> It's the kind of problem that requires incredible amounts of data, hardware, and theory to make any progress.
I wouldn't be surprised if the opposite was true, at least with the theory part. AI didn't really go anywhere for decades, because people focused too much on theory.
Otherwise, there's a lot of data and hardware at your disposal, even from the comfort of your home.
> Wouldn't it make more sense for him to join a cutting-edge team, like DeepMind or OpenAI?
You mean they guys that are training with videogames that people like John developed?
Do you really think that's what he should have said? Because that sounds like a vastly dumber thing to say, even if it's "honest".
The "rogue operation" story is "good enough". It's plausible. That doesn't mean it's true of course. It doesn't have to be. Nobody has to actually believe it. It's just good enough so that everyone involved can save the minimum amount of face required to move on.
The reason that the murder can be considered "a mistake" is that the fallout was far greater than estimated. Consider that the Saudis are killing political opponents all the time with very few issues.
It's probably the grandeur of the plot that made the difference here. If Khashoggi had just been shot in the streets by "unknown assailants" (Putin style), it wouldn't have been that big a deal.
It is plausible in the sense of "plausible deniability". Imagine MBS saying something like "That Khashoggi guy, I wish somebody hacked him to pieces!".
You can interpret that as an order, or you can interpret that as a more casual remark of frustration. If you're ordering a murder, you may well want that kind of ambiguity in the order.
Furthermore, sometimes an unwise underling implements the perceived will of the dear leader in a way that is counter to their interest, e.g the assassination of Pompey "on behalf" of Julius Caesar.
In a state like that it is about as plausible as a US general starting a war with Russia on their own. Anyway, I don't suspect we'll ever have a fair trial where these details will come out. Just more murders to cover up for the previous ones, possibly in the guise of executions by the judicial system or what goes for that in that country.
I don't think the power dynamics are that clear. The Saudi state is not at all under total control of MBS. Formally, his father is in power, there are other potential heirs to the throne. There is the possibility of intrigue.
In hindsight, it wouldn't make sense for MBS to order this killing, but it would make sense for a rival to blame the murder on him. That's hindsight of course. Equally plausible is the scenario where MBS is so self-absorbed, he orders the killing without regard to potential consequences.
An assassination team flew from one country to another, murdered a person in cold blood in an embassy and then exfiltrated the body in pieces. Such an operation does not get executed without orders from on high.
> Negative grid prices are the consequences of power plants which cannot shut down in times of oversupply, that is mainly coal and nuclear.
Riddle me this, then: Why did negative energy prices emerge along with renewables[1]?
The problem isn't coal/nuclear, because the variance in demand is far lower than the variance in output of renewables. Furthermore, both coal and nuclear plants can be adjusted to react to seasonal variance.
The reason is shows up with renewables is that the marginal cost of production for renewables is zero. And nuclear and coal plants have high shutdown/startup costs. One of these is a negative the other is not.
The marginal cost of renewables isn't zero, producing it causes more wear and tear than not producing it. Furthermore, if you pay negative prices as a result of a renewable surge, that might as well be considered into the marginal cost.
Even if the marginal cost was zero, you gave coal/nuclear as the cause for negative energy prices, even though they didn't exist before renewables came into the picture. You may claim that the situation is better with the renewables in the mix, but you can't claim that renewables aren't the ultimate cause of negative prices in this case.
Similarly, you may claim that coal/nuclear are slow/expensive to regulate, but then must also concede that wind/solar can't be regulated up at all and couldn't provide a stable supply without e.g. natural gas, or a buffering solution that currently doesn't exist (at scale) and isn't priced into wind/solar.
Actually they didn't. Even before renewables, excessive energy production had to be taken care of. In Belgium, all highways got street lights to use up the excessive nuclear energy at night. In many places, electricity was promoted for heating at night.
The negative prices paid on EPEX did not exist until 2007 on the German/Austrian market[1], which clearly coincides with the green energy rollout in Germany.
Insured by whom? You need to understand how insurance companies operate: They gather data, they calculate the risk vs. the cost of paying out, they add a markup - that's your price. There is not enough data to do this with catastrophic nuclear accidents.
I understand how they work and I have done work both for insurance companies and re-insurers, nuclear power plants are insured a little different.
Here are some facts on how it works.
