4 ms·
This seems like a cause and effect confusion. Because no one can solve the tech problem, no one wants to stick around for the point where the failure is obviou
by joe_the_user 6y ago
This seems like a cause and effect confusion.
Because no one can solve the tech problem, no one wants to stick around for the point where the failure is obvious.
Especially, for a while, self-driving cars have been at the level that impressive demos are possible but actual deployment isn't. So a fine career move is sticking around for long enough to create a demo and then leaving and blaming the failure on your successor.
Reason the tech can't deployed is that while 99% of driving problems involve just a complex, adaptive system, the 1% or .01% remaining involve "understanding what's happening", a far higher bar, one that requires a system well beyond what exists today.
- lumost 6y agoThis describes 99% of ML projects in industry. There is a reason the incentives are such that a science team creates a solution and then a theoretical engineering team deploys it. The Science team gets credit for the analysis/prototype but the prototype is so delayed getting to production that only the engineers take the blame or the scientists have moved on.
- joe_the_user 6y agoThis describes 99% of ML projects in industry. This may well be true given how fragile a construct most deep learning systems are. But there are still substantially different deployment issues for different systems. Google can deploy a question-answerer or a search-by-image-description system and lowered accuracy in practice doesn't really do much damage. Google can't deploy a self-driving car with low safety in the same way.
- bobsil1 6y agoLast .01% = “assume AGI…”
- echelon 6y agoI'm willing to longbets $5000/20 years: We won't have self driving cars that replace 30% or more of driving on existing roads and highways by 2040. The problem is too hard and everyone's drunk the ML kool-aid. (Just like everyone is drinking the NFT kool-aid right now.) Hype, unrealistic dreams, and spin. We'll be on Mars before we have generally available self-driving cars that do not require humans.
- avidiax 6y agoI think it will happen, but in China. We are trying to solve the wrong problem. When cars were introduced, people kept getting hit in the street. "Jaywalking" as a crime was created to shift the blame away from the technology (cars) to the victims (pedestrians). For self-driving vehicles, it suffices that we outlaw the edge cases, and enforce it to make them insignificant. The Uber ATC crash? Blame the jaywalker. Kid chasing a ball in the street? Blame the parents. Treat self-driving cars the same way that we treat trains: get out of the way, or you've no one to blame but yourself. What that leaves practically is preventing crashes with the static environment, self-driving cars, and "normal" road users (i.e. ones that can't be blamed if a collision occurs). This is much easier than the plastic bags and animals and cyclists and jaywalkers and wacky drivers that form most of the remaining unhandled 1%. The only country that can pull this off, IMO, is China. The mix of authoritarianism plus technical prowess allows the missing 1% of cases to be bulldozed in the name of progress.
- peytn 6y agoThe ruling party can’t even keep the factories north of Beijing from dumping pollution into the capital city’s air (and thus the lungs of the leadership and their families) for more than a few weeks at a time. I wouldn’t count on them here.
- 0xEFF 6y agoI'd consider taking the bet, but only against, "We'll be on Mars before we have generally available self-driving cars that do not require humans." Both seem unlikely.
- foobiekr 6y agoI think we should throw commercial fusion in there for the trifecta.