We couldn’t disagree more.:D

I know people like to harp on the self-driving promises…

“FSD” almost got me into a major accident. It had a tendency to ride in cars’ blind spots and when someone cut me off I didn’t only make an evasive maneuver but also fight “FSD” which locked the steering and did not brake.

…the technology not fraud at all, except for the promises of how soon it will be self-driving.

So the “idea of the technology” is not fraud, only the presentation, the selling, and the delivery of it. So like everything that is currently available:D

Cameras are cheap sensors now

You know what else are cheap sensors. Actual sensors. FFS they had a >$20K profit margin on each car but they saved $100 on sensors

and humans do succeed in driving based only on sight.

This is so disturbingly incorrect. We rely a LOT on our hearing, our vibration sense, our proprioception too. Have you tried driving with earplugs? It’s pretty dangerous

Now it’s a software problem - will we be able to create an ai that can drive?

It’s a safety issue. For any safety-critical system you apply redundancy.

I don’t know, but I don’t see other approaches doing any better, and that’s with much more expensive and ugly hardware.

Mercedes and GM have much better autonomous driving systems than Tesla, they just don’t market it as “”“FULL self-driving”“”. The fact that you’re unaware shows how incredibly effective tesla’s misleading marketing had been.

we’re not all jumping into the same hole that may or may not succeed.

It will succeed, but eliminating safety measures in half-baked technology will claim lives. And nowhere did I say self-driving can’t work, I’m saying that it won’t work within the product’s lifetime and eliminating redundant sensor data will make the process a lot more unsafe.

Let them try something different, and we have a better chance of something working

Again other companies are already ahead of tesla without the bullshitting involved.