Okla really seems like a meme stock. Their original design was rejected by the NRC, so they are very far from ever breaking ground.
I don’t understand why their valuation is so high. Why not just take all this money and build an existing, approved design?
I will be very surprised if Oklo makes it. Insiders have been selling a fair bit over the last couple years because my speculative guess is they know that they cant possible meet the expectations in the market for their product.
They essentially got a ton of traction because Altman was on the board (but since left) but most (not all) tech people don’t understand deep energy problems.
Basically it sounds like what happens in failed countries:
> “It’s not like the NRC asks for an extraordinary amount of information,” said a former nuclear official who was involved in reviewing Oklo’s failed application and spoke on the condition of anonymity to avoid jeopardizing their work in the industry. “The NRC asks three questions: What is the worst that can happen, what are the systems, structures and components in your reactor that prevents that from happening, and how do you know that?”
>“Oklo would only answer them at a very high level,” the person said. “They wanted to say nothing bad can happen to our reactor.”
>DeWitte said Oklo had planned a robust public rebuttal but claims that at the time, NRC officials “threatened us, in a retributional way, not to issue a response letter to correct the record.”
> “It’s not like the NRC asks for an extraordinary amount of information,” said a former nuclear official who was involved in reviewing Oklo’s failed application and spoke on the condition of anonymity to avoid jeopardizing their work in the industry. “The NRC asks three questions: What is the worst that can happen, what are the systems, structures and components in your reactor that prevents that from happening, and how do you know that?”
This...does not square with the successful hamstringing of the nuclear energy industry by regulation over the past several decades.
Are you saying that the NRC asks for more than that? That there was a different process in the past? The big complaint I've heard about the NRC are changes required mid-construction, which happened last in the 1980s.
In the 2000s the NRC adopted a new licensing scheme at industry urging. What "hamstringing" are you talking about?
Okla would sound a lot more reliable here if they would have fought back with lawsuits with their accusations, or if the would release the communication now that there's no chance of this supposed retribution. As it is Okla makes all the talk of "hamstringing" seem like people not doing their jobs and trying to blame others.
Without speaking to Okla specifically--I think it's completely reasonable (if not accurate or charitable) to assume they're avoiding as much compliance as possible--the simple fact is that, under the watch of the NRC, there have been a tiny number of licenses issued.
If your agency's job is to regulate something and you've done it so successfully that barely anybody has actually gotten a license--all while complaining about compliance costs--maybe you're the problem.
Had the fellow said "Oh, we have a really high bar for safety and compliance, and not everybody's able to handle that", it'd be fine. But, acting like "oh golly gee we're so easy to work with we don't ask for much" is brazen horseshit.
The data doesn't fit for NRC being the problem. Hell look at the Summer reactor that was approved alongside Vogtle in Georgia: construction failure, billions wasted, and none of it to do with the NRC.
Or any the other many many other reactors abandoned at various states of development:
There's an argument that the NRC could do things better, but placing all the well documented failures in the nuclear construction industry on the NRC doesn't make sense. Who are going to believe, the people who are always late and over budget, or the bystanders in the industry that have watched it all play out?
I'd like to see the source of your cost numbers, I've never seen a $/MW from the Navy's subs' and carriers' reactors that was defensible, even Construction Physics didn't want to come up with a $/MW number in their discussion [0]
But remember that with the Seawolf classes the cost was astronomically higher than in the Virginia submarine; high costs are very possible without the NRC and are frequent, and an excellent counterexample to show the underlying fallacy behind the "NRC must be reason costs are high" argument. And remember that the Navy can use highly-enriched fuel that we don't allow in civilian reactors, and that the military nuclear labor force usually gets the best and the brightest and that the civilian nuclear work force gets the leftovers.
The NRC could be the source of high cost, but if so there should be two clear pieces of evidence to show that: 1) clear examples of the NRC doing something to drive up costs, and 2) some example of what to do instead of the NRC, or differences with other regulatory schemes that we could adopt instead. In particular, I never hear the corrective action that people want to the NRC. Having the Navy license civilian power reactors does not seem feasible. The closest we got to suggested regulatory reform culminated with Vogtle and Summer's failure: combined licensing. The biggest benefit of the industry's request merely gave the builders enough rope to hang themselves with bad design and their own delays.
