Last I read you don’t even get an App Store (!) to install third-party apps on the device, so I’m not sure if that’s possible.
Interesting! Have you looked into variable-axis monospaced fonts at all for this project? https://monaspace.githubnext.com/
Exactly, it’d make the press’ work easier, and maybe enable more people or entities to look into this topic over all, too.
Good point, but do these settings need the latest M-series chip then, too?
Yes, of course, but couldn’t you have/make one in WASM too? (I have no clue how memory is allocated there)
(Not sure if this could be a genuine remark or is more of a joke, but) Why not vibecode a Go -> WASM cross-compiler, then?
> on par with deepseek atleast For chatbot-style usage maybe, can’t really attest to that; for long-running, long-context “agentic” work including planning, within a harness etc., definitely not (yet). > What I would…
> nonfiction is wilder than fiction Among other reasons, due to the very much nonfictional consequences you’d expect in a “real world” story like this: Had Graw been convicted/stripped of his titles, basically anyone…
Interesting. Doesn’t that shift the problem towards how you ensure alignment of the interpreter models, then?
The alternative is, I’m guessing, probably any other hosted git + CI/CD service that doesn’t offer a free tier to zillions of users (human and LLM alike) so that their computing capacity isn’t in a constant arms race…
Yeah, my first thought as well. How are they supposed to improve and iterate on this model?
Ah, much better, then – I mean it’s not like the US is already in debt or anything.
I wonder how the Helium [0] project solves this, since they also work off of an opinionated-ish Chromium base. [0] https://helium.computer
> normies The tinfoil hat explanation for this is that everyone at Apple is using iCloud for everything, and thus would never encounter a bug with locally-hosted services.
True that (I just recently saw an ad again promoting the equivalent of “grade and explain tasks, get paid”, clearly for training material). But I should’ve clarified, I meant the scraping aspect of it, since more and…
> LLMs cannot be pre-trained on LLM outputs because LLM output diversity is low. And with ever-more LLM usage both in labs and the real world, the task of finding new/“creative” out-of-average-distribution material will…
> it sucks Exactly. From the page about the requirements for the app under 1) in your link: > System requirements > Android 12 or iOS 16 So one of four methods is, by default, already completely dependent on having an…
You bring up an excellent argument there, however, haven’t you thought about the money? I mean, who cares about the developing minds of the next generation when there’s money to be made involved?
> And I also wonder what the hiring process was like to hire someone with that mental frame. Remember, this is only the company responsible for potentially building and aligning AGI. No need to be setting high…
A good while ago, I also read about a company [0] using a legal framework for the company structure/charter, the Public Benefit Corporation (PBC), specifically to ensure how they want “best interests” to be defined.…
> people do feel some tension about staring at their phones all day That, plus the crowd who would want or like to hack on a device like this, or at least use a non-Apple app or service, is probably far outnumbered by…
Really? You guys have laws that mandate profit-maximization?
Column-staggered keyboards are cool, but I think having one puts you in the vast, vast minority of all computer users. > I don't see a single objective benefit for choosing hjkl on its own. “j looks like a down arrow”…
While I agree on the premise – (written*) lawyer activity being reduced to putting signatures under things – I’m not so sure the gilded part is how it’ll actually play out in practice. It might also become a race to the…
> hairy realtime concurrent system programming On the other hand, I’ve heard a direct quote from the longtime (25+ YoE) colleague of someone I know, that in their mostly embedded/some realtime work, the LLM code quality…
Last I read you don’t even get an App Store (!) to install third-party apps on the device, so I’m not sure if that’s possible.
Interesting! Have you looked into variable-axis monospaced fonts at all for this project? https://monaspace.githubnext.com/
Exactly, it’d make the press’ work easier, and maybe enable more people or entities to look into this topic over all, too.
Good point, but do these settings need the latest M-series chip then, too?
Yes, of course, but couldn’t you have/make one in WASM too? (I have no clue how memory is allocated there)
(Not sure if this could be a genuine remark or is more of a joke, but) Why not vibecode a Go -> WASM cross-compiler, then?
> on par with deepseek atleast For chatbot-style usage maybe, can’t really attest to that; for long-running, long-context “agentic” work including planning, within a harness etc., definitely not (yet). > What I would…
> nonfiction is wilder than fiction Among other reasons, due to the very much nonfictional consequences you’d expect in a “real world” story like this: Had Graw been convicted/stripped of his titles, basically anyone…
Interesting. Doesn’t that shift the problem towards how you ensure alignment of the interpreter models, then?
The alternative is, I’m guessing, probably any other hosted git + CI/CD service that doesn’t offer a free tier to zillions of users (human and LLM alike) so that their computing capacity isn’t in a constant arms race…
Yeah, my first thought as well. How are they supposed to improve and iterate on this model?
Ah, much better, then – I mean it’s not like the US is already in debt or anything.
I wonder how the Helium [0] project solves this, since they also work off of an opinionated-ish Chromium base. [0] https://helium.computer
> normies The tinfoil hat explanation for this is that everyone at Apple is using iCloud for everything, and thus would never encounter a bug with locally-hosted services.
True that (I just recently saw an ad again promoting the equivalent of “grade and explain tasks, get paid”, clearly for training material). But I should’ve clarified, I meant the scraping aspect of it, since more and…
> LLMs cannot be pre-trained on LLM outputs because LLM output diversity is low. And with ever-more LLM usage both in labs and the real world, the task of finding new/“creative” out-of-average-distribution material will…
> it sucks Exactly. From the page about the requirements for the app under 1) in your link: > System requirements > Android 12 or iOS 16 So one of four methods is, by default, already completely dependent on having an…
You bring up an excellent argument there, however, haven’t you thought about the money? I mean, who cares about the developing minds of the next generation when there’s money to be made involved?
> And I also wonder what the hiring process was like to hire someone with that mental frame. Remember, this is only the company responsible for potentially building and aligning AGI. No need to be setting high…
A good while ago, I also read about a company [0] using a legal framework for the company structure/charter, the Public Benefit Corporation (PBC), specifically to ensure how they want “best interests” to be defined.…
> people do feel some tension about staring at their phones all day That, plus the crowd who would want or like to hack on a device like this, or at least use a non-Apple app or service, is probably far outnumbered by…
Really? You guys have laws that mandate profit-maximization?
Column-staggered keyboards are cool, but I think having one puts you in the vast, vast minority of all computer users. > I don't see a single objective benefit for choosing hjkl on its own. “j looks like a down arrow”…
While I agree on the premise – (written*) lawyer activity being reduced to putting signatures under things – I’m not so sure the gilded part is how it’ll actually play out in practice. It might also become a race to the…
> hairy realtime concurrent system programming On the other hand, I’ve heard a direct quote from the longtime (25+ YoE) colleague of someone I know, that in their mostly embedded/some realtime work, the LLM code quality…