j-pb

7 ms·

1.
A New Observable(observablehq.com)
2 points by j-pb 6d ago|0 comments
2.
by j-pb 17d ago
Self referential nonsense. You cannot have something that that is both a model of something and unprecedented. It'a even in their name, they model language, language is the thing that humans use to communicate, therefore they model com
3.
by j-pb 19d ago
What do modern LLMs act like then to you?
4.
by j-pb 19d ago
None of these animals breed on their own, they are mostly artificially inseminated. You truly have no idea what you are talking about.
5.
by j-pb 20d ago
It's having condciousnedd while it acts as if it's having consciousnes. It is powered while it is powered. Under general anesthesisa you are still alive, but you don't have consciousness.
6.
by j-pb 20d ago
You don't get it, do you? The "genocide" that you describe is happening constantly. None of the chicken we have today will be alive in a couple months. We kill them to eat them. If we cease to "use" them, we just ke
7.
by j-pb 20d ago
Wait... do you really think that if everyone goes vegan these animals will get killed, but if everyone eats them they won't? You are aware that all of these livestock animals didn't exist a few months to years ago? And that we go
8.
by j-pb 20d ago
So because you eventually die, it means that you were never alive? The weights of an llm obviously don't have consciousness. It's the process that potentially has it. Are you conscious on a nanosecond timescale?
9.
by j-pb 20d ago
The burden if proof is on you in this case. The philosophical zombie does not satisfy occams razor because it asserts an unmeasurable property that has no effect on the outcome. If it acts like it has consciousness, then assuming it to have
10.
by j-pb 1mo ago
There is no hidden inner monologue. Human inner monologues produce measurable micoactivations that are the same as the ones produced in speech. If it's not in the inner COT it isn't there.
11.
by j-pb 1mo ago
Sociopaths can feel emotions, they just lack empathy. So they can consciously manipulate others. LLMs don't have conscious processes that perform that kind of reasoning. If anything the fact that they are blackboxes to their own behavi
12.
by j-pb 1mo ago
I always feel like people leave out the third case of the analogy: indexing The article itself has decision trees for the compression explanation, which is also a lookup index. In each case you try to recognise (re)usable structure. Self-in
13.
by j-pb 2mo ago
I discovered that we got that back in a sense. If you task an llm with writing a CLI and have it use it, then it hides the "user interface" to a sufficient degree that you can pretty much build any kind of app/functionality i
14.
by j-pb 2mo ago
Ah that's just because the new training set also includes "The King in Yellow".
15.
by j-pb 2mo ago
Holy shit, thanks for mentioning it, the frontier labs truly have no idea what they are doing when it comes to software quality.
16.
by j-pb 2mo ago
Neat! Thank you for the in-depth! So you're also using succinct datastructures with Elias-Fano, you might want to check out this paper https://aidanhogan.com/docs/ring-graph-wco.pdf on WCO joins over succinct inde
17.
by j-pb 2mo ago
Cool stuff! I still always want to know the numbers per edge. Because there'a a reaon why these graph DBs try to hold the stuff in memory, everything else is slow as heck. We have ~30bytes per edge in our succinct indexes, which makes
18.
by j-pb 3mo ago
You don't know how much more vibrant the world is knowing that you still make games in it. Thank you for all of those hours. (And making the tools that help me teach my apprentices.)
19.
by j-pb 3mo ago
> movement can be executed more aggressively and quickly That's exactly the reason why it's usually a bad idea to run a classical robot on a neural controller. If atlas bumps into something you get a small bump and maybe a brok
20.
by j-pb 3mo ago
These robots operate on completely different principles. One can lift insane weights, has insane torque, and absurd precision, and can do the same movement millions of times with virtually no deviation. You program these with an exact movem
21.
by j-pb 4mo ago
I know these strengths, I've written Zig fulltime for ~1 year before switching to Rust, and I do miss comptime pretty much every day. Still in my experience the strengths do not outweigh the weaknesses. I'd also push back on the n
22.
by j-pb 4mo ago
Correct me if I'm wrong, but the three largest Zig project (by far, with a huge gap between them and the rest of the pack) are Bun, Ghostty, and TigerBeetle. A language so niche that it only has 3 major projects is a liability. Now it
23.
by j-pb 4mo ago
Read through the stuff yourself if you don't believe me: https://github.com/ziglang/zig/issues?q=is%3Aissue%20label%3...
24.
by j-pb 4mo ago
I'm not expecting the whole language to be stable, but I expect certain parts of it to be more stable than others. E.g. control flow vs. async. I'm not saying that they can't work that way, more power to them. But then having
25.
by j-pb 4mo ago
> not from the language they use for the implementation. Everybody has a plan until their compiler compiles an if-else wrong, and that's the level of instability that Zig had when I left it and Tigerbeetle committed to it. No amount
26.
by j-pb 4mo ago
I'm pretty sure they'll miss the full developer salary that Oven used to sponsor them, which they no longer do. I'd wager one doesn't do a rewrite like that, if you are in great personal standing with the language founda
27.
by j-pb 4mo ago
On one hand I kinda feel validated for having jumped ship on Zig 3+ Years ago[1] and moving everything to Rust[2], with the language simply being too unstable and unsafe in my eyes, despite my love for comptime and people arguing that Bun a
28.
by j-pb 4mo ago
On one hand I kinda feel validated for having jumped ship on Zig 3+ Years ago[1] and moving everything to Rust[2], with the language simply being too unstable and unsafe in my eyes, despite my love for comptime and people arguing that Bun a
29.
by j-pb 4mo ago
Very cool approach! We build something super similar, also going for content addressed storage and compare&swap as fundamental primitives. Also commit dag based, but we also wrote this whole knowledge graph / triple-store CRDT data
30.
by j-pb 5mo ago
Fair, it actually started out in JS, moved to Deno, then Zig and ended in Rust. If I ever find the time I'd like to back port what I have now, up the chain. It is supposed to be a RDF replacement so it will eventually have to happen, b

More ›