The Guardian highlights the "prominence" of YouTube recommendations because it's attention-grabbing and bury the actual number -- 4% -- in the fourth paragraph, but if the other 96% of the responses were from reputable medicals sites, and if the YouTube videos were produced by doctors or scientists who are careful about their information and sources -- there are lots of them on YouTube -- this would be fine.

The actually interesting part of the study results is that while 34% of responses came from trusted medical sources and 1% from government institutions and academic journals, 65% came from "websites that lack formal review or evidence-based safeguards". That 65% is a lot more concerning, IMO, than the 4% from YouTube.

What would make such a study really interesting is if the responses were vetted for accuracy and reliability, rather than just categorized by source. That would be a lot harder to do, of course.

They're trying to steer him, anyway. It works sometimes.

I think this is only true until AI-controlled fighter jets become a reality. Human-piloted fighters will not be able to compete. Perhaps you were including fighter drones in your definition of "conventional air superiority".

I wouldn't assume there is an endgame, or any real strategy at all. All of Trump's decisions and actions are based on what he feels at the moment. In a few areas he has deep-seated beliefs that provide a bit of consistency, not so much to the details of policy but at least to the direction. And occasionally the people around him manage to engage in some planning that goes beyond the current hour, but only when their plans are aligned with one of his core beliefs, or when they're doing something he isn't currently interested in.

Agreed.

The judge didn't throw that one out, though. He should have, but the jury took care of it.

They don't have to go down forever... just decline everywhere to where they already are in the rich world. That's enough.

I think it would have to be a man-made plague. I don't think naturally-occurring viruses or bacteria are likely to be effective enough, fast enough, because we're going to be really, really good at making vaccines. Some of the COVID-19 vaccines were developed in literally days. Then testing took most of a year, but when we get the mRNA carrier technology down so we don't need to test or validate it any more, only its payload, that can be greatly abbreviated. And the more severe the plague is, the less time we'll spend on testing, because if the disease is highly lethal we'll care a lot less about the real but almost certainly lesser risks of the vaccine payload.

I think you just have to look at the Ukraine war to see the shape of the future, and it doesn't look like it includes tanks, or even trucks. As drone warfare has developed, it's become nearly impossible to move large vehicles across the ground. It seems plausible that the same may happen to the seas as well, though maybe in that case we'll end up with drone carriers able to deploy intense protective drone and missile shells around them that allow them to survive and move... but maybe not.

What will happen in the air? Drone fighters seem destined to make human-piloted craft unsurvivable. Without having to limit performance to the constraints of frail human bodies, drone fighters seem likely to become impossible for even 5th-gen fighter aircraft to take on. But will that even matter? The purpose of fighters is to attack or defend bombers (or to be small bombers)... but bombers seem destined to disappear from the battlefield, too, replaced by long-range missiles and/or drones (not that there's really much difference between them).

It may well be that the industrial capacity that matters in WWIII is the capacity to produce large quantities of batteries, motors and electronic control systems, and the technology that will matter the most is effective small-scale AI, able to operate autonomously with a high degree of effectiveness. In Ukraine, radio-controlled drones have all but disappeared because they're too vulnerable to jamming. Instead, the land is being webbed over with miles-long spools of ultrafine fiberoptic cable so the operators can have reliable remote control, but that cable has its own problems and limitations, and the solution is clearly autonomy -- though fully-autonomous flying robots carrying large quantities of high explosives are a different sort of nightmare.

Anyway, it's possible that the sort of manufacturing capability that is needed for the next great-powers war is exactly the same thing you need to manufacture EVs.

Of course, the best option of all is to nip all this war stuff in the bud. The whole free world needs to come down on Russia in Ukraine like a ton of bricks, throwing them entirely out of Ukraine, including Crimea, and coming together to declare that wars of conquest ended with WWII. Obviously, that's a pipe dream when the "leader of the free world" is executing one war of resource extraction and contemplating a war of conquest against his country's closest allies, but it would be the best way -- convince everyone that unless they think they can fight the whole world, conquest is just off the table. Sigh.

If current trends continue and we end up at a global TFR of 1.5 or so (where the developed world is now), by 2150 the human population will fall below 6B, possibly below 3B. But the TFR could fall even lower. South Korea and Hong Kong have TFR around 0.7.