"Currently, owners of nuclear power plants pay premiums for $375 million in private liability coverage for each nuclear reactor they own. If there is an incident at a nuclear plant, and the $375 million in coverage is not sufficient, the owner’s coverage is supplemented by the second layer of protection, which is supplied by the nuclear power industry as a whole. Under the Price-Anderson Act, all reactor owners are committed to paying their share of any damages that exceed the incident reactor owner’s first-tier limit of $375 million—up to $111.9 million per reactor. Since are currently 104 reactors in operation, the amount that would be available in the industry pool to pay claims totals $12.6 billion (2011). "
It works because nuclear accidents are extremely rare and less dangerous than wind and solar let alone coal, oil and gas.
Thanks, this reminds me that I read this article half a year ago and forgot... my point is that it's not sufficient as the Fukushima Daiichi accident cost a lot more than $12.6 billion, let me just lazily link to my old comment on that: https://news.ycombinator.com/item?id=20014159
But what's the point though? Three Mile Island was fully paid by the insurance.
With regards to Fukushima the powerplant didn't kill anyone, the tsunami, on the other hand, did and of course, the public needs to pay for as it was force majure not the actual nuclear power plant that created the disaster.
And Tjernobyl was in a country which didn't have any private ownership so of course the public paid for it.
What other cases can you come up with where insurance didn't cover it?
It's pretty important to get things into proper perspectives here.
This isn't an insurance against catastrophic failure on the order of Fukushima or Chernobyl, which run into the hundreds of billions, if not trillions.
Of course you can have an insurance pool that covers arbitrarily small sums at arbitrarily low fees. That's missing the point.
Well you asked and I answered. In the case of Three Mile Island it was enough to pay.
"Of course, it is which is why there are two tiers. The other one paid by the whole industry.
But it's an absurd objection to begin with. All sources of energy have financial consequences if they go wrong, we use them anyway because the value they provide to society vastly outperform occasional disasters. And when it comes to human life, nuclear is far the safest, far safer than wind and solar. "
It's still not an argument against the safest and cleanest form of energy we have.
> All sources of energy have financial consequences if they go wrong, we use them anyway because the value they provide to society vastly outperform occasional disasters.
All energy sources aren't equal though. The value to society is cost/risk vs benefit. We're looking for the best deal here. The "catastrophic failure" scenario of wind energy is obviously completely different from that of nuclear power plants. That's a cost that's very difficult to estimate, but it is significant. You can't just ignore it.
> It's still not an argument against the safest and cleanest form of energy we have.
The fact that there's unfunded potential liabilities of enormous extent is an argument against nuclear energy. I'm not against nuclear energy at all, but let's be real here.
The energy wind and solar provides are also completely different. Capacity factor between 20-40 for wind and solar and 90 for nuclear. They provide less than 1% of the global energy and they don't have any other usage than providing energy plus the are unreliable and requires backup from other sources such as nuclear, oil, gas and coal.
I am not ignoring anything I am simply pointing to the fact that when we talk wind and solar the true cost isn't calculated whereas with nuclear it is.
And with regards to the insurance of catastrophes then Three Mile Island was paid by the insurance. Fukushima disaster happened not because of the nuclear powerplant but because of a tsunami, it was also the tsunami not Fukushima that killed thousands of people and Tjernobyl was owned and operated by the state so who else should pay for it?
So yes let's be real. No powerplant in the western world has yet cost taxpayers anything that wasn't covered by insurance and it has provided unparalleled clean energy which at least those who claim to worry about climate change as the biggest threat to humanity should support.
> I am not ignoring anything I am simply pointing to the fact that when we talk wind and solar the true cost isn't calculated whereas with nuclear it is.
I agree with you on wind/solar. I'm just going one step further and admit that with nuclear, the true cost isn't calculated either.
> And with regards to the insurance of catastrophes then Three Mile Island was paid by the insurance.
That convinces absolutely no one of anything. It's not what I was talking about as insuring a nuclear disaster. Don't waste your time making this point, it's worthless.
> Fukushima disaster happened not because of the nuclear powerplant but because of a tsunami, it was also the tsunami not Fukushima that killed thousands of people and Tjernobyl was owned and operated by the state so who else should pay for it?
It doesn't matter to the argument if a nuclear reactor blows up because of a tsunami, or an earthquake or Homer Simpson. It's a catastrophic failure case of enormous cost, and it's not insured, no money is being put into a fund to cover it, those costs are not taken into account when calculating the economics of nuclear energy. If such an insurance existed, it would be vastly more expensive for reactors built in a risk area like the Tsunami coast of Japan, so Fukushima Dai-ichi may have never been built.
I am not trying to convince anyone, I am trying to have a factual and principal discussion about insurance.
