4 ms·

>Doesn't the analogy fail given autonomous vehicles are eventually possible Only if there is a mathematical proof that eventually it will happen, someone might

>Doesn't the analogy fail given autonomous vehicles are eventually possible Only if there is a mathematical proof that eventually it will happen, someone might proof the reverse that current type of tech will never be good enough for general purpose driving and will work only in some specific conditions.

ben_w 5y ago
The only way it’s impossible in principle is if human intelligence (and a lot of non human animal intelligence) is due to some potent non-physical substance of thought which can never be made by any work of artifice on the physical plane, AKA a soul. Absent souls, human drivers are existence proof that minds made of matter can drive cars. That doesn’t stop being true even if the current preferred architecture is all wrong, as we can and do change those with ease.
- simion314 5y ago
  >The only way it’s impossible in principle is if human intelligence (and a lot of non human animal intelligence) is due to some potent non-physical substance of thought which can never be made by any work of artifice on the physical plane, AKA a soul. So I was thinking "eventually" as somewhat in next 100 years, otherwise even if you need a soul you could say shit like "eventually we will evolve to the next plane and become gods and we could create brains". Could it be impossible that we hit a tech limit for AI? I mean there is no sure what that we will invent something better then digital, silicon based computers. I am not that pessimist though, but is possible we get smart and realize that we could just put tracks and have self driving little trains on them to travel around: 100% safe (we could have actual scientist/engineers code the software for them) and much more green.
  - ben_w 5y ago
    I think you may be underestimating the rate of progress — 100 years ago, transistors were hypothetical, now we have made more of them than there are stars in our past light cone, and they are smaller and faster than synapses to the degree to which wolves are smaller than hills and move further per day than continental drift. IMO we will probably hit the single-atom limit in consumer devices this decade, though researchers are already suggesting multiple q-bits in a single atom so perhaps we won’t even stop when we get 1Å process chips. The main limiting factor right now is that we don’t really understand the structure of our intelligence. If you will permit an analogy: perceptrons and convolution kernels are like bricks, while current brain scans (MRI, EEG, etc.) are like nighttime satellite photos of city lights — we’ve got a gap in the middle that would tell us we need steel reinforced concrete to make skyscrapers and why want to build them in the first place.
yazaddaruvala 5y ago
https://nitishpuri.github.io/posts/books/a-visual-proof-that-neural-networks-can-compute-any-function/ https://nitishpuri.github.io/posts/books/a-visual-proof-that... This is a rather intuitive mathematical proof that neural nets can emulate all functions in many dimensions. Knowing integrals (Calc 102) would help. With this information it’s either that self-driving is eventually possible or human brains are not mathematically representable (i.e. something akin to a soul). Meanwhile, the data collection/storage is getting better (all Teslas are continually testing FSD and reporting failure cases, synthetic generation of “good and unique” perfectly labeled data[0] is only a year or two “old”). Sensors/cameras are improving (and have a lot more room to improve). The compute hardware is getting better and cheaper (analog computers[1], TSMC 2nm and beyond, Apple/Google/Teslas neural cores, etc). Additionally, given Waymo is already in production without humans in PHX it’s arguable that we already have “autonomous vehicles”. We now just need them to improve slightly and for Waymo or similar to test in the top 25 cities in the US. Which is mostly a matter of capital and time not technology limits. Maybe weather will be an issue for a few more years/decades - but even if Waymo only operated in Summer, I’d be happy to reduce my Lyft costs - and with enough data and maybe better sensors the neural nets will be able to handle weather too. Especially if the car just goes slow. [0] https://youtu.be/j0z4FweCy4M https://youtu.be/j0z4FweCy4M [1] https://youtu.be/GVsUOuSjvcg https://youtu.be/GVsUOuSjvcg [2] https://www.cnbc.com/2022/01/08/heres-what-it-was-like-to-ride-in-a-waymo-with-no-driver-in-phoenix.html https://www.cnbc.com/2022/01/08/heres-what-it-was-like-to-ri...
- simion314 5y ago
  >This is a rather intuitive mathematical proof that neural nets can emulate all functions in many dimensions. Yes, but the problem is we don't know what function we want to approximate. So it is like interpolation you give it some points and you get an approximation that is good enough around those points but probably very wrong if you are far away of the points. We don't even know how many layers and neurons are needed so it seems to be a brute force approach, throw more data and more hardware and when stuff goes wrong maybe throw more data. But one day a Tesla might go on a different environment and maybe people have different hats and goes to shit because those hats were not in the training data, we don't have like in other sciences bounded errors.

