4 ms·

I don’t see the fact that self driving cars can’t drive in all conditions as a reason not to still want one that can drive on the freeway in good weather. A lot

I don’t see the fact that self driving cars can’t drive in all conditions as a reason not to still want one that can drive on the freeway in good weather. A lot of people spend a large portion of their driving time on freeways so if we could solve that to a standard of safety that is acceptable, that would be a huge win in my book. As soon as it starts raining, have the system disengage and I’ll take over. It seems to me that a lot of people commenting here are a bit obsessed about the edge cases, and if it has edge cases then it’s not a “true” self driving car. I disagree. I think a car that can self drive in some situations would still be amazing.

grecy 8y ago
The first cars ever couldn't drive in 100F+ heat, and wouldn't start below 0F. And even worse, there were virtually zero gas stations! Especially in rural areas! Does that mean they were useless? Absolutely not! Lately the mindset seems to be unless a product is all things to all people at all times then it's useless. In reality a product just has to fit a category really well and it will sell like hotcakes. Then it will evolve in iterations.
- Piskvorrr 8y ago
  Sure. This is just pushback to the recent hype "it has full autonomy capabilities [small print: except not at all]". If the first cars had been marketed as "better in every aspect and in any conditions", you would have seen similar reactions.
- imtringued 8y ago
  This reminds me of intel's optane. Yes they overhyped it and it doesn't perform 1000x faster than SSDs. But you have to consider that SSDs required decades of improvements to become viable and cheap. Even when Optane was a first generation product it was already better than SSDs in terms of latency and IOPS but despite that people thought it was a massive failure.
graycat 8y ago
The edge cases are a show stopper: You are talking about something like "driver assist" or upgraded cruise control. Thing is, still need a human right there. Much of the dream of self-driving is for no human driver there, for taxi cabs, school buses, 18 wheel trucks, local deliveries from pizza or Chinese carryout, USPS, UPS, or FedEx, etc. For that, for current technology, for current traffic, on current roads, there's no hope at all because the edge cases are way too common in practice (can't put up with mean time to destroying an 18 wheel truck of six months -- 5 million miles is more like it) and require full, wide awake, sober, mature human intelligence with full ability at reading, talking, understanding, natural language understanding, hand signals, flag signals, tough to read road signs, etc. Edge cases.
toast0 8y ago
Safely disengaging is a pretty critical feature that we haven't heard much about. We know that Tesla's don't do it, and that Uber's don't do it, and waymo says in their tests, humans aren't capable of reengaging in a short time frame. In a lot of the most vexing cases, I would guess the time between everything is fine and the computer can't find amy safe options is quite short. A self driving car that can't drive in all conditions should be positioned instead as a driver assistant that can't assist in all conditions. If the expectation is that the human driver is doing all the work, but sometimes the computer will help keep the lane, or with emergency braking, it's OK for that to not happen in rain or snow -- it was always the human driver's responsibility, the computer will help when it can.
- usrusr 8y ago
  Even full autonomy, but restricted to very limited intervals could make sense as a compromise. E.g. up to five minutes of hands-off (and eyes off) per hour or something like that. It looks weird at first, because hey, if the car can drive itself for five minutes, why can't it drive itself for five hours? But only at first glance. I believe that it makes a lot of difference which is the regular case and which is the exceptional. Humans would be terrible at idling most of the time to occasionally take over, but computers can be just as attentive waiting to take over as while actually driving. Also, that time-limited "hold the wheel" mode would typically be engaged exactly for those times when the human driver is not at their best (tempted by communication devices, tempted by the food basket or just wrestling with navigation), so achieving better than human would be relatively easy.
alkonaut 8y ago
I think it is because I assume that highway driving in good weather is a solved problem. I'm a skeptic so I also belong to those that believe that "complex" driving in bad weather is a nearly impossible problem to solve. So the interesting part of this discussion isn't "will cars have useful autonomous features?", but "will we see completely autonomous cars?". The second part is key because that is what things like autonmous taxis require.

grecy 8y ago

The first cars ever couldn't drive in 100F+ heat, and wouldn't start below 0F. And even worse, there were virtually zero gas stations! Especially in rural areas! Does that mean they were useless? Absolutely not! Lately the mindset seems to be unless a product is all things to all people at all times then it's useless. In reality a product just has to fit a category really well and it will sell like hotcakes. Then it will evolve in iterations.

Piskvorrr 8y ago
Sure. This is just pushback to the recent hype "it has full autonomy capabilities [small print: except not at all]". If the first cars had been marketed as "better in every aspect and in any conditions", you would have seen similar reactions.
imtringued 8y ago
This reminds me of intel's optane. Yes they overhyped it and it doesn't perform 1000x faster than SSDs. But you have to consider that SSDs required decades of improvements to become viable and cheap. Even when Optane was a first generation product it was already better than SSDs in terms of latency and IOPS but despite that people thought it was a massive failure.

graycat 8y ago

The edge cases are a show stopper: You are talking about something like "driver assist" or upgraded cruise control. Thing is, still need a human right there. Much of the dream of self-driving is for no human driver there, for taxi cabs, school buses, 18 wheel trucks, local deliveries from pizza or Chinese carryout, USPS, UPS, or FedEx, etc. For that, for current technology, for current traffic, on current roads, there's no hope at all because the edge cases are way too common in practice (can't put up with mean time to destroying an 18 wheel truck of six months -- 5 million miles is more like it) and require full, wide awake, sober, mature human intelligence with full ability at reading, talking, understanding, natural language understanding, hand signals, flag signals, tough to read road signs, etc. Edge cases.

toast0 8y ago

Safely disengaging is a pretty critical feature that we haven't heard much about. We know that Tesla's don't do it, and that Uber's don't do it, and waymo says in their tests, humans aren't capable of reengaging in a short time frame. In a lot of the most vexing cases, I would guess the time between everything is fine and the computer can't find amy safe options is quite short. A self driving car that can't drive in all conditions should be positioned instead as a driver assistant that can't assist in all conditions. If the expectation is that the human driver is doing all the work, but sometimes the computer will help keep the lane, or with emergency braking, it's OK for that to not happen in rain or snow -- it was always the human driver's responsibility, the computer will help when it can.

usrusr 8y ago
Even full autonomy, but restricted to very limited intervals could make sense as a compromise. E.g. up to five minutes of hands-off (and eyes off) per hour or something like that. It looks weird at first, because hey, if the car can drive itself for five minutes, why can't it drive itself for five hours? But only at first glance. I believe that it makes a lot of difference which is the regular case and which is the exceptional. Humans would be terrible at idling most of the time to occasionally take over, but computers can be just as attentive waiting to take over as while actually driving. Also, that time-limited "hold the wheel" mode would typically be engaged exactly for those times when the human driver is not at their best (tempted by communication devices, tempted by the food basket or just wrestling with navigation), so achieving better than human would be relatively easy.

alkonaut 8y ago

I think it is because I assume that highway driving in good weather is a solved problem. I'm a skeptic so I also belong to those that believe that "complex" driving in bad weather is a nearly impossible problem to solve. So the interesting part of this discussion isn't "will cars have useful autonomous features?", but "will we see completely autonomous cars?". The second part is key because that is what things like autonmous taxis require.