3 ms·
>I'd love to know what the sticking points are probably the long tail of unique problems that any autonomous system in an open environment faces. It's kind of
by Barrin92 6y ago
>I'd love to know what the sticking points are
probably the long tail of unique problems that any autonomous system in an open environment faces.
It's kind of the crux of all these ML learning based systems. By definition you can't learn what's totally new and absent from data, but that is precisely where the intelligence lies when it comes to human drivers. Which is why I think genuine self-driving without fancy geofencing or whatever is almost an AI-hard problem.
- noneeeed 6y agoI live in Bath, an old-ish European city. It's dense, with narrow roads, many awkward junctions, and lots of onstreet parking that makes many roads one lane in places. I've often thought that FSD was basically impossible here without cracking general AI and some theory of mind just to tackle relatively simple cases of trying to negotiate with oncoming traffic in pinch points. But I'd always assumed that places like phoenix would be so much easier. It's interesting to see that even places like that are throwing up enough problematic situations that Waymo still don't feel confident launching.