I use the assistant, because it has so many models to choose from. I hope they can make a mobile app for it in the future

I’ve found Kagi has been good enough to justify the subscription price. I like that I can block certain sites, pin and promote others. It has some neat AI features but they only activate when requested and never replace actual results.

Yeah actually that makes more sense than what I originally said. The US is one of the main buyers of gray-market zero day bugs, way cheaper and less risk than trying to covertly implement bugs.

That would be too obvious and thus ineffective. In reality it is more likely that they have inserted bugs into various open source software covertly, like we saw with xz.

Not necessarily. If you trust the code running on your device then there is no backdoor they could install on a server that would break e2ee. They would have to backdoor the client where the keys are.

Yep. On kbin I think any user can too.

This is how I would describe my experience. Sometimes it’s crunch time and most of the time it’s fuck around time. After crunch time I always throw a tantrum about how if we only bothered with planning we could largely avoid it.

If it ends up being ruled that training an LLM is fair use so long as the LLM doesn’t reproduce the works it is trained on verbatim, then licensing becomes irrelevant.

If something like that were to work, a lot of effort would need to be put into minimizing the UI friction. I could see something like: uploaders add topic tags to their videos, and an AI runs in the background to generate and apply new tags based on the content (most people would not understand how to properly tag content). An AI would also be used to create a graph of related tags, where similar or closely related tags are nodes joined by an edge. Then, on first login the user is prompted to pick some tags to start with. Over time, the client uses the adjacent tag graph to fine-tune users’ tags, on device. The idea here is that we could get a decent algorithm that can recommend new stuff based on what the user watches, but keep that data processing of user-specific content local. Then, the client would also have an option the user could enable that would contribute their client’s tag information back to the global tag graph, improving the global tag graph for everybody. This data could also be combined with other users data at the instance level to somewhat anonymize the data, assuming it is a large multi-user instance. If you were to host a single user instance, you’d probably not want to contribute to the global tag graph unless you’re ok with your tag preferences being public.

It’s a bit tricky but I think a privacy preserving algorithm is possible. Simply put, the more data available, the better an algorithm can be.

I think the easy discoverability on these platforms is part of what makes them so popular. Using TikTok or similar, a user typically wants to be shown new things, it maintains a sense of novelty that keeps users constantly engaged. Having to do this manually would be a huge negative.

The algorithms are what makes these services. Most interactions aren’t searching and selecting something specific or intentional, they’re just opening a fire hose and expecting the algorithm to pick content they find entertaining for them. It requires the algorithm to have a lot of information, both about the specific user, and about similar users.

I use it to describe a variety of things, but usually it’s related to servers not being able to handle load rather than an outright crash, but I’m not strict about it. Laos balancer failures could be it, could also just be that something was really I efficient but wasn’t noticed until it went into production.

Because even if you’re not working, you’ll probably think about problems overnight

Where do I sign up?

Antivirus programs are way too inaccurate to be used authoritatively, especially for developers. It’s not uncommon that some virus will use a well-known open source library or packaging tool, and then the antivirus decides that any binary with that same library or stub from that packaging tool must also be a virus. When your program depends on it, if you can’t turn the AV off or make an exception, you’re just fucked. Also, programming is an iterative process. Make a small change, test, repeat. Requiring that developers upload and wait for a scan from some third party for software that they compiled locally and have no intent to distribute is a giant waste of everybody’s time, especially the developer’s. It’s a huge drag on productivity for the sake of bureaucracy.

As a silver lining, at least it’s terrible at it

Despite that, some languages make it easier to be wrong than others.

black shittyformatting.py

As far as free software goes, how does running free software on your own server that you allow others to communicate with using established standard protocols violate your freedom? Not saying you shouldn’t be able to be selective about federation, but why would Facebook specifically being one of the peers violate your freedom?