The world of computer vision is much bigger than multimodal LLMs. You'd run an ensemble of specialized models for 3d mapping, object classification, path validation, and so on. On a raspberry pi 5 8gb you can run what you need to self drive an RC car on an obstacle track at 10 FPS.
I don't think you realize how boringly normal it is in Europe for people to never have a car and take public transit to the hypermarket, the same Cathedral of Consumerist Capitalism with a combined supermarket, shopping mall, movie theater, often a stadium for concerts and sports... And you can go with a granny hand cart multiple times a week because it's just a few stops on public transit, where it is also often a hub. La Defense and Forum des Halles in Paris, Stratford City in London, Mall of Tripla in Helsinki, Vasco de Gama in Lisbon
I know what the comment said, thank you very much. They were conflating two senses of 'public' in two sentences. I was responding to the implication that because these are, in one sense of the word, public, that means that it is OK to treat them as if they are public in a different sense of the term.
This:
> If the room has an IP camera in it, it is by definition not private.
Does not necessarily mean this:
> Since cheap cameras have begun to appear everywhere I treat them all as if they were publicly viewable.
The implication is that if someone misconfigured or otherwise didn't know their camera was broadcasting to the world, anyone is morally and legally correct in doing whatever they want with it, and it is their fault because it is "public". That is wrong.
> If the room has an IP camera in it, it is by definition not private.
No. No. No. No. No. No. No. No. No.
So if I put an IP camera inside your bedroom without your notice or consent, and hook that up to the Internet, you'd be okay with that? Because it's public!
A lot of these are probably from default or misconfigurations. A lot of these people with IP cam feeds visible to the Internet probably do not know they are open.
How about you take that stick out of your ass and stop assuming that norms which have existed for generations are, what, just some kind of mass mental illness?
Because if not, people will game the system and submit the same paper to dozens of journals at the same time, and journals are for publishing new, original work. And because in a class, you are being evaluated based on your ability to produce new, original work.
Do you also get angry at a track and field meet, when you show up with a motorbike and get kicked out for violating the rules?
But I don't expect someone who writes that much of an aggressively pompous yet uninformed statement to understand. Chesterton's Fence much?
Noble goal, but tell a bunch of scientists and startups that you've got a grand vision and $500 million in cash to burn, and they're always going to tell you a story about how it could be possible if you give them that money. And your sycophantic AI you use to research and vet will also always tell you there's a chance, if that's what you seem to be wanting.
If it is needing to generate that many tokens to do the same tasks, then it probably has higher inference costs. So (for you) the model is bad, the plan is the same plan.