Given the ease of implantation of end to end encryption now, it’s a reasonable assumption that anything not e2ee is being data mined. E2ee has extensive security benefits, for example even if your data is dumped the info is still useless. So, there has to be a compelling reason to not use it.

People haven’t really changed. As always, power corrupts. When the rewards are great enough, it seems people are often enough willing to compromise their integrity.

Key detail in the actual memo is that they’re not using just an LLM. “Wallach anticipates proposals that include novel combinations of software analysis, such as static and dynamic analysis, and large language models.”

They also are clearly aware of scope limitations. They explicitly call out some software, like entire kernels or pointer arithmetic heavy code, as being out of scope. They also seem to not anticipate 100% automation.

So with context, they seem open to any solutions to “how can we convert legacy C to Rust.” Obviously LLMs and machine learning are attractive avenues of investigation, current models are demonstrably able to write some valid Rust and transliterate some code. I use them, they work more often than not for simpler tasks.

TL;DR: they want to accelerate converting C to Rust. LLMs and machine learning are some techniques they’re investigating as components.

I have LTS and zen kernels installed in addition to the default Arch one, that should prevent this yes?

What do you mean by “this stuff?” Machine learning models are a fundamental part of spam prevention, have been for years. The concept is just flipping it around for use by the individual, not the platform.

If by reliably you mean 99% certainty of one particular review, yeah I wouldn’t believe it either. 95% confidence interval of what proportion of a given page’s reviews are bots, now that’s plausible. If a human can tell if a review was botted you can certainly train a model to do so as well.

Cool it with the universal AI hate. There are many kinds of AI, detecting fake reviews is a totally reasonable and useful case.

KDE Connect and Syncthing do the trick for most stuff. For all else, all hail the USB C M.2 NVME enclosure.

Well this is a tremendous step in the wrong direction. The economic problem is the ad supported model in the first place, no matter how it’s run. This is the same thing Google does, they keep user data to themselves and sell the ad placement. So now Mozilla has the same economic incentives as Google. Unfathomably bad move.

If you read carefully this is actually very similar to the Steam news. I doubt Valve or GOG care, but generally the games are “sold” by the publisher as non transferable licenses for you to play them. So the part that matters isn’t up to them.

Have you tried recent models? They’re not perfect no, but they can usually get you most of the way there if not all the way. If you know how to structure the problem and prompt, granted.

Them using Google indexes anonymously isn’t intending to solve the problem you think it is. It’s more about incentive structures. Google’s “free” search optimizes for ad revenue now. The API access doesn’t as much, and Kagi certainly doesn’t have an ad incentive. So privacy is a nice bonus, but the real benefit is a customer serving incentive structure.

Funny how the DOS equivalent of ls is dir, so before the GUI folder metaphor.

A major caveat I’ve noticed some people misunderstand: it’s corporate CLAs that are problematic. The Apache Foundation also requires contributors sign a CLA, but it’s to provide a legal fail safe and a way to update to say Apache 3.0 if need be one day. Apache’s non profit, open source mission aligns with respecting the rights of contributors and the community. Corporations, on the other hand, not so much.

Codeberg is run off of donations, they have no service contract revenue. Nobody, much less a volunteer, wants to commit to a 5 or 10 year service plan like that, it’s not sustainable for a small project from a non profit.

CLAs can be abusive, but not necessarily. Apache Foundation contributors need to sign CLAs, which essentially codify in contract form the terms of the Apache 2.0 license. It’s a precaution, in case some jurisdiction doesn’t uphold the passive licensing scheme used otherwise. There’s also a relicensing clause, but that’s restricted to keeping in spirit, they can’t close the source.

The idea that “it’s ok cause we’d do the same” is ridiculous. There is no comparison: China is an authoritarian government and the parent company is practically an arm of the state. There are legitimate criticisms of American tech companies obviously, but they’re ultimately subject to the market and democratic governments. We shouldn’t be doing any business with authoritarians in the first place, much less inviting them to control a significant social media app in the guise of a legitimate business.

It runs great now. Most importantly, it supports extensions like ublock.

So this is probably another example of Google using too blunt of instruments for AI. LLMs are very suggestible and leading questions can severely bias responses. Most people using them without knowing a lot about the field will ask “bad” questions. So it likely has instructions to avoid “which is better” and instead provide pros and cons for the user to consider themselves.

Edit: I don’t mean to excuse, just explain. If anything, the implication is that Google rushed it out after attempting to slap bandaids on serious problems. OpenAI and Anthropic, for example, have talked about how alignment training and human adjustment takes a majority of the development time. Since Google is in a self described emergency mode, cutting that process short seems a likely explanation.

Arch for stuff I have physical access to. Nothing’s ever gone wrong, so it’s worth it for the immediate updates and consistency with my other systems. For VPS I use Debian though, occasionally the unstable/Sid branch if I really need the latest updates. There are almost always Debian images available on a VPS.