Again the three disasters we've had the one in the west that was the fault of the powerplant was paid by the power plant.
Of course you can't calculate the risk of being run by an opressive regime like the soviet union or ignorance/incompetence by those who granted Fukushima rights to build.
However they don't change the fact that nucelear reactors do in fact get insured and as far as I can see sufficiently. The other two other cases where either the state as the owner and thus insured properly, and the state allowing something to be built where it shouldnt.
Neither is an argument against nuclear and it's an unrealistic standard to set from my perspective and it's not a problem in the west.
I do however appreciate that we can disagree about this in a civilized manner.
Look, as far as the broader argument for or against nuclear goes, we are one the same side. I actually want you to have strong position.
> Of course you can't calculate the risk of being run by an opressive regime like the soviet union or ignorance/incompetence by those who granted Fukushima rights to build.
You can not by decree avoid ignorant people and/or incompetent people and/or less-than-ideal political frameworks. You need the wrong people to do the right thing.
> Neither is an argument against nuclear and it's an unrealistic standard to set from my perspective and it's not a problem in the west.
This is really just moving the goalpost. What if due to incompetence/negligence some reactor in a politically/economically deteriorating France blows up? Are you going to say "it's not a problem in America"?
You just keep repeating the mantra that it "isn't an argument". In reality it is an argument that "the other side" uses all the time. You're hurting your own credibility. You may disagree that it's a good argument, but it is an argument.
Now here's my standpoint: We can't by law require operators to purchase insurance that cannot exist. This is an impossible demand of "the other side", presented as a "cost" argument. It just doesn't help to say "but there is insurance!" when that insurance doesn't cover what we're actually talking about.
However, while we cannot calculate an insurance plan, we can pit the actual costs of either wind/solar against nuclear, including potential disasters and including CO2 prices and including the required buffering. I'm fairly confident that nuclear could come out on top even with one or two trillion-dollar disasters in the estimation. Remember, "the other side" just made a cost argument.
At another level, the argument for nuclear changes as the technology changes, newer and safer designs should have exponentially less dramatic worst-case failure scenarios. If "the other side" keeps factoring in speculative cost reductions in both production and storage facilities, so can we.
Insurance is a fringe discussion compared to general opposition to nuclear. Lack of insurance against catastrophic events is an even more fringe discussion and simply isn't a normal argument against nuclear. The normal argument against it is that it's dangerous, waste issues, expensive and so on.
I already stated my position on insurance of catastrophic events and someone isn't able to participate in good faith on that it wouldn't matter what I said anyway and my credibility would be tainted simply because of my position.
And to be frank, I've debated this enough and convinced enough people that nuclear is better than it's rumor. This is actually the first time anyone even mentioned insurance against force majure or communist regimes.
Anyway thanks for the discussion. I don't agree with you but but I appreciate the good faith argument.
> Insurance is a fringe discussion compared to general opposition to nuclear. Lack of insurance against catastrophic events is an even more fringe discussion and simply isn't a normal argument against nuclear.
Dissatisfaction with the way that insurance for catastrophic events is handled is in fact the main argument against expanding nuclear from the faction whose opposition is decisive in the US: the nuclear industry.
All other issues surrounding nuclear in the US are peripheral pieces of the debate over liability and insurance and whether or not the industry’s desired changes on that front should be met.
> You might assume this is just some techno-weenie blogger that's getting too excited about one of his hobbies, but if you look at the official submissions from companies aiming to build nuclear in Australia you'll see that this worrying lack of interest in objective truth is endemic to the industry.
It is endemic to human nature. There's no reason to believe that any industry is being particularly honest about their own raison d'être.
The issue is that you won't find an insurance company willing to insure a risk that they can't calculate. There's too little data to find a price here.
In the end, the government always bears the cost, because any catastrophic event will likely bankrupt the operator involved.
I think it's fair to count these catastrophic costs as subsidies, but they're not "untold gazillions". Estimates for the Fukushima disaster go up to about a trillion[1]. At the other side, the costs for Germany's "Energiewende" will run up over half a trillion[2], with relatively little to show for it[3].
I wouldn't be surprised if the opposite was true, at least with the theory part. AI didn't really go anywhere for decades, because people focused too much on theory.
Otherwise, there's a lot of data and hardware at your disposal, even from the comfort of your home.
> Wouldn't it make more sense for him to join a cutting-edge team, like DeepMind or OpenAI?
You mean they guys that are training with videogames that people like John developed?