No. It's inherent. If you give a system a goal, it will work to achieve that goal. If there's just one goal, it will ignore other costs. And if it's smart enough, it will notice that being shut down will (usually) prevent it from achieving its goal.

So the problem is we're designing AIs with stupidly dangerous goal-sets. E.g. obeying a human is extremely dangerous. He might ask you to produce as many paper clips as possible, and there goes everything else.

Happy to: It has NO VALUE.

The problem here is that your definition of "value" is invalid. Your claim is equivalent to saying that shares of Apple stock have value because we have built systems to ensure that ownership of them is unique and non-duplicable, and that the supply is finite. But that is not what makes shares of Apple stock valuable. What makes them valuable is that they represent partial ownership of a company that produces and sells hundreds of millions of useful devices every year, among other things, and generates hundreds of billions in profits.

Shares of stock, dollars and bitcoins are all non-duplicable, finite-quantity tokens, but the shares of stock and dollars are tokens that represent something in the real world, while bitcoins represent nothing beyond proof that some electricity and some hardware was diverted to create them and transact them, rather than put to a productive purpose.

Yes, it's true that the guarantees of the non-duplicability and finite quantity are stronger for the coins than the others, which could theoretically make bookkeeping easier and reduce the need for auditing of transactions, but it turns out that the cryptoasset guarantees are hugely more expensive than the old way of bookkeeping and auditing, and the non-reversibility of crypto transactions is a fatal flaw for serious real-world use (though it's a boon for scams). If cryptoassets were easier and cheaper to exchange than traditional methods, and if there were a good way to address the reversibility gap, it would make sense for countries and companies to switch to using crypto-tokens rather than audited database entries. Maybe someday someone will invent a crypto-token design that achieves that, and then we'll trade shares of Apple stock by exchanging tokens on a blockchain (or similar). But even if we do that, it will still be the case that it's the ownership of a productive enterprise that provides value, not the token, no matter how elegant the cryptography.

Note that I'm not saying this because I don't understand the cryptography behind cryptoassets. I'm a mathematician and a professional cryptographic security engineer; this stuff is literally my day job, and has been for most of my 35-year career. I deeply understand the (actually quite simple, though cleverly assembled) cryptography and precisely what it does and doesn't do. It strongly ensures that there is a limited supply and that coins cannot be double-spent, and it does this in a decentralized way. That's it. The limited supply and non-duplicability are properties required of any asset, but aren't what give an asset value.

Most people fundamentally don't believe in markets. Look at how they react when prices rise due to a supply bottleneck or similar. They shout "Greed, we're being gouged!", even when there's plenty of competition to keep prices down and rising prices is actually the optimal solution to suppress demand in the short term and encourage increases in supply in the long term. Conservatives are worse than liberals these days, since conservatism/liberalism has evolved to be divided in large part by education level, and market mechanics are really non-obvious and counterintuitive unless you've had at least some level of formal (even if auto-didactic) education in basic economics.

I think markets are, to a first approximation, the ideal solution for every problem that involves optimization of large-group human behavior. Only when markets clearly don't or can't work should you even start to think about non-market alternatives, such as direct regulation, or even subsidies.

But shockingly few people agree. Carbon pricing immediately makes them want to know who's getting the money and why they should get it, even though that question is irrelevant to the primary goal.

I have great confidence in market-based solutions to rapidly move us toward optimal solutions to the climate problem. I have near-zero confidence that it's what we'll do. Heavy-handed, imprecise and mildly counter-productive regulation is the strategy we'll take, if anything.

More like 80% for me. I use Rust.

(In Rust, an expression without a semi-colon at the end of a block -- including at the end of a function -- is the return value for that block. This is used heavily in idiomatic Rust, which means there are lots of lines that do not end with semi-colons. The more you use short, single-purpose functions and the more you program in a functional style, the fewer semi-colons you use.)

I feel I can be outraged. I have a Samsung 24" display that I brought into work and it stopped working. Fortunately work was willing to replace the display that I had sourced with one in work's inventory, but it was very annoying that a 1920x1200 display with DVI and HDMI inputs had its drivers removed and would only run at 640x480.

...then there's not a lot of difference compared to the status quo.

