Ironic that it's OpenAI that stopped the trend.

Hey, we need protecting from AI, only one company can get this right.

AI genius discover brute forcing... what a time to be alive. /s

Like... bro that's THE foundation of CS. That's the principle of The bomb in Turing's time. One can still marvel at it but it's been with us since the beginning.

It's quite a safe case if you maintain provenance because there is a ground truth to compare to, namely the untranslated paper.

SteamDeck, System76, Pine64, Slimbook, Tuxedo, etc there are PLENTY of Linux devices to buy in all form factors.

Source : I'm using at least 2 of these and use Linux on my desktop daily, for years. Spent maximum 15min total caring about drivers and yes I do also game.

> Doing ANC properly is very complicated. It’s much harder than taking the input from a microphone, inverting the signal, and feeding it into the output. There’s a lot of computation that needs to be done continuously.

Neat, any recommended reading on the topic?

Wrong title, if it's "File System Access API (still Chrome-only as far as I can tell)" then it should read "A browser is the sandbox".

At the risk of sounding obvious :

- Chrome (and Chromium) is a product made and driven by one of the largest advertising company (Alphabet, formally Google) as a strategical tool for its business model

- Chrome is one browser among many, it is not a de facto "standard" just because it is very popular. The fact that there are a LOT of people unable to use it (iOS users) even if they wanted to proves the point.

It's quite important not to amalgamate some experimental features put in place by some vendors (yes, even the most popular ones) as "the browser".

I stand by a policy that if a feature in one of my projects can only be implemented in Chrome, it's better not to add the feature at all; the same is true for features which would be exclusive to Firefox. Giving users of a specific browser a superior experience encourages a dangerous browser monoculture.

Not writing the feature makes sense, but pushing Firefox and Safari to add support would be pro-social if you're up for it. The most common reason for browsers not to add support is something like "this can be done in other ways, and it has maintainability/security/bloat downsides". Running into a feature you can't build is evidence on the "this can be done in other ways" question (but of course the other downsides could still be big enough that it's not worth doing).

I say the following as a firefox+ubo user:

There are many useful things that can only be implemented for Chromium: things like the filesystem API mentioned in this post, the USB devices API used to implement various microcontroller flashing tools, etc. Users can have multiple browsers installed, and I often use Chromium as essentially a sandboxed program runtime.

SOME users can have multiple browsers installed. Some can absolutely not. In fact, 1.6 billion users can only have one installed and it's not Chrome or Chromium based.

Assuming you're talking about iOS: and their OS won't let them install your app to manage files or flash microcontrollers anyway. It's not your problem when they choose an actively hostile platform.

Firefox is only a few percent market share. You are hiring your users for not improving their user experience because it's not compatible with one of the a web browsers on a few percent of people's computers.

Chrome add these features because they are responding to the demands of web developers. It's not web developers fault if firefox can't or refuses to provide apis that are being asked for.

Mozilla could ask Claude to implement the filesystem api today and ship it to everyone tomorrow if they wanted to. They are holding their own browser back, don't let them also hold your website back. In regards to browser monoculture there are many browsers built on top of the open source Blink that are not controlled by Google such as Edge, Brave, and Opera just to name a few of the many.

Shouldn't the EU then escrow keys?

Was using them just this morning. I've been using them since they are out. Great device but battery is quite limited, ~2hrs top with ANC on.

But then again what won't? Non tech stocks?

Gold stocks, basic materials, MSCI world and emerging market indexes. Look at their prices and see how very smart people are positioning their money.

Yes, but that’s not because of AI, it’s because of the Orange Wizard of Tarrifs.

Isn't it included in "Non tech stocks?"?

Yes, lots of other companies would be affected to a greater or lesser extent (even non-tech stocks), but specifically any company that relies on manufacturing all their product in Taiwan will be affected most of all.

Industrial military complex and government contractors.

Don't they also depend on chips for a lot of components?

Probably a lot more from TI and Intel than Taiwan.

I'd be curious how many of the design and verification (using computer vision) tools used at TI and Intel rely on on farms of stock GPUs thus chips still made in Taiwan. They might have in house chips just for such part of their workflows though, any insight appreciated.

Jets, tanks, drones and data centers for intelligence services, even design, are full of electronics but what's the share of those not made in Taiwan?

They’re not cutting edge chips, they’re made all over, including the US, Korea, and China.

So the share is 0%? I have a hard time imagining that but OK, any source I can read on the topic please?

The whole economy will crash. Probably won't be due to China invading Taiwan though. More likely because the president decided to delete their country's world reserve currency status (which is another word for a trade deficit).

Doubt it, code will be generated to pass tests, not the intent behind the tests.

A million times, this. Sometimes they luck into the intent, but much more frequently they end up in a ball of mud that just happens to pass the tests.

"8 unit tests? Great, I'll code up 8 branches so all your tests pass!" Of course that neglects the fact that there's now actually 2^8 paths through your code.

What makes you think the next generation models won't be explicitly trained to prevent this, or any other pitfall or best practice as the low hanging fruit fall one by one?

I think we agree - getting the llms to understand your intent is the hard part, at the very least you need well specified tests.

Perhaps more advanced llms + specifications + better tests.

if you can steer an LLM to write an application based on what you want, you can steer an LLM to write the tests you want. Some people will be better at getting the LLM to write tests, but it's only going to get easier and easier