There's only one real reason to be concerned about whether the rest of the world can compete with China, and that's national security. Specifically, if consumers shift to buying Chinese cars and the result is that everyone else's carmakers go out of business or permanently shutter most of their manufacturing capacity, that means that in the event of a war those manufacturing facilities aren't available to repurpose for military production. This, of course, assumes that repurposing them is useful. That was definitely the case in WWII, when the United States' massive manufacturing capacity was the #1 contributor to the allies' winning the war. But internal combustion-powered passenger vehicles are a lot more similar to the tanks, trucks and planes that were needed then than EVs are to whatever might be needed today (mostly missiles and drones? Plus platforms for both).

And there is another way to preserve manufacturing capacity other than forcing your citizens to spend more on transportation than they would otherwise have to: government funding for building and maintaining unused manufacturing capacity, including training for enough personnel that more could be trained when needed. That would require a lot of taxpayer money. Whether it's more or less than burdening taxpayers with paying higher prices for cars is an open question. Perhaps some economist has tried to work it out.

There is another national security risk to buying all of our cars from China, of course: given how computerized cars are today, it's possible that the cars themselves could be weaponized against us in the event of a conflict with China. We might be able to find ways to mitigate that risk, though. The main one is the manufacturing capacity question.

Maybe, maybe not. Our track record is bad, but it's not that bad. There are many environmental problems that we've tacked and fixed, and before they became severe. The ozone layer, acid rain, fire rivers, sewage dumping, industrial waste... a lot of these things were pretty serious problems that we've addressed just in my lifetime, at least in the US and other rich-world countries. Even minor stuff, like trash on the side of the road, has been massively improved (if you think roadsides in the US are trashy now, you should have seen what they looked like in the 70s before the "don't litter" campaigns and the "adopt a highway" programs).

We're obviously not doing enough on climate change, but we're also not doing nothing. Climate change is by far the hardest environmental problem we've created and need to fix, and shouldn't be taken as the sole example of how humanity deals with such issues.

At least where I live, I think we could solve water use problems easily, just by establishing a uniform water market that doesn't advantage farming or industry. Water taken from the aquifers is already regulated, and has been for decades. And, assuming the EPA is allowed to continue doing its job (not a given!), we've pretty much already solved water pollution. I'm sure other areas would require other sorts of interventions, but I don't think water is climate-change hard (though climate change will change water distribution and water needs, making it harder).

Nope. Google is all about the big hits. Unless everyone knows about it, it's not big enough to be worthwhile for them. And mostly they don't advertise. There are exceptions but in general if products can't get to 100M users with word-of-mouth, it's not something Google wants to continue.

Here's the context: Stadia was a Google game-streaming platform. Games ran on servers in the cloud, and streamed through a Google TV dongle device to your TV. So buying a Stadia kit meant getting a dongle and one or two controllers. The controllers were typical gaming controllers, but pretty high-quality. The game library was pretty good, and while you might think that latency would be a killer for this sort of application, it was completely fine for basically everything other than first-person twitch shooters.

I used Stadia and really liked it. I bought a number of games individually and subscribed to the Ubisoft library, giving me all of their games for, IIRC, $15 per month.

But it apparently became clear that it wasn't going to hit the 100M user mark, so Google shut it down. They handled the shutdown reasonably well, refunding 100% of the money spent on the kit and games, and with reasonable notice. This left users with useless. Also a Google TV dongle, but that could still be used as a general streaming device.

So Google released a tool that could be used to reflash the controller firmware, making the controllers work as normal Bluetooth controllers. I updated my controllers and have used them to play games on multiple platforms, simple and reliable. But Google announced up front that the tool would only be available for a limited time, and they eventually followed through and took it down.

Try plugging the route in to https://abetterrouteplanner.co... It offers a slider that allows you to choose whether you prefer fewer, longer stops or more, shorter stops.

