llama.cpp isn't configured for this mode, there are forks that can work though, this is a good explanation of what's different https://www.youtube.com/watch?v=bcGO7xre46o

If there is a BBC documentary to go with this, I'd love to see it.

This is supported in AOSP now. No it isn't. GrapheneOS is directly based on AOSP, so it includes everything AOSP does. GOS devs have made it clear they want you to use gestures, and they only include the navbar at all because it's still in AOSP. Inverted navbar is in wide a range of other AOSP custom ROMs, because those developers chose to incorporate it. But it isn't in AOSP by default. This doesn’t seem like a mature thing to say just because you want to flip buttons and are annoyed that the privacy and security-focused project didn’t add a feature to flip buttons. It's an accurate take on their general attitude to polite requests on their github. And the reason the requests are so excessively polite is because people are treading on eggshells due to their diva reputation - albeit, this reputation is somewhat drowned out by their fandom. It's also not mature to downvote people just because you disagree with them. And furthermore, the reason I use the navbar is because I want a force close long-press gesture on the back button, so I can more easily close apps and stop their background processes. To me, that's a fairly important feature for privacy and security. That's also likely a big reason Google pushed away from that, and to see GOS devs align themselves with Google in such a manner raises my eyebrow against them, at the very least.

There's always a bigger bully. Maybe a taller child needs to take their glasses while they're still recording and put them up somewhere high where they can't reach them? Could be hilarious and earn said child some popularity.

This is further illustration that raw model capability isn’t really the most interesting thing at this point. The model and the harness should be viewed as a single system. I expect that we’ll be moving towards neurosymbolic systems where LLMs act as a heuristic engine within the boundaries of a symbolic logic system.

How are all those things you just described not how neurons turn inputs into outputs? Of course a brain is greater than the sum of its parts, so it an LLM. LLM's aren't black i/o boxes either you can read the values of each individual neuron and it's connections. I still don't understand how brains can be more than just a collection of statistical predictions. It's not magic, they still function according to the mathematical world we live in.

My crackpot theory is that a lot of Western tech bros grew up reading Tolkien, and now they all think of themselves as Sauron forging the One Ring