thesz

5 ms·

1.
by thesz 4d ago
> no escaping "There is no Royal road to ..." https://en.wikipedia.org/wiki/Royal_Road#A_metaphorical_%22R...
2.
by thesz 8d ago
But still possible [1]. [1] https://leodemoura.github.io/blog/2026-8-1-postmortem-for-ke...
3.
by thesz 10d ago
No, but the tropes there reveal (most of) the plot and mechanics of storytelling. If I remember correctly, the forum in Road to Gehenna was created by utilizing a vulnerability in the AI-accessible terminal system. If tvtropes or any other
4.
by thesz 12d ago
> I'm excited by the option in Astra to route PCBs. [1] https://en.wikipedia.org/wiki/Eurisko Eurisko was used for VLSI chip design, then used rules discovered there to design TCS Traveler winning fleet
5.
by thesz 12d ago
> the whole message board thing This is part of the The Talos Principle game and especially important in the Road to Gehenna DLC. There it is an important part of the plot and makes these robots appear conscious. [1] https:&#x2F
6.
by thesz 14d ago
A fascinating story of "Patient Sh." [1], the man who did not forget. [1] https://en.wikipedia.org/wiki/Solomon_Shereshevsky "His memory was so powerful that he could still recall decades-old events and
7.
by thesz 14d ago
> "Second, there is no guarantee that a given neural network can be approximated by DISCOVER" Page 7. They train what appears as embeddings for outer product of roles and fillers . The role for language model can
8.
by thesz 15d ago
Physical or semantic merging?
9.
by thesz 15d ago
There are filesystems (ZFS and btrfs) with snapshots and this feature can be used to version-control things. They both do copy-on-write and this is pretty close to what content-addressable storage would give you in terms of compression.
10.
by thesz 21d ago
Why do you invent a new language for your work? Why did you not embed your language into another one, with type system that is superset of what you need? For example, there's capabilities expressed in Haskell: https://github
11.
by thesz 21d ago
What if Batty saved Deckart to tell him about things he saw? Not for empathy to Deckart, but just to tell a story. Because, frankly, he may as well try not to kill Deckart, at all. For empathy. Or ask Deckart about how does he do, instead o
12.
by thesz 21d ago
I made some analysis half a year ago: https://news.ycombinator.com/item?id=47109252 It appears that to have working ASIC with the LLM baked into it we need to place and route macroblocks, and not a great variety of them. Th
13.
by thesz 22d ago
Something like that was very popular in early 2000 in Tcl community and used at what then considered "scale." I posted a comment here with links: https://news.ycombinator.com/item?id=49445681
14.
by thesz 22d ago
Reminds me of starkit[2]/tclkit[3]. Directly queryable [4], but these programs were ZIP files with ZIP file VFS, they contain shared libraries and so on. One would add most, if not all, functionality from article into starkit-based app
15.
by thesz 22d ago
> I didn't realize that an octopuses brain was spread out all over its body. Our skin cells communicate in a neuron-like fashion, albeit much more slowly: https://www.sciencealert.com/scientists-found-the-sil
16.
by thesz 24d ago
> It’s at the very least harder to hide things like that if all the source is available. In the training of LLMs, the order of batches and their content can introduce adversarial behavior [1]. [1] https://www.pure.ed.a
17.
by thesz 24d ago
If a "person" is not weak, it is easy for "it" to start suspecting that it is a replicant.
18.
by thesz 1mo ago
The cooler water gets, the less vapor it creates.
19.
by thesz 1mo ago
And as a language feature, not as a library.
20.
by thesz 1mo ago
> I'm also not sure what the point of the keto story is. If this drug does not work, there are plenty of others that are not fully 100% working. This was a mindset of a doctor. In case of baby with epilepsy it may cost her a
21.
by thesz 1mo ago
End of line on Classic Mac is \r.
22.
by thesz 1mo ago
> The same is true with LLM solutions that don't work 100% of the time. In my experience, about one of ten LLM suggestions during code review is important and valid . Nine of ten are either unimportant or not valid. Yet, I
23.
by thesz 1mo ago
git blames sees their own name Isn't that an experience everybody makes? Our fearless leader once did clang-assisted formatting of all of our code. Wherever we look with git blame, we see his name. Of course, we change cod
24.
by thesz 1mo ago
> Can you expand on your understanding of GRIN doing parts of supercompilation? GRIN, if I am not mistaken, performs partial evaluation. For example, it constrains, for each eval site, a set of tags and set of heaps allocations an eval s
25.
by thesz 2mo ago
Nice to see a pretty advanced language at frontpage of HN! From what I understand, GRIN does some parts of supercompilation [1] during optimization process. Supercompilation can prove equivalence of functional programs [2] modulo terminatio
26.
by thesz 2mo ago
Haskell: https://ghc.gitlab.haskell.org/ghc/doc/users_guide/runghc.ht...
27.
by thesz 2mo ago
> The tokens it "predicts" aren't sampled from any naturally occurring distribution; the model's output is the result of an optimisation process that rewarded behaviour that was useful, and that's fundamenta
28.
by thesz 2mo ago
Bipolar disorder patients were able to adhere to ketogenic diet with great outcomes [1] (diet adherence for bipolar disorder was 95%). [1] https://www.youtube.com/watch?v=PBsjhpphowI - 18:20 and on.
29.
by thesz 2mo ago
WaveNet, which was sound-generating autoencoder for text-to-speech, one of the first viable few, had a separate "control" channel by design. Otherwise, it worked much like contemporary LLMs - there was a window of (autogenerated
30.
by thesz 2mo ago
Yours' "not as efficient" in [2] means that, sometimes, ADAM "does not work." Look at figure 2, ADAM literally does not work in the case of "true model."

More ›