_superposition_

5 ms·

91.
by _superposition_ 2mo ago
Melancholy ellipsis...
92.
by _superposition_ 2mo ago
Geocities is broken otherwise thanks!
93.
by _superposition_ 3mo ago
This. People forget most food we eat is at least 50% water.
94.
by _superposition_ 3mo ago
Antacids did this to me. I was convinced it was electrolytes, would take supps etc but I believe it came down to calcium absorption. Taking a break from the antacids immediately stopped the cramping.
95.
by _superposition_ 3mo ago
I do the same thing with pickle juice. Works for me.
96.
by _superposition_ 3mo ago
I would recommend listarr as well. Woke up to "disclosure day" this morning without even asking.
97.
by _superposition_ 3mo ago
The only good thing about this in my setup is I can add shows/movies to my watchlist from their advertisements and it gets picked up by my arr stack.
98.
by _superposition_ 3mo ago
The cost of business nowadays? At what point does it become apparent that in today's world software needs continuous pen testing and scanning? If you don't your attackers will.
99.
by _superposition_ 3mo ago
I don't necessarily buy this view as plausible other than the fact that in my experience the majority of large companies are dysfunctional politically and technically.
100.
by _superposition_ 3mo ago
I've wondered with all the generated code, will llm provenance be a thing in the future? I can't imagine the DoD/DoW wants qwen writing its code.
101.
by _superposition_ 3mo ago
Shakes fist at cloud
102.
by _superposition_ 3mo ago
It was good enough that I read it anyway.
103.
by _superposition_ 3mo ago
We shouldn't need AI to do this
104.
by _superposition_ 3mo ago
Funny one of the very first things I tried when I got my hands on chatgpt was to play the game "zork" with me. I was initially floored by the fact it could recreate the game on the fly, seemingly accurate, but it fell off the rail
105.
by _superposition_ 3mo ago
That was way too much work...
106.
by _superposition_ 3mo ago
Agreed, this will save us all money in the long run.
107.
by _superposition_ 3mo ago
I think we have a need to revise the old let me Google that for you thing Click the link to view conversation with Kimi AI Assistant https://www.kimi.com/share/19f6b96d-fdd2-8589-8000-0000daada...
108.
by _superposition_ 3mo ago
The whole point is that a dsl syntax is constrained enough to fit in the context window and not rely on training data. LLMs are damn near built for this exact use case.
109.
by _superposition_ 3mo ago
It's NIMBY all the way down. Murica...
110.
by _superposition_ 3mo ago
Or an incoming PR ;)
111.
by _superposition_ 3mo ago
I think it all comes down to tight, context efficient, deterministic feedback loops. Pre commit hooks work well for this type of thing. Ideally I think a set of those hooks should run on every file edit, however I haven't gotten around
112.
by _superposition_ 3mo ago
I get what your saying, however we don't need to be stuck with the Claude code harness. You may find this interesting. https://www.langchain.com/blog/tuning-the-harness-not-the-mo...
113.
by _superposition_ 3mo ago
Hit the nail on the head. I find it surprising people give a fuck any more. Who likes writing documentation? Us devs being horrible at and hating to write documentation was a standard trope before llms. They do a better job and dont complai
114.
by _superposition_ 3mo ago
I know it's a type of blasphemy here, but deterministic workflows such as what you describe is where langgraph really shines imo.
115.
by _superposition_ 3mo ago
I do have a gripe already... The readme says macos, however as of yet I don't see the actual dependency. Sort of defeats the *nix spirit as well, no?
116.
by _superposition_ 3mo ago
I really like this idea and the way you mapped the concepts to unix primitives. Indeed llms are already "unix native". I've been experimenting with similar event driven workflows using k8s primitives but that's one leve
117.
by _superposition_ 3mo ago
I can't really explain it but when I am working with a system in my head it's not represented as text. It has an almost physical dimension to it similar to imagining how an engine works or what have you. This is obviously beyond t
118.
by _superposition_ 3mo ago
This looks awesome. Great work.
119.
by _superposition_ 3mo ago
I can't prove it but I have strong beliefs that the logic and intuition required for abstracting for future changes is not possible in a stream of predicted tokens. Mental models aren't built from text. Something's missing th
120.
by _superposition_ 3mo ago
And they claim naming things is hard...

More ›