3 ms·

I see clear pattern. Some guys now tell you: self driving is around a corner, the cars are safe. We know how it ended with chemicals from 30 years ago. DDT is a

by lnsru 3y ago

I see clear pattern. Some guys now tell you: self driving is around a corner, the cars are safe. We know how it ended with chemicals from 30 years ago. DDT is a good example of it both for Western world as well for soviet union. 30 years from now we will shake the heads thinking about 2 ton vehicles with camera trying navigate by themselves. Looks like that it’s in the human nature to believe things that are too good to be true.

HPsquared 3y ago
Delayed effects are a bit different.
pas 3y ago
DDT as used is safe for humans though. it's estimated that DDT saved 500 million lives. so you know, it's not black and white. and FSD works, just not efficient if the safety parameter is set to high, because "suddenly" human drivers will start to bully it
- kuhewa 3y ago
  It's not so much about "as used", but how it was produced by Montrose Corp and dumped straight down the municipal sewer by the ton, decades later leaving a signature in every marine animal off southern California and sometimes much farther away.
Shorel 3y ago
Self-driving is very easy to evaluate. We humans are potentially very good drivers, and in practice we all can say if someone is driving safely or not. Chemical and biological interactions are very difficult to evaluate and predict, so I don't see the parallels here. One is a very hard research field, while the other requires complex hardware, but it is readily applicable if/when it works.
- wkat4242 3y ago
  Part of self driving is dealing with those human drivers who are terrible drivers though. This is what makes the whole thing so hard. If we could have dedicated roads for self driving cars it would be easy peasy. No need to interpret road markings, signs, other vehicles and their intentions. Because it would all be communicated digitally over radio. Steering a car is total peanuts for a computer. It's the unexpected stuff that makes it hard. Like that Uber car that killed that lady that crossed the road with her bike in the dark at a totally unexpected spot. A human is better at those things because they're caused by other humans who aren't logical. But such matters are really hard to test and evaluate for precisely because they are unexpected.
  - Shorel 3y ago
    All you say is true. Even then, that's many orders of magnitude easier than predicting all possible chemical interactions of a given compound.
    - hiatus 3y ago
      Predicting the actions of humans is easier than determining chemical interactions of a given compound?
      - Shorel 3y ago
        Yes, but only because in the general case, the actions are very limited by the controls of a car. Steering wheel, two pedals, that's only three axes of possible inputs.
- otikik 3y ago
  > Self-driving is very easy to evaluate. I disagree. We can say that someone is driving safely in one given moment. A drunk driver could drive quite safely through a main road while we watch them, and then turn around a corner and run over someone the moment they leave our sight. We can only extrapolate the "goodness" of a driver by expanding the time we spend observing their driving. 2-3 hours of evaluation will be quite solid. It would be even more solid if we are inside the car with the driver, and if we have received training about how to evaluate a driver. And the result would be even more reliable if the evaluator is financially independent from the driver. And for safety, the car should have controls that allow the examinator to take over if needed. So yeah, I am describing a driving test. As far as I know we don't have those for self-driving cars. We should. They should be way more thorough than than the ones we give to a human, because machines lack the shared context that humans have with each other. Tests like driving offroad, driving while there's tiny (fake) animals around, behavior in a (staged) accident, etc. Things that we know a human will have enough context to deal with, but for a machine it's not a given. One advantage is that these tests could be done in parallel. You can load the same AI in 50 identical vehicles and have 50 examinators run tests on it at the same time.
  - tanseydavid 3y ago
    >> We can only extrapolate the "goodness" of a driver by expanding the time we spend observing their driving. 2-3 hours of evaluation will be quite solid. Millions of miles driven with extensive telemetry for every instant in time and you do not think it amounts to a meaningful level of observation?
    - PakG1 3y ago
      Personally, I'd be more comfortable when get to trillions that cover all aspects of the globe in all seasons 24 hours per day over 365 days per year for perhaps a decade. Edge cases are likely insane everywhere, we already know that. Are my estimates overkill? Perhaps. But it's what I'd like.
    - otikik 3y ago
      Not if my son has to share a road with them, no.
pegasus 3y ago
Sure, we're shaking our head about DDT, but we're still very much relying on DDT's less obviously harmful successors. I hope it won't be the same with self-driving cars. Where there's a will, there's a way, and technology does advance. And at some point, the only way to improve things further will be to remove manually driven cars from the streets. I, too, am skeptical that wide-scale adoption of autonomous vehicles will be an overall good for humanity. I'm afraid it's going to make us humans even more dependent on hyper-complex technology, hence more vulnerable to issues like over-concentration of capital and break-downs of the global supply chain system. That it will atrophy some of our collective native capabilities (for example, the skill to assess risk and take action based on our innate grasp of physics of motion and psychology of others). We would in effect be outsourcing our physical safety to large corporations. Even if more safe, comfortable and efficient, I don't think that's a good idea.
biorach 3y ago
poor analogy

by lnsru 3y ago

HPsquared 3y ago

Delayed effects are a bit different.

