IshKebab

Thanks for not using AI to write this. So much more pleasant to read.

There is about 150 years of cognitive science experimentation in animals

Yeah that would be relevant if AI were an animal...

As I said, it's clearly intelligent, but a quite different intelligence to that shared by animals.

there’s no “intelligence” there BUT

There is clearly intelligence there. We have no way to recognise intelligence other than the appearance of intelligence and this very clearly displays that.

It's also quite clearly different to human intelligence in some notable ways, but not in any that preclude describing it as intelligent. At least for normal non-pedantic definitions of the word.

I think you misunderstood his (quite obvious IMO) point.

...without actively changing the code so that you start trusting a new key.

Yes but the problem is there's nothing stopping them from doing that. If Apple/Meta rely on the argument that they technically cannot read people's messages then they will eventually lose because it's simply not true.

Do you mean Supernote? How do they compare to the Remarkable tablets?

This is for epaper tablets, just tiny ones.

Remarkable looks like it has a pretty active hacking community: https://github.com/rehackable/awesome-remarkable#launchers

The Remarkable tablets are pretty big and have a large community.

ESP32-S3 is really too underpowered for larger e-ink devices.

You're suggesting they purposely put a backdoor into all their custom methods? Why?

I'm not sure what you mean by "custom methods", but I'm not saying they have bypassed the e2e encryption - I'm just saying that they technically could.

And as for why they would do that, they might get compelled by a government to do it secretly. As far as I know that hasn't happened yet but I see no reason it couldn't and it would take a whistleblower to find out.

Security wise it's amazing. If a company's admin cannot take your data it's excedingly hard for a hacker to do so.

I agree, it is the best option available. But Apple/Meta are technically lying when they say it's impossible for them to read your messages.

How would that work for closed source apps like iMessage and WhatsApp?

It's probably impossible to get reliable statistics about that given how both groups are trying to keep everything secret. But you can consider how many paedophiles there are vs how many horny teenagers there are. Based on that it would be extremely surprising if he was wrong.

The real question is what happens when a horny teenager sends another a nude. There definitely have been insane cases where they get stitched up for creating child porn. I don't know if that's the normal outcome today though.

Yes that's exactly his point. E2E is often sold as preventing the owners of the server from being able to read the messages at all, even if they are evil and misleading you.

That's obviously only the case if they aren't also the sole providers of the "ends".

I think you've imagined this faulty understanding. There are many mechanisms by which Apple could actually decrypt the data despite pinky promises not to. You listed one. There are others.

Yeah and it's surely in the training data by now. Long past time to stop.

None of them.

Beautifully written!

This is written by AI. Maybe whoever prompted it has a point but I wouldn't say AI can reach the heights of beautiful writing yet, and this definitely doesn't.

Very clearly. Even if you have no qualms with AI I don't understand why people think writing in this cliched style is ok.

On the other hand people thought it was fine to use cliched titles like "The unreasonable effectiveness of..", "... for fun and profit", "The rise and fall of ...", "All you need is ...", etc. etc.

So I guess there's just a large portion of writers that are immune to cliche? Pretty annoying anyway.

Yeah it makes sense for their use case. I meant it was an odd thing to have if you were using it as a general purpose language like Lua.

The submission is "Nativ: Run frontier open models locally on your Mac" and on the page it says "Nativ puts frontier intelligence on your desk."

That's not what people are normally referring to when they say "frontier models". It means the most capable models full stop. Not the most capable that you can run locally.

I wish there was something better than Lua and just as "obvious" a choice. Static types, real arrays, zero-based indexing, etc.

Closest thing I've seen so far is Roto: https://roto.docs.nlnetlabs.nl/en/stable/index.html

Basically everything about this looks great - Rust-style syntax, static type checking, sane type system, easily embedded. The only slightly odd things are that it has built in types for IP addresses, AS numbers, and `filtermap`. But I guess you can just ignore those if you don't need them.

and an honest account of the limitations.

Gotta be trolling :-D

It's written by Claude already so it probably wouldn't make any difference I guess?

Technically nothing, legally a lot more.

So many words to say so little... Don't bother reading this.

From the photos they mostly look a bit like light SSOA. Obviously some devs overdo it, but that's not SSOA's fault. And we know SSAO isn't physically based. Of course it's not going to perfectly match reality.

The issue isn't that they haven't added a load of modern features; it's that they never fixed the old janky broken features.

Yeah I'm also wondering what the point of Bun is when Deno already exists. Especially since it added Node compatibility. Genuine question, why should anyone use Bun if they're already using Deno?

Presumably they don't pay $500k anymore...