You’re still putting words in my mouth.

I never said they weren’t stealing the data

I didn’t comment on that at all, because it’s not relevant to the point I was actually making, which is that people treating the output of an LLM as if it were derived from any factual source at all is really problematic, because it isn’t.

You’re putting words in my mouth, and inventing arguments I never made.

I didn’t say anything about whether the training data is stolen or not. I also didn’t say a single word about intelligence, or originality.

I haven’t been tricked into using one piece of language over another, I’m a software engineer and know enough about how these systems actually work to reach my own conclusions.

There is not a database tucked away in the LLM anywhere which you could search through and find the phrases which it was trained on, it simply doesn’t exist.

That isn’t to say it’s completely impossible for an LLM to spit out something which formed part of the training data, but it’s pretty rare. 99% of what it generates doesn’t come from anywhere in particular, and you wouldn’t find it in any of the sources which were fed to the model in training.

That simply isn’t true. There’s nothing in common between an LLM and a search engine, except insofar as the people developing the LLM had access to search engines, and may have used them during their data gathering efforts for training data

Except these AI systems aren’t search engines, and people treating them like they are is really dangerous

I’ve heard it’s actually really difficult, because a good translator doesn’t do it literally.

They’re supposed to say something which gets the same meaning across, which often isnt what you’d get just by translating each word.

That leaves people translating trump with a problem: you can’t generally turn his long rambling speeches into something with a clearly understandable meaning without putting words into his mouth, or summarising so aggressively that you’d only say a couple of sentences for every few minutes of speech

No, it couldn’t. That’s pure misinformation.

Kessler syndrome is only a possibility in orbits high enough that atmospheric drag is negligible. Starlink, by design, is at an altitude where the atmosphere is still thick enough to bring any debris or old satellites down to earth in a timely fashion rather than building up like Kessler syndrome requires. (To be clear, the air is still so thin that you’d need sensitive instruments to detect it at all. It’s just enough to produce a tiny amount of drag, which adds up over weeks or months to lower the debris’ orbit so that it meets thicker air)

There are plenty of perfectly legitimate objections you can raise to starlink without resorting to Kessler syndrome

For an emergency ascent, they’d probably have dropped more than two. They also probably wouldn’t have taken the time to type a message to the surface if it were going wrong that quickly.

It seems more likely to me that they were controlling their rare of descent. I’d expect them to lose a little buoyancy as the vessel compresses, so it seems reasonable that they’d drop the occasional weight as they descend.

It’s unlikely to cause anything to outright fail, but it will certainly be creating bottlenecks and inefficiencies

They started launching in 2019, according to a quick look at Wikipedia. They told the general public (and regulatory agencies, I think) that the lifetime of the satellites was on the order of 5 years. The plan was to replace them frequently enough to maintain the constellation with that kind of service life (i.e. to launch the whole constellation worth of satellites every 5 years)

Now, here we are 4 years later. It’s not terribly surprising if some of the early satellites are starting to reach the end of their lives.

It’s going to be very expensive for them, but not an unexpected cost. This is the reason they’re so keen to start launching them on Starship

Personally, I’m alarmed that a bear only merits closing half. Did the guests in the other half do something to make themselves unpopular?