Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[martinuzz]

It's hard to see that actually producing surplus energy. To achieve that pressure and temperature, how much coal is burned / nuclear waste generated?

[Starver]

It's a repackaging thing. All that surplus energy from the desert sunlight/mountain torrent/brineous straights/orbital microwave-beamer can be theoretically squeezed into into a relatively stable high energy-density form that all vehicles except Ford Pintos can safely carry with them for many more miles of travel than trying to do the same by just storing the electricity directly into battery storage, even if that's also rather safe (barring the Ford Galaxy Note 7) and maybe a little bit more efficient.

Plus, it sounds green. Whatever colour it really is.

(And dibs on refilling pumped-out salt-domes, so that the next civilisation to populate Earth get the chance to make all the same mistakes as we did!)

[martinuzz]

On the bright side, when turning sewage into oil, we can introduce the crapton as an official metric.

[Reelya]

Well it's probably not any less efficient than digging special rocks up from inside the ground and burning them. There's a lot of overhead in any of that. To get to the coal, you're digging through a whole lot of not-coal, and then you have any costs associated with that, plus you've lost money on whatever else that land could be used for.

Turning sewerage into fuel has the opposite impact to mining coal: it costs money and land to get rid of sewerage, so the cost of making biofuels out of it is offset by the savings of not treating it the way you normally do. Also, biofuels are worth money, whereas regular sewerage treatment doesn't produce a commodity.

So, even if the net energy-in and energy-out of this process are identical or it's lossy, there are other savings that need to be factored in.

[SalmonGod]

In some immediate down-to-earth / going to be having real effects on us in the very near future type-news... the very first commercial automated semi-truck delivery was made about a week ago.  Delivered some thousands of cases of beer over a little over 100 mile distance.  And I think the truck was owned by Uber?

Anyone else notice?  This is pretty big news.  Truck driving has been projected to be the next industry to be hit hard by automation for years now, and this is finally it for real.  Implementation should be easier than many industries that have seen automation digging in, I imagine.  Don't need to custom design a machine and do a bunch of re-building for every operation replaced by automation.  You just buy the truck and fire the driver.  Once the things are trusted enough to make runs without a human on board monitoring, that is, which I can't imagine will be too long, given the obvious financial incentive. 

Trucking is a huge, huge employer, so this is expected to have a major economic impact.  Could it be the point where popular opinion sways towards the need to ditch capitalism as we know it, and consider things like basic income?

[Reelya]

...and heated to 660 degrees Fahrenheit,

Well, that's convinced me. It now sounds like really hot s...tuff.

It's probably a rounding error. The exact optimal temperature is 666 degrees f.

[Reelya]

The funny thing is, we think of driving a car as something you need human-level intelligence for. But perhaps we're wrong about that in the same way that people once said "AI won't work a computer could never beat a human at chess". That's real BTW.

Driving a car on paved roads is a thing with some pretty detailed set of rules about what you can and can't do, what's expected of you. So perhaps it's very well suited as a field for automation.

People went from viewing chess as beyond computers to something inherently suited to computers. This also affected how we view what computers are capable of. Perhaps automated driving will have the same sort of synergy?

[martinuzz]

In some immediate down-to-earth / going to be having real effects on us in the very near future type-news... the very first commercial automated semi-truck delivery was made about a week ago.  Delivered some thousands of cases of beer over a little over 100 mile distance.  And I think the truck was owned by Uber?

Anyone else notice?  This is pretty big news.  Truck driving has been projected to be the next industry to be hit hard by automation for years now, and this is finally it for real.  Implementation should be easier than many industries that have seen automation digging in, I imagine.  Don't need to custom design a machine and do a bunch of re-building for every operation replaced by automation.  You just buy the truck and fire the driver.  Once the things are trusted enough to make runs without a human on board monitoring, that is, which I can't imagine will be too long, given the obvious financial incentive. 

Trucking is a huge, huge employer, so this is expected to have a major economic impact.  Could it be the point where popular opinion sways towards the need to ditch capitalism as we know it, and consider things like basic income?

I believe the truck was a Google project. The chips for it were made by the company I posted about a page or two back, that was bought for billions of dollars by the american chip designer, it was one of their big showcases on automotive chips.

[SalmonGod]

Driving has always seemed to me like one of the things most obviously best suited for a computer.  Faster reaction times, no variation in awareness levels, 100% adherence to rules, etc.

Also just read that apparently the system that drove that truck can be easily retro-fitted onto any truck with an automatic transmission?  It's only designed for driving on the interstate, so a human driver is intended to take over when driving in the city and making delivery.  It automatically pulls over at pre-determined locations for the human driver to take over.  So... instead of having a driver on-board for the whole trip, you can just have staff at off-ramp hubs who hop in, make the delivery, and are back at the hub in 20 min.  At the very least, this completely eliminates the need to ever have team drivers for non-stop overnight hauls.

It was actually Uber [link]

[PTTG??]

Looks like metallic hydrogen is a thing now.

Nextbigfuture is hyperventilating and assuming it's metastable. And then it's assuming that metastable metallic hydrogen will be a safe superconductor that won't explode when, say, jostled slightly.