Over the last decade the US Navy has commissioned 1-2 new nuclear reactors per year. They currently operate 100 reactors, more than any other org on the planet.
The US Navy has managed a total of 273 nuclear reactors, 6200 reactor-years, over 177 million miles, averaging 4 new reactors per year over 70 years.
They have done this with a perfect safety record. Zero accidents. Zero injuries, zero deaths, zero environmental pollution.
US Navy Cost: $2 billion for 2 x 400 MW reactors in Ford-class aircraft carrier
NuScale: $10 bn for 500 MW reactor
Westinghouse: ~$8 bn for 1000MW reactor
Military reactors should be more expensive, not less, because they operate under harsher conditions. But they aren't, because the US Navy doesn't have to go through the NRC.
Naval reactor power ratings are for thermal output. You can assume that about 1/3 of the thermal output can be converted to electricity when steam from the reactor is used to drive an electrical generator:
Assuming that your cited numbers are correct, "$2 billion for 2 x 400 MW reactors in Ford-class aircraft carrier" translates to 267 megawatts of electrical output for $2 billion. Or $7.5 billion for 1000 megawatts of electrical output. This is not much cheaper than "Westinghouse: ~$8 bn for 1000MW reactor."
The expectation is that their close-to-Trump investors will push for the dismantling of the NRC, which is something Republicans wholly support, which will of course make their rejection moot.
Maybe but the underlying tech still needs to perform which, as i understand it from public docs, has not. No amount of clear runway will make up for an airplane not being able to take flight.
The NRC is now mostly composed of Trump appointees. They’ve been quietly doing that. His most recent appointee was just made chair. Expect permits for friends of the Trump family and heavy regulation for competitors.
Anyone could’ve picked up the mantle of fixing the NRC, which is an obviously broken agency. France transitioned the majority of its grid to nuclear back in the 80s. Clinton, Bush, Obama, Biden, anyone could have picked up this low hanging fruit and fixed the problem. Nobody even tried.
Great post. I work on two large codebases. One is structured much like the example from the post, and the other is a mess. LLMs care much better at understanding the organized code.
> It is almost aways a failure of the technical infrastructure previously created in the company. An AI will solve the trivial aspects of the problem, not the real problem.
This is so true. Software that should be simple can become so gnarly because of bad infra. For example, our CI/CD team couldn't get updated versions of Python on the CI machines, and so suddenly we need to start using Docker for what should be a very simple software. That's just an example, but you get the idea, and it causes problems to compound over the years.
You really want good people with sharp elbows laying the foundations. At one time I resented people like that, but now I have seen what happens when you don't have anyone like that making technical decisions.
> what is definitely not inevitable is the monetization of human attention. It's only a matter of policy. Without it the incentives to make Tiktok would have been greatly reduced, if even economically possible at all.
This is not a new thing. TV monetizes human attention. Tiktok is just an evolution of TV. And Tiktok comes from China which has a very different society. If short-form algo slop video can thrive in both liberal democracies and a heavily censored society like China, than it's probably somewhat inevitable.
Radio broadcasting and newspapers monetized it even before TV. China is hyper-capitalist too, what is restricted is mainly political speech so that doesn't make much difference. If anything the EU is probably where advertisement is the most regulated. We can easily envision having way more constrains on advertisement and influencing, that would reduce drastically the value of human attention. Not sure many would get in the streets to protest against that.
The monetization of attention was a side effect of TV, not the primary purpose.
TikTok and other current efforts have that monetization as their primary purpose.
The profit-first-everything-else-never approach typical in late-stage capitalism was not inevitable. It is very possible to see the specific turns that led us to this point, and they did not have to happen.
> The real news is that it's also slightly happening in other developed countries too, another rhetoric point towards Steven Pinker's concept that as nations get richer they become more environmentally conscious, cause they can afford to care about it.
I'm not sure it's environmentalism. It's efficiency. From the article.
> In richer countries, where farming has become more efficient, deforestation has slowed or even reversed
You simply don't need as many people living in villages, farming marginal land. New England re-forested because the land was never that good for farming, and it made a lot more sense to work in factories.
This is a great idea. I always thought that if there has to be online gambling, it should be a government monopoly, and it should be managed by the most incompetent employees.
reply