It’s not even copyright, they’re suing for using things they patented, but their patents are extremely general. I kid you not, they have a patent for MOUNTING CREATURES, something hundreds of games have done.

Abstract: In an example of a game program, a ground boarding target object or an air boarding target objects is selected by a selection operation, and a player character is caused to board the selected boarding target object. If the player character aboard the air boarding target object moves toward the ground player character automatically changed to the state where the player character is aboard the ground boarding target object, and brought into the state where the player character can move on the ground.

I’m no lawyer so I can’t tell you how well this would hold up in court but it’s ridiculous. See more: https://patents.justia.com/assignee/the-pokemon-company

It’s weird that they’re adding E2EE on voice but not in private text DMs, which is probably everybody’s biggest concern when it comes to security on Discord. Better than nothing I guess.

What an overreaction. Also FYI the free file limit was 8MB two years ago before they raised the limit to 25MB, so if anything it’s gotten better.

I guess you missed the part where he said “Oh you said X but you’re actually implying Y? Did you mean Y? Please confirm you actually meant Y.”

Can you confirm that you think with L3, the result would look completely opposite and the summaries of the AI would always beat the human summaries? Because it sounds like you are implying that.

Lemmy users try not to make a strawman argument (impossible challenge)

No, that’s not what I said, and not even close to what I was implying. If Llama 2 scored a 47% then 3.1 would score significantly better, easily over 60% at least. No doubt humans can be better at summarizing but A) It needs someone that’s very familiar with the work and has great English skills and B) It needs a lot of time and effort.

The claim was never that AI can summarize better than people, it was that it can do it in 10 seconds and the result would be “good enough”. People are already doing AI summaries of longer articles without much complaints.

I read the article. I’m aware it’s an older study. Point still stands.

The most promising model, Meta’s open source model Llama2-70B, was prompted to summarise the submissions

Llama 2 is insanely outdated and significantly worse than Llama3.1, so this article doesn’t mean much.

It’s pretty well known at this point, they lied about their initial goals and nothing they’ve done in ages has been “open”. It’s all been closed research that they won’t share because they don’t want anyone else to catch up.

Also they had a lot of friction with their AI safety department because caring about safety would slow down business.

From the article

The Justice considers that the lack of moderation, cooperation with the forces of law and order and the tools offered by Telegram (disposable number, cryptocurrencies, etc.) makes it complicit in drug trafficking, paedo criminal offences and fraud.

But a lot of people are speculating they just fabricated claims to arrest him because Telegram is russian.

Even more annoying is how many people in the Linux community often recommend distros that are terrible for beginners. People who constantly try to tell newbies to download base debian or arch should be removed from the conversation instantly.

Pretty much sums it up. JPEGXL could’ve been the standard by now if Google would stop kneecapping it in favor of its own tech, now we’re stuck in an awkward position where neither of them are getting as much traction because nobody can decide on which to focus on.

Also, while Safari does support AVIF, there are some features it doesn’t support like moving images, so we have to wait on that too… AVIF isn’t bad, but it doesn’t matter if it takes another 5+ years to get global support for a new image format…

He’s the co-founder of a wildly successful company though, not just a regular employee. It’s definitely not normal for him to bail and work for the rival company.

Co-founder of OpenAI left the company and went to a rival AI company that currently has the best model. This could signify that he thinks OpenAI is a sinking ship.

Translating entire codebases with LLMs? What could POSSIBLY go wrong?

I also don’t see how it would ever be possible to directly translate C to Rust. They’re so fundamentally different that things are bound to not work the same.

Not all cookies are harmful and some websites don’t work properly without cookies. Having cookies off by default also usually means user preferences wouldn’t be saved when you leave and return to a website.

If this thing needs to be present, the option should be there to toggle on, not off.

This is my takeaway in general. The idea of this sounds fine, but the fact that they opted everyone into this experiment is really stupid considering a huge chunk of people use Firefox are privacy-conscious and care deeply about this stuff.

AI was a promise more than anything. When ChatGPT came out, all the AI companies and startups promised exponential improvements that will chaaangeee the woooooorrlllddd

Two years later it’s becoming insanely clear they hit a wall and there isn’t going to be much change unless someone makes a miraculous discovery. All of that money was dumped in to just make bigger models that are 0.1% better than the last one. I’m honestly surprised the bubble hasn’t popped yet, it’s obvious we’re going nowhere with this.

Still interesting to know just how expensive that thing is and what it takes to run it

shocking users

Are they shocked though? It’s getting common for piracy sites to disappear or move to different domains, there are always alternatives

People said the same thing when the original Vision Pro came out. Aside from some rich people flexing that they own one, I haven’t heard anything about it after one week of release.