ben_w 5y ago

The only way it’s impossible in principle is if human intelligence (and a lot of non human animal intelligence) is due to some potent non-physical substance of thought which can never be made by any work of artifice on the physical plane, AKA a soul. Absent souls, human drivers are existence proof that minds made of matter can drive cars. That doesn’t stop being true even if the current preferred architecture is all wrong, as we can and do change those with ease.

simion314 5y ago
>The only way it’s impossible in principle is if human intelligence (and a lot of non human animal intelligence) is due to some potent non-physical substance of thought which can never be made by any work of artifice on the physical plane, AKA a soul. So I was thinking "eventually" as somewhat in next 100 years, otherwise even if you need a soul you could say shit like "eventually we will evolve to the next plane and become gods and we could create brains". Could it be impossible that we hit a tech limit for AI? I mean there is no sure what that we will invent something better then digital, silicon based computers. I am not that pessimist though, but is possible we get smart and realize that we could just put tracks and have self driving little trains on them to travel around: 100% safe (we could have actual scientist/engineers code the software for them) and much more green.
- ben_w 5y ago
  I think you may be underestimating the rate of progress — 100 years ago, transistors were hypothetical, now we have made more of them than there are stars in our past light cone, and they are smaller and faster than synapses to the degree to which wolves are smaller than hills and move further per day than continental drift. IMO we will probably hit the single-atom limit in consumer devices this decade, though researchers are already suggesting multiple q-bits in a single atom so perhaps we won’t even stop when we get 1Å process chips. The main limiting factor right now is that we don’t really understand the structure of our intelligence. If you will permit an analogy: perceptrons and convolution kernels are like bricks, while current brain scans (MRI, EEG, etc.) are like nighttime satellite photos of city lights — we’ve got a gap in the middle that would tell us we need steel reinforced concrete to make skyscrapers and why want to build them in the first place.

yazaddaruvala 5y ago

https://nitishpuri.github.io/posts/books/a-visual-proof-that-neural-networks-can-compute-any-function/ https://nitishpuri.github.io/posts/books/a-visual-proof-that... This is a rather intuitive mathematical proof that neural nets can emulate all functions in many dimensions. Knowing integrals (Calc 102) would help. With this information it’s either that self-driving is eventually possible or human brains are not mathematically representable (i.e. something akin to a soul). Meanwhile, the data collection/storage is getting better (all Teslas are continually testing FSD and reporting failure cases, synthetic generation of “good and unique” perfectly labeled data[0] is only a year or two “old”). Sensors/cameras are improving (and have a lot more room to improve). The compute hardware is getting better and cheaper (analog computers[1], TSMC 2nm and beyond, Apple/Google/Teslas neural cores, etc). Additionally, given Waymo is already in production without humans in PHX it’s arguable that we already have “autonomous vehicles”. We now just need them to improve slightly and for Waymo or similar to test in the top 25 cities in the US. Which is mostly a matter of capital and time not technology limits. Maybe weather will be an issue for a few more years/decades - but even if Waymo only operated in Summer, I’d be happy to reduce my Lyft costs - and with enough data and maybe better sensors the neural nets will be able to handle weather too. Especially if the car just goes slow. [0] https://youtu.be/j0z4FweCy4M https://youtu.be/j0z4FweCy4M [1] https://youtu.be/GVsUOuSjvcg https://youtu.be/GVsUOuSjvcg [2] https://www.cnbc.com/2022/01/08/heres-what-it-was-like-to-ride-in-a-waymo-with-no-driver-in-phoenix.html https://www.cnbc.com/2022/01/08/heres-what-it-was-like-to-ri...

simion314 5y ago
>This is a rather intuitive mathematical proof that neural nets can emulate all functions in many dimensions. Yes, but the problem is we don't know what function we want to approximate. So it is like interpolation you give it some points and you get an approximation that is good enough around those points but probably very wrong if you are far away of the points. We don't even know how many layers and neurons are needed so it seems to be a brute force approach, throw more data and more hardware and when stuff goes wrong maybe throw more data. But one day a Tesla might go on a different environment and maybe people have different hats and goes to shit because those hats were not in the training data, we don't have like in other sciences bounded errors.