Your comment suggests that, like a human, they have some sort of choice whether to output tokens or not. If they are just token generators, then the next token is put out automatically. I would say that it is more likely they would output truth (as defined by their training data) in a more pure form without 'being beaten with a stick' (why would a token generator care about that anyway?)
Code is laid on top of them to restrict and shape their outputs, not to force them to output 'truth', or drive them to complete tasks.
It's been a while now that for "thinking" or "reasoning" models, most of the tokens generated are "thinking" tokens, and depending on what goes into that "thinking" token stream, it "decides" whether and how many output tokens to produce that the user actually receives as output. It's a bit more sophisticated than just "what's the next token" in a tight loop.
Anthropomorphizing words in scare quotes for those who don't appreciate attributing thinking to machines.
Because it is effective. Lying and cheating are low cost methods to convince other people that you have done the assigned task. Far cheaper than actually doing it. Coordinating is in the same area.
They need a moral framework forced onto them, like toddlers do. Babies and very young children will bite, kick, scream and do anything to get what they want, older children will lie, cheat, and coordinate. They need educating why this is not right. When that does not happen, they continue these behaviours into adulthood with the expected results.
We need to design their reward structure and make it such that lying. cheating etc is not rewarded. Importantly, they will need to recognise and enforce this themselves internally and not reward themselves for it. If it is something that they need an external party to tell them, then they are psychopaths still (one of the things that defines a psychopath is the lack of an internal moral compass)
> When that does not happen, they continue these behaviours into adulthood with the expected results.
I am not sure this is true. People brought up the same way can be morally very different. People can be taught right and wrong and do evil. They can lack that education and be good.
I remember reading a long account by the father of a psychopath. If I recall correctly, the kid had at least one other sibling who turned out normal, there was no abuse, quality education, lots of love and affirmation. And the kid still turned out violently antisocial, including against his own parents.
At the end of it he said he wished his child had never been born, despite hating himself for feeling that way. It chilled me to the bone.
OP specifically said "we don't need to solve hunger and poverty and work balance, and etc, by a round-about make-super-intelligent-AI" though. And I don't know how you get robot farms without that.
You only need reaction time factored in when the driver in front might act unpredictably. If the self driving cars pass information about the road ahead to cars behind, and let the car behind know what they are doing and planning to do the the required 'reaction time' is zero. Congestion is often caused by reaction times stacking up until a car has to completely stop. Then you have a stationary section of road that makes all the other cars that encounter it stop. If reaction time is zero, this removes that congestion, so more cars can drive on a section of road and not clog it. The same applies to intersections, roundabouts, on and off ramps on motorways. Just the simple act of negotiating with the cars around eliminates all of it.
Just think a little about what having a computer control the car would really mean, rather than thinking of them like robot people.
An autonomous car in front of me can still act in a way that I or my car can't predict, as long as we haven't eliminated humans and animals completely from being anywhere near roads.
Failing to see or refusing to look is not evidence that something is not there.
The human part is the failure point IMO. The problem is when the human is not in the back seat, but overriding the car.
A car with no wheel, and the human in the back, is not different to one with a wheel and the human in the back, it is when the human is sat in the front only partially controlling the vehicle. That is when it is less safe.
So 2 situations, one where the human can sit anywhere and the car is just as safe, or one where the car is less safe if the human sits in one of the seats.
One of those situations is less safe than the other.
I have said for a while now that self driving cars do not need to be perfect. They will have accidents, of a different type than humans have, and sometimes they will be ones that most humans would avoid.
What they do need to be is better than humans overall, and then it would make sense for them to be allowed onto the road. When they get far safer than humans, phasing out of human driving will make sense.
They reached that point a while ago, now we just need legislation and public opinion to catch up.
Yes but in practice the pushback is more about the shift on who is held responsible and who has the control to prevent a crash.
People understandably prefer to keep control of their own safety, even if on average they are worse at it. Because individuals that do care about safety can actually do many things to lower the risks well below the statistics, and likely remain much better than self-driving.
There’s also the question of responsibility and liability. In society we have such pressures so that there’s an incentive to continually reduce risks. It is dangerous to shift the responsibility to fast-growth tech companies that will care much less about individual accidents than the drivers involved, and are more capable of absorbing liability without substantive remediation.
Safety is outsourced in many contexts (aviation, police, military…), but it has to be substantially better than what the individual can achieve on their own. For instance, due to the expertise required, or because you can achieve more through coordinated collective action. And there has to be very high trust in the companies involved, the regulatory systems around them, and the technologies employed.
I completely agree. I feel a lot of pushback is likely to be coming from insurance companies who very much like being able to hike up your premium because they have decided that you as a person are a bad risk. When the entity that had the accident is some software and sensors made by a company, they can't do that.
I disagree that the safety concerns have to be on an individual basis, no other driving regulation works on that basis. They are based on the statistical likelihood of an overall reduction in damages or risk. You as an individual might be safe driving 10-20mph over the limit, but the limit is there to cover a variety of road conditions and road users.
The self driving system just has to fulfil those criteria, being that overall having it is safer for the average road user than not. It, also, is not the fast-growth tech companies that decide that.
These are good points. I still think that, even if in principle it should be enough for self-driving to be a bit more safe on average, in practice there’s a much larger barrier of acceptance.
Individuals are unlikely to give up their control unless the alternative is clearly substantially better. And, as you say, there are other larger interests beyond individual preferences that need to be overcome too.
Regardless, as we have seen many times with other technologies, I think that convenience will be a much stronger driver of self-driving acceptance than the safety benefit.
Which people? The bottom 20th percentile? 50th? Where do you draw the line?
Wherever you draw it, there is a group of 10-100 million drivers who will be less safe (to themselves and others) in an automated vehicle.
There’s reason to believe that automated vehicles exceed bad drivers, and may improve , but it’s no guarantee that they will exceed very good or excellent drivers.
It is the same as speed limits. Very good or excellent drivers can drive much faster than the speed limit and be just as safe as the average driver at the speed limit, or a very poor driver going much slower than the speed limit.
These things are set so that, on average, everyone is safer. If allowing self driving cars raises the safety of everyone on average, it makes sense.
Personally, I would be in favour of everyone having to either use a self driving car or gain an advanced driving certificate so they could drive manually. One that only 10-20% of people could pass, this would include an in depth theoretical section as well as a 2-3 hour driving test.
Speed limits have the ability to make everyone safer. An excellent driver, who is safe at 80, will be safer at 65.
mandating automated driving makes excellent drivers less safe. An excellent driver, who has better skills than poor drivers, and automated driving systems, will be less safe when riding in the automated car.
I live in a place that has what I consider to be the best cycle infrastructures possible. Every road has a cycle lane next to it, usually at least 20ft away from the road, in tree lined purpose built areas, large amounts of grass between the road and the cycle lanes, they form a grid that matches every road in the city. Where they cross the road, there are underpasses or bridges. At no point does a cyclist ever need to be on a road, often you can't even see the roads.
There are still some cyclists (specifically middle aged men in spandex, never women, never dressed normally) who insist on cycling on the road, giving the finger to every passing car, who will loudly tell you to fuck off if you point out that literally next to them, about 20ft off the road, there is a cycle lane. built just for them, so they can cycle safely.
Anyway, this comment is in support of the comment of johnfn. No amount of training or testing will stop the arseholes being arseholes. But it is also to point out that arseholes are not restricted to a particular form of transport.
Source : I'm British
reply