense lawyers. Once Trump dies of natural causes I assume some intrepid journalists will go through the mountain of evidence and dig out some exciting snippets of wha
19.
by NortySpock 2mo ago
Similar interest here, possibly including if qwen 3.6, Gemma4 or DiffusionGemma (with the largest quants that will fit in a single card) will offer, say, 50 tokens-per-second (fast enough for interactive human-in-the-loop code research, p
20.
by NortySpock 2mo ago
Have you seen https://news.ycombinator.com/item?id=48892015 and yoloai? https://github.com/kstenerud/yoloai Seems like both projects are following very similar approaches.
21.
by NortySpock 2mo ago
"integrated graphics processor, using system memory" had its name dragged through the mud for decades. So we had to rebadge it to "unified memory". Curious if we'll ever see some old integrated graphics processor &q
22.
by NortySpock 3mo ago
I find it to be a useful tool for summarizing things, creating examples, and as a tutor for explaining a topic using analogies. Plus it can generate and iterate on code snippets. Like, I personally find python pandas documentation unusable
23.
by NortySpock 3mo ago
You can work in the technology department of a non-tech company. I am enjoying being a part of an employee-owned company so far.
24.
by NortySpock 3mo ago
Ok, and? You can live with that if there are more important things to deal with. I've stared at ugly LLM code, that I had just had generated, and worked well enough for my purposes. (generally, some quick recursion into a nested pytho
25.
by NortySpock 3mo ago
I have been looking at this sandbox tool, but haven't quite figured out how I want to deploy it, and anyways I haven't found an agent I like yet. https://github.com/kstenerud/yoloai
26.
by NortySpock 3mo ago
I play with my brother and an old friend of mine. I'm getting close to forty. There's no "leagues", it's just lobbies you join, and you gain or loose OS (Elo) per battle. Party up with your friends, find a lobby tha
27.
by NortySpock 3mo ago
Speaking as an occasional contributor: there's a roadmap towards that, but there's a few components that need to finish baking before we release on Steam. e.g. the current lobby protocol isn't capable of handling Steam levels
28.
by NortySpock 3mo ago
BAR does have a series of scenarios for singleplayer / offline play. Or in multiplayer you can arrange a co-operative game with humans against AI opponents, which often has substantially less flaming involved, especially when playing a
29.
by NortySpock 3mo ago
Knowing what to build (and that it hasn't already been built or bought elsewhere in the company) requires bits of information / person-to-person networking / visibility into the state of the company that not all managers or V
30.
by NortySpock 3mo ago
You can also observe this in games like Dyson Sphere Program, (which is all workers and queues and buffers) where adding a buffer storage section of a belt only hides the fact that you are under-producing one of the components required. The

More ›