Social media in its current form is being used to unravel whole societies.
"The tyrant will always find a pretext for his tyrany, and it is useless for the innocent to try by reasoning to get justice, when the oppressor intends to be unjust."
I know it was just a convenient pretext for a learning journey, but do not come away from this thinking llama.cpp needs to be compiled on Windows before use. The GitHib project has a cornucopia of pre-built artifacts to use.
Just say I'm a sender who wants to send an email to someone on this system, and who also holds zero BTC and is justifiably deeply skeptical of anyone pushing it, exactly what steps would I need to take with this system vs hitting "Send" from Outlook?
Something this "shape" has been coalescencing since the first tool calls were done. To draw another Star Trek parallel, this reformulation is what Brent Spiner is during the little stares and pauses made before answering a complicated but constrained problems on the show. Onward!
The numbers matter. The thermal budget a satellite is an tightly controlled thing. Large modern ones are in the order of a few to a couple of 10s of kilowatts, so something like a few to several low 10s of modern GPU compute power. Even with thousands of yet to be designed or launched satellites, it's going to have trouble competing with even a single current DC, plus it is in SAPCE for some reason, so everything is more expensive for lots of reasons.
I don't think that was the point being made by GearSkeptic, the video creator. It was a demonstration to the lay person who may not be familiar with what 5W "looked like".
I binged ST:NG before it went away again on Netflix. The more I heard from Data, the more he sounded like where Ai should be heading: Quick, thorough reasoning but followed by explicit, tagged verification from external ground truth.
There needs to be a more meta, layered approach to reason. Different personalities viewing the output with different hats on: "That's a bold claim, champ. Search required." But I guess the current real-time, interactive nature of these systems makes it difficult to justify.
If sabotage it will be plain as day to a trained eye. I await the report. That break could also be explained by the rail heading away in that photo snapping at that point because the train pushed it out, noting the rail has rotated 90 degrees clockwise -- something did that work, and it was probably the train going out and over. I'm not a rail tie expert (nor is anyone likely to be on HN) so I don't know if this is an unusual failure mode. But there was a line change point intersection immediately south of the crash. My money is there was a fault (accidental or deliberate) there, not at this snapping point.
Assuming no foul play, it's going to be a Points Failure, isn't it? Like Potters Bar (2002) where most of the train makes it through, but rattles/breaks some weak point that was just holding on, and the last carriages change tracks. But at 250mph. Shocking stuff.
I tried to parse the history and it goes back to 2021 and apparently touches upon pronouns, NFTs, online unmasking and various online battles around the same. Sigh. Can we put this HN thread out of its misery, please?
My understanding of JavaScript is cursory, but my reading of that webpage is the UI is just smoke and mirrors, and it is just waiting for the whole thing to be processed in a single remote API call to some back-end system. If the back-end is down, it will always stop at 90%. The crawling progress bar is fake with canned messages updated with Math.Random() delays. Gives you something to look at, I guess, but seems a little misleading. Might be wrong ...
Indeed. I've always had trouble picturing how to efficiently "unmix the cake" too. CO2 is rare and throughout the whole atmospheric column. What kind of concentration gradient can you get going to meaningfully pull it out from everywhere in human timescales? (Sorry if this nerd-snipes someone stronger with calculus than me.)
Not quibbling at all, but I recall some discussion somewhere saying that the history of the Squeak impl itself (not the name) traces back via saved base images to the original Smalltalk implementations, including via customs at-rest transformation tools when backwards incompatible changed where made in the primordial days. Base images, at least back when I was toying with Squeak, where never rebuilt from scratch, just modified, transformed etc. In some sense, at least for the image, they were decades old.
This port of Fuzix to the Pico was done by David Given. In 2021 he also screen-recorded and narrated basically his entire effort porting Fuzix to similarly-sized ESP8266. Really very interesting if you are in to that sort of thing!
I have a colleague whom refers to the emails he sends as 'letters'. It was strange at first, but it also implies some effort being taken. I read the 'essays' terminology the same way. If it isn't some textual version of habitual food/travel/life logging, and has some cohesive topic, it's an essay.
If you replace JSON with XML in this model it is exactly what the "document store" databases from the 90s and 00s were doing -- parsing at insert and update time, then touching only indexes at query time. It is indeed cool that sqlite does this out of the box.
I can't comprehend how that's even an issue. Like it's the sort of thing you might read in an old bug report online and go "wow, that must have been an awkward few days for everyone" but to hear that it is "normal"? Wild. Utterly unacceptable.
I agree. Park the handle with a polite "You will never find a more wretched hive of scum and villainy" note and a suggested list of other places for more fruitful discussion.
I still tidy my house using the Settlers resource movement algorithm, moving things closer towards where they need to be even if they don't go all the way to their final destination.
Another worthy mention in this space is Linus Åkesson's dialog language[1]. From its description:
Dialog is a domain-specific language for creating works of interactive fiction. It is heavily inspired by Inform 7 (Graham Nelson et al. 2006) and Prolog (Alain Colmerauer et al. 1972).
An optimizing compiler, dialogc, translates high-level Dialog code into Z-code, a platform-independent runtime format originally created by Infocom in 1979.
Development seems dormant at the moment, but it feels more like Inform 7 'done right' to me. If my brain was a little bigger and calmer I'd be all over it. It has excellent documentation too. Very portable -- I compiled it locally under Termux on my phone with nothing but Clang.
Small meta observation: Gemini Flash 2.5 summarises this discussion as a problem with how ML was used and it's flaws, whereas Gemini Pro 2.5 emphasizes that it mainly a cultural communication problem. Both topics are here, but "fast Vs slow thought" sees different things.
Haha but seriously, Trump is just starting to ramp up full kleptocracy mode. Each tariff change is going to be associated with billions in trades made with foreknowledge of the move. His robber baron friends will fund him and his regime forever. They can do whatever they want now. We might as well tear down the White House and replace it with a Putin style gilded palace for Oligarchs. Oh wait.
ID checks, driven by prudishness, are an absolute gift to the big social media companies. They're the only entities whom (a) already know the check's answers, and (b) have the resources to keep hackers largely at bay.
I am not surprised these laws are landing with such little resistence.