pas 3y ago

DDT as used is safe for humans though. it's estimated that DDT saved 500 million lives. so you know, it's not black and white. and FSD works, just not efficient if the safety parameter is set to high, because "suddenly" human drivers will start to bully it

kuhewa 3y ago
It's not so much about "as used", but how it was produced by Montrose Corp and dumped straight down the municipal sewer by the ton, decades later leaving a signature in every marine animal off southern California and sometimes much farther away.

Shorel 3y ago

Self-driving is very easy to evaluate. We humans are potentially very good drivers, and in practice we all can say if someone is driving safely or not. Chemical and biological interactions are very difficult to evaluate and predict, so I don't see the parallels here. One is a very hard research field, while the other requires complex hardware, but it is readily applicable if/when it works.

wkat4242 3y ago
Part of self driving is dealing with those human drivers who are terrible drivers though. This is what makes the whole thing so hard. If we could have dedicated roads for self driving cars it would be easy peasy. No need to interpret road markings, signs, other vehicles and their intentions. Because it would all be communicated digitally over radio. Steering a car is total peanuts for a computer. It's the unexpected stuff that makes it hard. Like that Uber car that killed that lady that crossed the road with her bike in the dark at a totally unexpected spot. A human is better at those things because they're caused by other humans who aren't logical. But such matters are really hard to test and evaluate for precisely because they are unexpected.
- Shorel 3y ago
  All you say is true. Even then, that's many orders of magnitude easier than predicting all possible chemical interactions of a given compound.
  - hiatus 3y ago
    Predicting the actions of humans is easier than determining chemical interactions of a given compound?
    - Shorel 3y ago
      Yes, but only because in the general case, the actions are very limited by the controls of a car. Steering wheel, two pedals, that's only three axes of possible inputs.
otikik 3y ago
> Self-driving is very easy to evaluate. I disagree. We can say that someone is driving safely in one given moment. A drunk driver could drive quite safely through a main road while we watch them, and then turn around a corner and run over someone the moment they leave our sight. We can only extrapolate the "goodness" of a driver by expanding the time we spend observing their driving. 2-3 hours of evaluation will be quite solid. It would be even more solid if we are inside the car with the driver, and if we have received training about how to evaluate a driver. And the result would be even more reliable if the evaluator is financially independent from the driver. And for safety, the car should have controls that allow the examinator to take over if needed. So yeah, I am describing a driving test. As far as I know we don't have those for self-driving cars. We should. They should be way more thorough than than the ones we give to a human, because machines lack the shared context that humans have with each other. Tests like driving offroad, driving while there's tiny (fake) animals around, behavior in a (staged) accident, etc. Things that we know a human will have enough context to deal with, but for a machine it's not a given. One advantage is that these tests could be done in parallel. You can load the same AI in 50 identical vehicles and have 50 examinators run tests on it at the same time.
- tanseydavid 3y ago
  >> We can only extrapolate the "goodness" of a driver by expanding the time we spend observing their driving. 2-3 hours of evaluation will be quite solid. Millions of miles driven with extensive telemetry for every instant in time and you do not think it amounts to a meaningful level of observation?
  - PakG1 3y ago
    Personally, I'd be more comfortable when get to trillions that cover all aspects of the globe in all seasons 24 hours per day over 365 days per year for perhaps a decade. Edge cases are likely insane everywhere, we already know that. Are my estimates overkill? Perhaps. But it's what I'd like.
  - otikik 3y ago
    Not if my son has to share a road with them, no.

pegasus 3y ago

Sure, we're shaking our head about DDT, but we're still very much relying on DDT's less obviously harmful successors. I hope it won't be the same with self-driving cars. Where there's a will, there's a way, and technology does advance. And at some point, the only way to improve things further will be to remove manually driven cars from the streets. I, too, am skeptical that wide-scale adoption of autonomous vehicles will be an overall good for humanity. I'm afraid it's going to make us humans even more dependent on hyper-complex technology, hence more vulnerable to issues like over-concentration of capital and break-downs of the global supply chain system. That it will atrophy some of our collective native capabilities (for example, the skill to assess risk and take action based on our innate grasp of physics of motion and psychology of others). We would in effect be outsourcing our physical safety to large corporations. Even if more safe, comfortable and efficient, I don't think that's a good idea.

biorach 3y ago

poor analogy