I strongly disagree. Every response longer than one line of code or longer than 1-2 sentences on a non-trivial task, has at least inaccuracies or mistakes. When working with something AI has created, I had to go in and fix it manually in 100% of those cases. That’s why I limit the use of AI to only type-ahead suggestions, as I can easily verify them and don’t waste more time than creating the result manually.

Just a few posts down my lemmy stream:

https://literature.cafe/post/31346850

That’s what some people claim. If I remember correctly, it was determined that these large scale prestigious events are a net negative.

Depending on the source DNA seems to have a half life of 500-1500 years. This source calculates with slightly over 500 years. Permafrost samples may be more on the upper end of the scale but the message would be similar:

A study of DNA extracted from the leg bones of extinct moa birds in New Zealand found that the half-life of DNA is 521 years. So every 1,000 years, 75 per cent of the genetic information is lost. After 6.8 million years, every single base pair is gone.

https://www.sciencefocus.com/the-human-body/how-long-does-dna-last

If you say it like that, I’ll happily abandon science to go velocitaptor jousting with you!

Even under the best preservation conditions, there is an upper boundary of 0.4–1.5 million years for a sample to contain sufficient DNA for sequencing technologies.

https://en.wikipedia.org/wiki/Ancient_DNA

They are both fans of Anker.

a young, attractive woman

with strong teeth

Waaaait… double blind?

Or ASS in German 😁

When I used LaTeX many years ago, I loved this. Or was practically impossible for to overlook that I had opened a draft version that I didn’t want to send.

The ones I’ve handled and bought are more robust than most other glassware. They can break if dropped on hard floor, but that’s normal for anything out of glass.

Just install a panic button. The microphone is not reasonable since there are way less intrusive options available.

No, that’s ridiculous.

This Regulation does not apply to the processing of personal data: […] © by a natural person in the course of a purely personal or household activity;

As is stated, the call is processed locally in the user’s device. If that holds true, there is no recording and no third party processing going on. Your point does not make sense.

That’s a real world issue. AIs training on each other’s output and devolving because of it. There will be a point when vendors infringing on user content and training their AIs with it will leave them worse off.

It’s easy to train a model to do exactly what you want and have the seeming “personality” that you want. It’s just incredibly expensive. You need to vet and filter everything that you use to train the model. That’s a lot of person hours, days, years. The only reason the models act the way they do is because of the data that went in to train them. If you try and fit the model after the fact, it will always be imperfect and more or less easy to break out of those restrictions.

Also, revolt self hosting is broken. The web call functionality (WebRTC) is being rewritten but that effort is stale and out of the box it simply does not work. There is no real documentation about this either. It just won’t work and you need to invest a lot of effort to figure out why. The moment self hosting properly works, I’ll give it another shot. Not being able to connect without a fat client is a show stopper for me. There’s no way I can get enough traction for my groups if the barrier to switch is higher than a sheet of paper.

When self hosting all the shortcomings you mentioned are perfectly acceptable for me.

If you use much of the software that is included in the support package, then the price seems reasonable. No way you could get the same price if you went to each provider individually. If all you use is bare bones openshift, then you’re right.

Don’t shoot the messenger. The regulations are pretty draconic. I have to ensure the training for that every year.