My point was to reduce the dependence on it, not to try to eliminate it. At this point we've invented so many uses for the leftover oil byproducts from the fuel refining process that I doubt we'll be able to truly eliminate them, but if the petroleum wax, or the polystyrene coating is net reduction in the quantity of oil required compared to making the full-plastic bottles.

Also fund research into those non-petroleum plastics and create and enforce labeling laws, so that consumers can know with certainty that if they're buying these alternative plastics what exactly goes in to them.

I'm also a fan of market solutions. Most people aren't, though, and they're especially challenging with an international problem, so I don't think carbon taxes or warming fees is how we'll do it.

I think you present a false dichotomy. Several of the proposed geoengineering schemes are relatively inexpensive, though of course a lot of investment will be needed to demonstrate their practicality and effectiveness. I expect that geoengineering will be much cheaper than getting to net zero.

However, that doesn't mean we don't also need to get to net zero, because the geoengineering proposals only address warming, not the other effects of high atmospheric CO2. Ocean acidification and whatever is downstream of that is probably the biggest one, but there are others.

Geoengineering should be viewed as a short-term mitigation strategy, not a long-term solution. I think we'll need that mitigation, though. Assuming AI doesn't kill us all and render the whole question moot :-).

You're treating this like it's a binary proposition.

The GP post is correct, if it's left in the ground then the carbon in the hydrocarbon soup remains contained, sequestered. The GP post didn't say anything about all of the byproducts that society has become accustomed to, you brought up all of that.

A great many products that people buy and use only exist or came to exist because of a desire to find a use for the byproducts of crude oil refining for fuel purposes. There are other ways of accomplishing the tasks that many of these products do. Hell, it's not a stretch of an argument that products made from wood, where the growth and harvest of timber is managed, are closer to zero-carbon since forests get their carbon from the air itself. So even if the wood products are burned at the end of their usefulness, they're just releasing the carbon taken from the air right back into the air again. And they don't even have to be burned, they can be landfilled, leaving the carbon captured from the original timber further captured.

We would do well as a planet if we reduced our dependence on crude oil for transportation and heat, and stopped making excuses for why we need cheap plastic crap that was invented to use the byproducts left of from producing petroleum-fuels as justification for pumping oil. We may never stop pumping oil as it has proven itself useful, but removing it from the ground simply to burn it when there are other energy sources available is wasteful.

I'm not sure. I myself made a similar argument about GUI management tools when the CLI tools were still required for troubleshooting, advanced configuration, and other similar things. I was concerned that the use of the 'dumb GUI tool' would lead to generations of IT engineers that couldn't fall back on basic principles.

In my thought experiment I was proven wrong, but only out of those GUI tools failing to provide a complete enough solution to begin with that the IT engineers could forego the CLI. They simply haven't. The CLI is still required because either the GUI tool can't do everything, or the GUI tool can't troubleshoot for shit, or the GUI tool itself breaks. So I guess technically the jury's still out on my own little thought experiment, but if AI works similarly to the GUI tool then there's going to be a lot of tinkering to make things operate anyway.

yes. It would seem that when it comes down to it, actually doing the work is what causes one to learn the material and basically shortcuts don't result in learning the material.

For what it's worth I didn't even like taking notes on a computer, I liked taking them by hand. And don't get me wrong, I don't like handwriting and my penmanship leaves a lot to be desired. I just found that if I wrote out my notes on paper that I learned the material, I did not even have to refer back to said notes most of the time. The act of writing them down helped me learn far more than typing the notes did.

Can you quantify how that might be a problem? ie "The wing tips were 3 feet higher than normal limits." From what I know, many parts of the 787 like the wings are made of composites like carbon fiber. The wings appear to bend more than on the 787 than on the older 777 which used traditional aluminum alloys. The new 777X will have composite wings and will bend more.

Yeah I would love to see an NFC standard that didnt extend the range, but instead specified where in the (*^^*&^#@ the reader was on the POS system, or at least a lit up nfc logo where you want me to tap.... Every POS I go up to has it somewhere else, and finding it is a pain

While Boeing recently had issues with safety and manufacturing, this plane (registration VT-ANB) was built in 2010 and delivered to Air India in 2014. Air India has been flying this particular plane for over 10 years. The root cause might be a design defect; however, it does not appear to be a manufacturing defect at the moment.