You can make 3He here on Earth, though. That's probably more practical than setting up a remote mining operation on the moon (where the 3He isn't exactly plentiful either)
Not exactly - most of what the press has decided to call AI is machine learning of some form or another
That's just because ML is a popular way to implement AI. Nearly all AI advances in the last couple of years come from machine learning, so it's not strange that the press calls it AI. If researchers had produced AI using different methods, the press would be calling that AI.
Sounds like an area where machine learning could help in the near future. You know the goals, and you can run design through simulations to see how close you get to the goal.
Compilers are hard, but still, they're generally written by a very small software team.
Compilers for hardware targets are a lot harder than for a general purpose CPU, because the hardware offers much more degrees of freedom in implementing a design.
In an industry that already spends billions of dollars on design and manufacturing of chips, as per the example of $500 million for a single SoC, what are you going to do with a measly $100 million ?
Everything that's called "AI" today is just advanced pattern recognition
That's what intelligence is all about: the act of recognizing patterns and applying them in different ways. After that, it's just a matter of how complicated the patterns are.
It's not nit picky. It's a demonstration of your pervasive sloppy thinking.
Even if we remove the word 'fundamentally' it makes no sense. Human brains developed from non-thinking species by a bunch of small random changes that slightly improved survival, without anybody being there to understand the mystery of biological brains.
It will come with an ejector mechanism to dump all your groceries over the fence, or in your driveway. And then a legal fight with UPS will ensue because they've patented that mechanism.
You don't need AI ethics anyway. Just have the car follow the rules of the road, and use appropriate speed for the visibility so that it will not cause an accident.
Yeah, so lets do both, instead of just one, like you're suggesting.
A good idea in principle, except that we don't have the electricity to power both. It will be hard enough for electricity generation to keep up with the growth rate of electric cars that we need (as well as replacing other fossil fuel based applications such as home heating or cooking).
Also, even if we do both, we need a method that's not based on the idea of hydrogen fuel.
Programs that pass limited versions of the Turing Test have been around for a very long time.
That's why Alan Turing, didn't specify a "limited" version in his paper. No program today comes even close to passing a full test like Turing proposed, with high quality judges asking good questions.
Anyway, none of this really matters for what OP said and my response to that.
That's why I said "too early". While we still have most cars producing CO2 from fossil fuels, it makes no sense to start recapturing. Recapturing combines an inefficient process at one end (cars generate a lot of CO2 for little energy output) using an inefficient process at the other end (use a lot of energy to recapture a small amount of CO2).
When all the low hanging fruit is gone, we can start worrying about recapture, preferably using a process that produces something more useful than hydrogen.
Slashdot Mirror
Studies have shown that excessive cleanliness increases chances of developing allergies.
Good plan, except for the fact that a moon colony is a giant useless waste of effort.
If they want to be profitable, they should focus on cheap LEO/GTO, not on silly moon missions.
You can make 3He here on Earth, though. That's probably more practical than setting up a remote mining operation on the moon (where the 3He isn't exactly plentiful either)
Except these things already exist, and chips are still really expensive to make.
Not exactly - most of what the press has decided to call AI is machine learning of some form or another
That's just because ML is a popular way to implement AI. Nearly all AI advances in the last couple of years come from machine learning, so it's not strange that the press calls it AI. If researchers had produced AI using different methods, the press would be calling that AI.
Insane ? It's got 144 extremely limited cores, with tiny memories, small registers, no floating point, no cache, and no DDR memory interface.
It's low power, but that's the only good thing.
In this case, the article seems fairly level headed. How about reserving the pedantry for the cases where it's actually needed ?
Why not ? Make better tools, charge more money for them.
Sounds like an area where machine learning could help in the near future. You know the goals, and you can run design through simulations to see how close you get to the goal.
Compilers are hard, but still, they're generally written by a very small software team.
Compilers for hardware targets are a lot harder than for a general purpose CPU, because the hardware offers much more degrees of freedom in implementing a design.
If a $100 million effort can make a $500 million SoC design cost "a lot less", then these projects would already have been done.
In an industry that already spends billions of dollars on design and manufacturing of chips, as per the example of $500 million for a single SoC, what are you going to do with a measly $100 million ?
Everything that's called "AI" today is just advanced pattern recognition
That's what intelligence is all about: the act of recognizing patterns and applying them in different ways. After that, it's just a matter of how complicated the patterns are.
4) cancel the project after a few years.
It's not nit picky. It's a demonstration of your pervasive sloppy thinking.
Even if we remove the word 'fundamentally' it makes no sense. Human brains developed from non-thinking species by a bunch of small random changes that slightly improved survival, without anybody being there to understand the mystery of biological brains.
Something cannot be *fundamentally* impossible *until* a certain event, unless you have a very strange definition of "fundamentally"
This guy retired in 2011:
https://www.youtube.com/watch?...
It will come with an ejector mechanism to dump all your groceries over the fence, or in your driveway. And then a legal fight with UPS will ensue because they've patented that mechanism.
You don't have to have some sort of AI "ethics"
You don't need AI ethics anyway. Just have the car follow the rules of the road, and use appropriate speed for the visibility so that it will not cause an accident.
It doesn't mean you have to pass the test to be intelligent.
GP seems to think it's a requirement.
Yeah, so lets do both, instead of just one, like you're suggesting.
A good idea in principle, except that we don't have the electricity to power both. It will be hard enough for electricity generation to keep up with the growth rate of electric cars that we need (as well as replacing other fossil fuel based applications such as home heating or cooking).
Also, even if we do both, we need a method that's not based on the idea of hydrogen fuel.
In order to avoid 2C of warming by 2100, we need to have negative annual CO2 emissions by 2050
That may be true, but it will be challenging enough to convert most of our ICE vehicles to electric by 2050.
Programs that pass limited versions of the Turing Test have been around for a very long time.
That's why Alan Turing, didn't specify a "limited" version in his paper. No program today comes even close to passing a full test like Turing proposed, with high quality judges asking good questions.
Anyway, none of this really matters for what OP said and my response to that.
That's why I said "too early". While we still have most cars producing CO2 from fossil fuels, it makes no sense to start recapturing. Recapturing combines an inefficient process at one end (cars generate a lot of CO2 for little energy output) using an inefficient process at the other end (use a lot of energy to recapture a small amount of CO2).
When all the low hanging fruit is gone, we can start worrying about recapture, preferably using a process that produces something more useful than hydrogen.