DennisP

On the one hand there's stuff you think is worthwhile to learn about, and on the other there's the dopamine brain candy that youtube throws at you along the way. They just want to keep you clicking, they don't care whether you feel good about it afterwards.

Do a good job training it and this isn't much of an issue, but give it an inch...

Plus a good business can often get better returns by reinvesting profits than its investors are likely to get on their own. Investors always have the option to sell shares when they think that's not the case.

People should really look at cash flow statements instead of just earnings. Anthropic published one and their cash flow was positive for the first half of this year.

That doesn't mean all the circular financing won't be a problem, but people keep drawing the wrong conclusions from the negative earnings, which are relevant to investors but not so much to the viability of the business.

I saw a comment saying it is, and having a coding agent able to modify a running environment is a game changer, but I haven't wrapped my head around that yet.

Bad for hyperscalers. Good for Nvidia.

Help with health issues is turning into a big one.

My mother-in-law has a complicated set of medical conditions and sees half a dozen different doctors. Fable with max thinking came up with a bunch of interactions between drugs and conditions that her doctors hadn't considered. We've run its recommendations by one doctor so far and he agreed with everything.

My brother has one friend who used a frontier AI to diagnose his wife's uncommon cancer, and insisted on a test for that. It was correct and she's fine now since they caught it early. Brother has another friend whose wife had back pain after surgery; medical team didn't find what was wrong, he uploaded the scan, and the AI said "oh they left in a staple," which turned out to be correct.

I've started using it to work on a cookbook for the medically prescribed diet I'm on.

Wall Street Journal is now reporting that Musk ordered the restriction.

https://www.cnbc.com/amp/2026/10/20/trump-musk-starlink-ukra...

I would classify training and research more as capital investment than overhead.

Contrary to popular belief, plenty of long-term profitable businesses have no moat. It's just that if you use Warren Buffett's investment strategy and want to buy stocks to keep for decades, companies with moats are the surest way of doing that.

That seems an uncharitable assumption. I for one strongly object both to today's abuses and the treatment of Snowden, and GP may well feel the same.

That's Starshield, which is not under Musk's control. Reportedly the WH did make the decision.

https://www.reuters.com/world/europe/order-limit-starshield-...

It's the same tech but after Musk took flack for making a decision on Ukraine's military use, they split off a portion for the US government to control.

The Starshield portion of Starlink is already controlled by the government. It was Starshield that restricted access to Ukraine, and the decision reportedly came from the White House.

https://www.reuters.com/world/europe/order-limit-starshield-...

OP's article of course is about spectrum allocation in the US and potential trade retaliation, which isn't all that related to Musk's control of Starlink.

This is a superhero origin story if I've ever seen one.

Yes, and since so many apparently edible things are quite harmful, it seems reasonable to be cautious about new ones.

For the first few years they were just doing design work, and it's hard to say how much effort they were putting in since their focus back then was Falcon; the first Falcon 9 landing was 2015, and the first Falcon Heavy flight was 2018. For Starship they didn't even decide on stainless steel until 2019.

Comparing apples to apples, this seems to be the appropriate starting point for N1:

> Studied and designed by OKB-1 since 1959, it was the counterpart to the US Saturn V

https://en.wikipedia.org/wiki/N1_(rocket)

First launch was 1969. Starship's first launch was in 2023, so it took one extra year from initial concept to launch.

Starting from first launch, N1 did four launches in three years and never succeeded. Starship did 15 launches in three years, had multiple mission successes in which they achieved orbital velocity but had planned a suborbital trajectory, and then orbited on the 15th launch. Their first mission reaching orbital velocity was in 2024.

Also worth mentioning that a lot of the Starship work focuses on full reusability, a quite difficult problem which N1 did not attempt.

They might not consider the cold gas thrusters to be particular trade secrets. They're very simple devices.

Well I'd say they're ahead of N1 considering that N1 never made it to orbit.

Because if each neuron represents a small set of similar concepts instead of several completely different things, it should be easier than expected to look inside the neural net and see what it's really thinking.

It might still be quite difficult since the network is huge, and just because you know what the individual neurons represent doesn't mean you understand the complicated thoughts they're building. But easier than expected is still good news.

Also, here is a list of his papers and citations.

https://www.semanticscholar.org/author/Eliezer-Yudkowsky/254...

Check out this 1970s project that proposed extracting electricity from thermonuclear explosions.

https://en.wikipedia.org/wiki/Project_PACER

> A series of 50-kiloton bombs would be dropped into the cavern and exploded to heat the water and create steam. The steam would then power a secondary cooling loop for power extraction using a steam turbine. Dropping about two bombs a day would cause the system to reach thermal equilibrium, allowing the continual extraction of about 2 GW of electrical power

Checking Amazon, several of the available editions are illustrated, which I suppose would have the same problem.

In the US, courts enforce the Fourth Amendment by excluding evidence gathered in violation of it. Good thing too, because we're terrible at holding cops accountable for anything. To do that we have to rely on the executive branch, which is where the cops are.

Parallel construction happens and is hard to disprove but reportedly, the police were just lucky. The Oklahoma cop felt suspicious of a car with California plates, which sounds to me like his own personal or political prejudice but I don't know. If the cops there do this sort of thing a lot, then we wouldn't be hearing about all the times they found nothing, so maybe the find wasn't as unlikely as it seems.

I think you're pattern-matching into a different discussion. The people most affected by what?

By eating less beef? The people hardly using resources aren't eating much of that anyway.

By biodiversity loss? We'd do a better job with that if we stop looking at in terms of effect on people and start caring about wildlife for its own sake, and that's the perspective I take when I say the planet is crowded: not just crowded for people, but people overcrowding the rest of nature.

But if you insist on focusing on people, we're all going to regret the loss of fish from the oceans.

Side note: I'm not sure who you meant but in some of those third-world countries that "hardly use resources," there are people doing major damage to biodiversity by burning down rainforests and hunting endangered animals. Their countries just aren't burning much fossil fuel.

That would definitely help. But if people have to stop eating what they like because there's not enough space, I would call that crowded.

Well then we differ. I have insurance, and I'm willing to suffer inconveniences if that's what it takes to avoid an Orwellian surveillance state.

There are lots of other ways we get tracked, and I oppose those as well.

The judge went further than that. He ruled that the Flock search itself was a Fourth Amendment violation, and "called the technology a form of 'indiscriminate mass surveillance.'" He suppressed both the evidence from the car search, and the data from Flock.

So the choice is between major biodiversity loss, or eating a lot fewer burgers and steaks. Either way, I would call that crowded.

> In 1983, the Supreme Court held unanimously that ‘a person travelling in an automobile on public thoroughfares has no reasonable expectation of privacy in his movements from one place to another’. And in January of 2026, a federal judge in Virginia ruled in a challenge by two civil rights activists against Flock specifically that: ‘An individual scan of a license plate, a vehicle identifier that generally must be displayed under state law, does not invade any privacy interests’.

And a week ago, a federal court in Oklahoma threw out the evidence after a cop with no warrant checked a month of Flock's travel history on a car, used that as grounds for suspicion, searched the car and found drugs.

https://www.lawcommentary.com/articles/flock-license-plate-s...

I subscribe to an olive oil club, which buys a planeload of olive oil bottles each quarter from producers wherever it's harvest season, and ships them out to US subscribers. They send three bottles each time, including a mild which is often Arbequina. So thanks, I'll start using that one first :)

Doesn't seem that impractical since good olive oil producers pick by hand anyway. One example is Woodpecker Trail from the article.