As for more, shorter breaks instead of fewer longer ones, that's actually the right way to minimize total charge time because you take on energy faster during the first minutes of charging, though you do spend a couple of minutes getting from the highway to the chargers, pulling in and hooking up, so there's a point of diminishing returns (which is why the "shorter stops" option in the tool isn't also the "quickest time" option). For example, here's a route I plotted between a couple of random cities in Texas, about 800 km apart. It has four stops, of 5, 12, 13 and 5 minutes. Personally, I'd probably make the midpoint stop an hour long and sit down for a nice meal, which would reduce it to three stops; 5, 60 and 5 minutes -- and I'd make those same stops in an ICEV.

In practice I don't use a route planning tool since I upgraded to an EV with 400 miles of range. The Tesla Supercharger network attempts to place a charger every 75 miles, but there are some areas where they're up to 150 miles apart. With a 200-mile range, that means you have to completely fill the battery to make those longer gaps, which takes a lot of time, which makes pre-planning worthwhile. With a 400-mile range, I just let the car tell me where to stop and charge. The only planning I do is guess where I'll be around meal times, and then look for a charger in the surrounding area that is near restaurants, then add that as a waypoint.

In my opinion, and with the experience I have of owning and driving (including long-distance) EVs and hybrids, I think EVs are better. They're simpler, quieter, more powerful and just nicer to drive. The range thing just isn't a problem with EVs with 300+-mile batteries and reasonable fast charging networks.

Not enough range or towing capacity.

I've made a lot of long-distance trips with an EV, and they're really just fine. You do want to have an EV with 300+ miles of range. I did some long-range trips with my 2014 Model S, which had only 200 miles of range, and that was painful, especially since in regions with sparse chargers (like mine) such limited range means you often have to charge the battery to full in order to reach the next charger, and charging to 100% is slow. With a bigger battery you can avoid charging over 70% (except at mealtimes or overnight when you're stopped anyway), which makes your charging stops much shorter.

Apparently you consider a 20-minute stop to be an "extended" break, which isn't the term I'd use, but, yeah, that sounds about right. Except probably two stops if you start with a full battery and can charge at your destination. 800 km is about 7 hours of travel time. You should be able to drive for the first 3-4 hours without stopping, then stop to charge for 20 minutes, drive another two hours, stop for 20 minutes, then drive to your destination.

Did you try plugging this plan into an EV route planner? What was the actual route?

Infrastructure for driving an EV across Texas isn't bad. I'm not sure what exact route Tom was looking at, but Texarkana to El Paso is 814 miles and Google Maps says it would take 11:51. That's assuming no stops at all, of course. More realistically, even if you're peeing in a bottle you're going to have to stop for gas at least twice, so it's more like 12:15.

According to abetterrouteplanner.com, my Tesla would take 13:36 to drive that same route:

Texarkana to New Salem: 3h 52m, 259 miles, charge for 38 minutes
New Salem to Colorado City: 2h 18m, 170 miles, charge for 38 minutes
Colorado City to Pecos: 2h 24m, 175 miles, charge for 18 minutes
Pecos to Van Horn: 1h 12m, 90 miles, charge for 20 minutes
Van Horn to El Paso: 1h 39m, 119m miles.

In practice, using more realistic driving patterns, I doubt the ICEV would get there more than a few minutes ahead of the EV, if any.

Personally, I like to stop for a real meals on a long road trip. That erases a lot of the potential ICE advantage. In this case, stopping at noon and 5 PM, for 45 minutes each time, adds only 9 minutes to the total route time, increasing it to 13:45, because the car charges during those stops. Adding the same 90 minutes to the ICEV route means it takes 13:21. Add another 10 minutes to refuel onto those meal stops, and it's 13:31. Figure another 10 minutes for bathroom breaks along the way and the ICEV time is 13:41.

The bottom line is that if you drive like a normal person, long-distance ICEV trips arern't significantly quicker than long-distance EV trips, and that's true even in Texas, at least along major highways.

Where this becomes untrue is with pickup trucks when towing, which cuts your range in half. IMO, there aren't any usable electric pickups yet. Barring significant reductions in battery prices (and maybe weight), I think the closest that we'll see is hybrids. The 2026 Dodge Ram 1500 REV looks interesting: ~150 miles EV-only means it can be pure electric for daily driving, but ~650 miles of electric + gas range means that you have enough range even when towing. I'd really want a 3/4 ton, though. When it's been on the market for 3-4 years to work the bugs out, and there's a 2500 REV, I'll buy one, unless there's an even better option.