As such, it raises concerns that AI systems deployed in a real-world situation, say in a driverless car, could malfunction when presented with dynamic environments or tasks.

This is currently happening with driverless cars that use machine learning - so this goes beyond LLMs and is a general machine learning issue. Last time I checked, Waymo cars needed human intervention every six miles. These cars often times block each other, are confused by the simplest of obstacles, can’t reliably detect pedestrians, etc.

It’s best to use specialized tools for this. A knife this small is basically useless.

If you think the type of person who signs up for a Facebook product will flock over to the “real” fediverse the moment they are seeing ads (which they are most likely seeing everywhere else on the Internet, since this kind of low-information user is usually not even aware of the possibility of blocking ads), then I got a bridge to sell to you.

Also wars, future pandemics, any kind of global cooperation that depends on the White House not being a madhouse, which is a lot.

And make sure it comes with a seat belt cutter.

This paragraph is the most ChatGPT of paragraphs:

As Tesla continues to refine its products, it’s crucial to address these concerns and ensure the reliability of its vehicles. The abandoned Cybertruck incident serves as a reminder that even the most innovative technologies can encounter setbacks, emphasising the importance of rigorous testing and quality assurance.

After Elon started Elonning

I’ll steal that one. You can have “the Xührer” in return.

I’m guessing they’ll comply with India’s censorship demands in order to not get locked out of a large and potentially lucrative market. The fact that state governments can also demand ISP follow their own censorship rules might make things a bit more complicated:

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

The Model X is notorious for this. I know a doctor who just had to fix his Model X’s suspension for “thousands” (his words). He also complained that you can never ever reach anyone at Tesla, that parts are impossible to get and that you have to expect repair appointments to be pushed back at least a few times, often mere days beforehand.

At least the batteries seem to last forever, even on his very early Model S, even after hundreds of thousands of kilometers. There’s barely any degradation. Then again, this mirrors other EVs, even much cheaper ones without active cooling. EV battery packs usually outlast the cars they are in, which is why there’s a thriving market of second-hand batteries that are used for all sorts of applications, from converting normal cars into EVs to storing solar power at night.

This reminds me: In countries like Russia and China, it’s not unusual for police to just randomly stop people and search their phones, at which point even locally stored data isn’t safe anymore. This could happen in America as well.

Except that I know first-hand that German government institutions are already using this exact tool in order to make up for the chronic lack of translators. They are translating texts into languages they don’t speak, which means there’s no going over the output to correct for mistakes.

That’s a very good answer.

If I’m getting this right, this was a novel that you perhaps mentioned to your loved one, but a language barrier prevented them from reading it. They then suggested the use of an LLM to translate it, which you used as foundation to build upon. If I may ask, which story did you translate (it has to be good if you spent this much work on it) and which LLM did you use?

I can’t see anything wrong with this. I’ve used this kind of approach using all sorts of machine translation tools going back over 20 years (not for entire books though). Let the computer do its thing, then fix mistakes - but this was always noncommercial, private use for myself, friends and relatives, as well as the occasional friendly online community. Although, I’ve also done entirely manual work, with no machine translation at all in situations when I wanted the best possible quality or where complexity and nuance made anything else impossible - like with a long list of “whisper jokes” from Nazi Germany, subversive jokes that people told each other under the punishment of death that require a ton of context no translation tool could possibly have.

The point here is though that this is very different from a publisher doing this commercially - and you and I both know that these companies will not even allow for the bare minimum of time spent fixing mistakes made by the translation tools.

Did you inform your readers that most of the translation was done by the LLM?

Technically it does, but not locally in the age of national governments. Before you’re saying it, the moment it stops being a local movement, it would work even less and lead to the organized repression I mentioned. To support my point, see how harsh government reaction has been to activists merely gluing themselves to the street (not to mention, how most people were happy about this crackdown).

And no, I doubt “The Revolution” that magically solves all of our problems (unlike most revolutions) will be started by anti-AV riots.

How would self-driving taxis do this any better than taxis that already exist and aren’t relying on large tech corporations?

Singular acts of violence don’t work, organized violence doesn’t work either and will only lead to organized repression in response. The actual solution is to elect local representatives who are willing to prevent the nightmare scenario from the video from happening.

If you want to see a real-world example of this: Toppling over rental e-scooters didn’t get them removed from cities, but petitioning municipal governments to ban them did.

This is precisely the kind of niche, but vital use case that even places that have otherwise already completely banned cars (like certain islands) allow cars for. Nobody will ever take this away.

What matters is that TI has an effective monopoly on scientific and graphic calculators in the US in particular, which means that it’s the platform that matters to most. It’s irrelevant whether or not some alternative is better. It’s also extremely widely supported by software and tutorials.

I can’t imagine there being many comparable jobs for aerospace engineers in India.

There is one crucial difference between image editing software like Photshop and Gimp vs. 3D software suites like Maya and Blender: My hypothesis is (and feel free to pick this apart) that you can totally teach yourself to use the former rather competently without any outside help, not even documentation and tutorials, but I would argue that this is nearly impossible with the latter due to their far greater complexity. This in turn means that people will look up guides and tutorials and learn the idiosyncratic UI patterns that way, which is why Blender with its extremely nonstandard controls managed to gain a foothold far beyond the broke hobbyist sphere.

Thanks for gently letting me down on RISC OS. I guessed that there wasn’t much going on with it, but I wanted to be sure.