I (maybe naively) believe a healthy society could find a way to build a robust public transport network and still accommodate the minority of enthusiasts who drive and work on cars for fun.
Engineers aren't just dry husks of people, robotically creating solutions to meet needs. The drive to create cars, planes, and motorbikes, which have significant technical overlap with trains, buses, and mobility aids, is at least partially borne from the thrill of piloting machines that extend human capabilities.
I have a Quest 2 VR headset that I use for playing sim racing like Assetto Corsa, and flight sim on Xplane 11. To use that I have to open up Meta's Quest app, connect the headset to the computer over the WIFI, and it sorta functions like a monitor. In that I can view the whole Windows desktop environment on a virtual screen floating in VR space. When you open a VR game like Xplane you stop seeing the floating monitor, and it takes over the whole VR eye space for the duration you play it.
Is this type of thing also possible on Ubuntu? If so, I'll shitcan Windows ASAP.
I once naively used Windows file copy utility to transfer my huge MP3 library to an external hard drive and later lost the originals. I came to find out it silently failed to copy any songs containing certain nonalphanumeric characters. To this day I’m still traumatized when I try to locate some song and find it’s not there. Burn in hell Windows.
Fair enough but it just seems like a fluffy distinction.
And I don't think they "tweak the algorithm" so much as generate a load more training data of that one specific task to get it up to spec.
In any case, humans make mistakes on lots of stuff too, so if the criterion for "true" understanding is to make no mistakes then humans cannot be said to understand either.
Meaning what? It needs Cartesian Dualist qualia floating around between its wires and transistors, or else it's just a word vending machine? What's the demonstrable test for understanding vs "understanding"?
I'm reminded of that Futurama episode where the gang logs onto year-3000 VR internet and is immediately assaulted by a vicious swarm of flying viagra ads.
I made a kind of deal with myself that if I wanted baked goods and sweets I had to make them myself. Since then I've learned to make brownies, cookies, ice cream, sorbet, chocolate ganache tarts, pancakes, and more. It's fun, allows you to be creative, and the extra work of having to make if yourself keeps you in check.
I actually really like all those designs. They're bold, playful, distinctive. Much more interesting than the dreary crossovers of today that all look identical - bloated hatchbacks with unnecessarily high ride-height, angry anime eyes, and oversized grilles. Full-width tail light bar like a dollar-store Porsche completes the look.
I like Gnome because it looks sexy and sleek, and comes default on my Ubuntu. I have a little experience with XFCE and LXDE on Proxmox and Raspberry Pis, and they're perfectly functional and great, so I don't want to besmirch them. But they give me a kinda uneasy sensation like I'm using a tamagotchi or something. I don't know if this is only because I'm using them on low-power potato computers or without proper display drivers, but they just look a little crude by comparison.
What you're describing is a Battery Management System (BMS), whose job is to monitor some key parameters of the cells and make sure they remain balanced. There's no intrinsic reason for it to be tightly integrated into an overarching system that performs surveillance or other high-level functions in a "smart" vehicle. This video by Great Scott explains the basic principles and he even builds a simple one from scratch, that would be suitable for something like an e-bike https://www.youtube.com/watch?v=rT-1gvkFj60
Sufficiently motivated people have been building highly performant DIY electric cars for several years with no Big Brother tech in the OpenInverter community https://openinverter.org/wiki/Main_Page
I have high hopes for concepts like Toolformer where the model has to learn to use external APIs and resources like Wikipedia or Wolfram to get answers, rather than relying on the inscrutable and garbled soup of knowledge absorbed from the text training corpus directly. Systems plugged into knowledge graphs could have the best of both worlds - able to generate well-written novel text outputs AND the added rigor of "classical AI" style interpretability.
I (maybe naively) believe a healthy society could find a way to build a robust public transport network and still accommodate the minority of enthusiasts who drive and work on cars for fun.
Engineers aren't just dry husks of people, robotically creating solutions to meet needs. The drive to create cars, planes, and motorbikes, which have significant technical overlap with trains, buses, and mobility aids, is at least partially borne from the thrill of piloting machines that extend human capabilities.