Stepping back a bit from the over-excitability threshold... it's still really, really cool. Even if it's only stable at cryonic temperatures and moderately high pressures, it might be a fantastic way to store and use hydrogen. One big problem with hydrogen as a rocket fuel is that it leaks through everything. A stable solid form would have no such problem. Then all you need to do is dump solid cryonic hydrogen metal powder into a nuclear reactor and you have a two-stage nuclear thermal rocket.

Or maybe spray hydrogen chaff into a liquid oxygen stream and you have a top-tier chemical rocket.

Lots of options.

[Skynet]

The funny thing is, we think of driving a car as something you need human-level intelligence for. But perhaps we're wrong about that in the same way that people once said "AI won't work a computer could never beat a human at chess". That's real BTW.

Even during that era, there was at least one person who realized that human-level intelligence isn't necessary for chess-playing. I would like to refer to the  1971 Lighthill Report, a report that was responsible for massive cuts in AI spending in Great Britain and caused a mini-AI Winter.

[C]hess is a complicated enough game so that in a contest between a computer and a human player the computer's advantages of being able to calculate reliably at a speed several orders of magnitude faster need by no means be decisive (the number of possible positions being incomparably greater) and so there is real interest in whether or not they are outweighed by the human player's pattern-recognition ability, flexibility of approach, learning capacity and emotional drive to win. Another good reason for investigating chess-playing programs is that the long-term interest of the big international computer manufacturers in bringing about some spectacular achievement of machine intelligence against such a well developed human intelligence as an able chess player, in order to assist in selling more generally their products' potentiality for superseding human intellectual activity, has been an incentive to the devotion of quite considerable resources to producing an effective program.

It is interesting to consider the results of all this work some twenty-five years after the researches aimed at chess-playing programs began: unfortunately these results are discouraging. The best programs play chess of only experienced amateur standard characteristic of county club players in England. Chess masters beat them easily.

More important, progress on constructing chess-playing programs has been made solely by heuristic methods. The programs seek to maximise in what may be called the foreseeable short term a complicated evaluation function; this function, constructed entirely from human knowledge and skill, represents an evaluation of a position, depending on large numbers of different measurable features of it with different weights attached to them. What relatively modest success the programs have achieved is a measure primarily of human skill and experience in the concoction of this evaluation function. The computer's contribution is primarily rapidity in looking a few moves ahead and finding a line that produces a position change good on the basis of that evaluation. The intelligence contribution is human; what the computer offers is its speed, reliability and biddability. By contrast, learning programs are not considered applicable to computer chess at present.

To sum up, this evidence and all the rest studied by the present author on AI work within [robotics] during the past twenty-five years is to some extent encouraging about programs written to perform in highly specialised problem domains, when the programming takes very full account of the results of human experience and human intelligence within the relevant domain, but is wholly discouraging about general-purpose programs seeking to mimic the problem-solving aspects of human CNS activity over a rather wide field. Such a general- purpose program, the coveted long-term goal of AI activity, seems as remote as ever.

Today we have been able to build programs that are extremely good at chess, but only chess. Of course, now we actually do use learning programs (machine learning), so maybe...this time is different?

[Reelya]

Today we have been able to build programs that are extremely good at chess, but only chess.

That's just not true. A min/max algorithm is 100% good at every game that has a set of finite states, given enough processing power, and it only needs a finite amount of memory. Could you say the same for a human, that any human could solve any game trivially given enough time?

The reason we make chess solvers is because they're more efficient than min/max - a hard-wired solution is just going to be faster than one that can consider all possible games. But the domain-specific one is a lot harder to write, not easier. It's needed because computers have very limited processing compared to a brain.

If we threw the amount of CPU power that the brain has at basic un-optimized min/max then could we still beat that?

[Akura]

Found a good one:

https://hardware.slashdot.org/story/16/11/05/0434217/a-new-process-turns-sewage-into-crude-oil

The U.S. Department of Energy's Pacific Northwest National Laboratory has found a way to potentially produce 30 million barrels of biocrude oil per year from the 34 billion gallons of raw sewage that Americans create every day... The raw sewage is placed in a reactor that's basically a tube pressurized to 3,000 pounds per square inch and heated to 660 degrees Fahrenheit, which mimics the same geological process that turned prehistoric organic matter into crude oil by breaking it down into simple compounds, only...it takes minutes instead of epochs... The end product is very similar to fossil crude oil with a bit of oxygen and water mixed in and can be refined like crude oil using conventional fractionating plants.

But honestly, who'd actually want to run their car on this shit?

Fossils fuels in the last century reached their extreme prices because of their inherent utility: they pack a great deal of potential energy into an extremely efficient package. If we can but sidestep the 100-million-year production process, we can corner this market once again.

The thing about this isn't creating a new energy production, it's creating new energy storage.

[Reelya]

Well it's more than that. It also circumvents a waste problem. And if you look at history, innovations often come from rethinking a waste product. Gasoline for example was considered an unpleasant byproduct of kerosene production, so they used to dump it into rivers.

[Emma]

Driving has always seemed to me like one of the things most obviously best suited for a computer.  Faster reaction times, no variation in awareness levels, 100% adherence to rules, etc.

It's always seemed like that to me as well, most of the people around me seem to disagree though. They seem to think that humans are inherently better at absolutely everything.