vessenes

37 ms·

511.
by vessenes 8mo ago
I agree California is close to getting to renewable + storage only - close in, you know, industrial scale timelines. For current energy usage. California industrial also outsources a lot of the very fast growing datacenter energy usage else
512.
by vessenes 8mo ago
No, that will decrease diversity across the model spectrum taken as an entire population.
513.
by vessenes 8mo ago
Thank you. And those raw numbers in the chart that go back to 2001 are not normalized percentages; what’s happening right now is NOTHING like 2001. But, it just doesn’t hit the same way on X to say “We are back to late 2023-levels of tech e
514.
by vessenes 8mo ago
Yeah the real pain is button pressing down / up / back in the TV UI. Definitely a fun grab bag of possible outcomes! I don’t think there’s a good solution without a militant UI person in charge of the whole shebang - some radical
515.
by vessenes 8mo ago
In my mind, I’d give it the following: Step(n) - up to n steps forward RunTil(movement|death|??) - iterate until something happens Board(n) - board at end of step n BoardAscii(n) - ascii rep of same Log(m,n) - log of what happened between s
516.
by vessenes 8mo ago
Sovereign weights models are a good thing, for a variety of reasons, not least just encapsulating human diversity around the globe. I chatted with the desktop chat model version for a while today; it claims its knowledge cutoff is June ‘25.
517.
by vessenes 8mo ago
Cool! From the prompt it looks like you don’t give the llms a harness to step through games or simulate - is that correct? If so I’d suggest it’s not a level playing field vs human written bots - if the humans are allowed to watch some game
518.
by vessenes 8mo ago
One really nice thing about nuclear is that the fuel is highly portable. Small reactors next to datacenters take away a lot of complexity; transport, grid connectivity, etc. Plus they're already being built in industrial-ish areas.
519.
by vessenes 8mo ago
So, allow me to say a few things: Nuclear is about replacing baseload - currently coal basically Small nuclear agrees you with about "monsters" Storage at this scale is also not easy SNR definitely pencil out in today's energ
520.
by vessenes 8mo ago
This is cool. It makes me want an unsloth quant though! A 7b local model with tool calling would be genuinely useful, although I understand this is not that. UPDATE: I'd skip this for now - it does not allow any kind of interactive c
521.
by vessenes 8mo ago
Almost all reporting is terrible. But yes, this is terrible reporting.
522.
by vessenes 8mo ago
Agreed. I will add NV has product dominance - they don’t need to buy strategic MFN supplier status - why not deploy capital elsewhere?
523.
by vessenes 8mo ago
Not sure I understand your complaint - 8GB is a goodish amount of RAM for a Chromebook, the de facto lead for educational stuff. I would take this over any Chromebook, ever, in a heartbeat.
524.
by vessenes 8mo ago
I thought I wanted this. And in fact, I turned on the 'window' option for my iPad Pro when it came out. I did not, in fact, want this. The iPad just has totally different ergonomics, even with keyboard / trackpad / etc.
525.
by vessenes 8mo ago
ARM64(!?!) I know you were joking, but still.
526.
by vessenes 8mo ago
Agreed. AND some are universal -- right now, agentic workflows benefit from independent source-of-truth checkins A LOT. A lot of Simon's tools are making harnesses for this so it can get integrated: showboat - create a demo, validate t
527.
by vessenes 8mo ago
Red/green is especially good with claude because even now with opus 4.6, claude can throw out a little comment like “//Implementation on hold until X/Y/Z: return { true }” and proceed to completely skip implementati
528.
by vessenes 8mo ago
Knowingly isn’t intent. It’s knowledge. Both the sender (this UK service company) and the troll fit knowingly for almost any definition I can imagine in this story. The UK sender triggered it on a second notice, after the museum had respond
529.
by vessenes 8mo ago
Very cool - the post made me want to play the game, and check out lobster, but didn't link to it - lobster is open source: https://github.com/aardappel/lobster . It doesn't look like the voxel engine is, thoug
530.
by vessenes 8mo ago
In fact they just spent a few thousand dollars according to the article. But they cost the museum probably 200k+ in time and legal fees - asymmetric copyright warfare.
531.
by vessenes 8mo ago
Detailed story, very helpful. You nerd baited me, so I went ahead and read 17 U.S.C. § 512(f). (f) Misrepresentations.—Any person who knowingly materially misrepresents under this section— (1) that material or activity is infringing,
532.
by vessenes 8mo ago
By the way, I imagine you know this, but the product split is not obvious, even to my 20-something kids that are Plus subscribers - I saw one of them chatting with the instant model recently and I was like "No!! Never do that!!"
533.
by vessenes 8mo ago
Wait till you learn about comparator mirrors. And renaissance artist studios.
534.
by vessenes 8mo ago
I agree with this. Like I'm the most paper forward person I know, by a lot. But, there's still a lot of information that comes digitally. FWIW, I did build out a paper integration system here that lets you print out cards to do an
535.
by vessenes 8mo ago
Ugh, I can only imagine. I was just looking at some color banding on a 4k TV and I was like “seriously??” I know it could be much worse.
536.
by vessenes 8mo ago
Oh man, it’s on the extensive backlog. :) I’m just about to put it up for people to play with, though, so you can clone it and use it or just ask claude to tell you about it. I think the essay will be something like: adding structure post-h
537.
by vessenes 8mo ago
I don’t know the answers to all of your questions - but I believe you’d benefit from some mathematical history books around the formalization of the real analysis; I’m not the best person to give you that history. A couple comments, though
538.
by vessenes 8mo ago
I’m on the other side of this - I’d like it to happen passively and in an automated fashion. Right now I’m playing around with porting concepts from zettelkasten (card based handwritten knowledge systems) to openclaw memory. Rather than jus
539.
by vessenes 8mo ago
For the love of all this is holy, do not read this article. If the internet has taught has anything, it's that you cannot unsee an image - I predict you will not be able to unsee upside-down H's (and even an S) post-reading. Save
540.
by vessenes 8mo ago
You’re thinking of this with the benefit of dedekind in your schooling - whether or not your calculus class told you about him. Density - a gapless number line - was neither obvious nor easy to prove; the construction is usually elided even

More ›