Wait, uneducated people can't get high now? Things sure have changed since I was young. Back then, it was mostly the uneducated kids getting high. :-)

I mostly do heavy multi-disciplinary engineering problems (these I also use for checking an AI, as they're typically not good at these sorts of problems), complex coding, OS analysis, stuff like that. However, sometimes I do throw the occasional odd-ball - I've used ChatGPT to propose a workable quantum mechanics that will cope with Doctor Who canon, for example, and to produce an outline for a story in which symphonic metal appears in 1964 that is compliant with current sociological and psychological models of behaviour.

Claude Opus 4.6 was coping surprisingly well with just about everything I threw at it (but ran out of credits fast), but Opus 5 is churning out incoherent babblings to the point I'm worried I may have accidentally summoned Cthulhu.

But Gemini, Grok, and DeepSeek got hopelessly confused on just about everything past a very low level of complexity. They can handle large problems, yes - Gemini has a huge context window - but complex interactions baffle them.

ChatGPT is able to identify issues correctly, but can only outline solutions, it's just not good at depth. 5.6 is a lot better, but still not good at deep answers. ChatGPT is also prone to agreeing for the sake of it, which makes me nervous about trustworthiness.

That is fair enough. I've been trying out Kimi on the free model, and have subscriptions to Claude and ChatGPT. If Kimi is actually as good or better than ChatGPT, at the pro level, then it might be worth my while moving over as ChatGPT has become very disk-hungry of late and I'm pushing right to the very limits on what it can reliably process.

Although the Chinese models are all considerably smaller, it's not obvious to me that they're as good or better. They could be, but I'm not convinced the benchmarks that currently exist are stressing the models enough to establish where models actually are any good.

You use the can of grue repellant in the lake to get past those.

That humans have sufficiently directional hearing is perhaps the most impressive part. Once you have that, then the rest really just follows,

However, this continues the unexpected senses in humans, one of the first discovered was that humans have a weak magnetic sense.

The potential, both in fact (if you can "see" walls and gaps by clicking then you can presumably navigate a cave even if your light source goes out) and in fiction (this one should really be obvious), is considerable.

Not sure you quite get how dopamine works.

Heh. You reminded me that something I ordered got mis-delivered to some random unknown house, and since they don't provide delivery coordinates, there's no way to know where. So thanks for reminding me that I needed to reorder and reach out to Amazon Custom Service for a refund on that. :-)

This matches with my experience. Amazon has a wider selection than Walmart, but often slightly worse prices.

And yeah, Amazon's search engine is absolute trash, designed to promote what will make Amazon the most money over what you're asking for. Alexa shopping (on Amazon's website) is often better, and doing a Google search with site:Amazon.com is also often better. As another poster pointed out, odds are the seller on Amazon was just buying it at Walmart and selling it on Amazon at a markup. :-/

They didn't. The GP is very, very incorrect.

When a remote operator takes control of the vehicle, you'll know it, because it will mean that the vehicle has already been stuck for a minute or so without being able to figure out how to proceed, and will be sitting there for additional minutes while the remote operator looks at camera/LIDAR data and draws a proposed path through the problem area on a map.

It is not physically possible, for any amount of money, to do what the GP is claiming. Sending a signal over a cellular network from the Bay Area to the Philippines and back takes a quarter of a second on average. That's up to a 33% increase in response time compared with a driver in the vehicle. If you tried to fake self-driving with a remote operator in that way, they would crash into things constantly.

Even if the remote operator were in the same city, the latency spikes would still likely cause the cars to crash into things constantly.

LEO satellites also won't work, because these things mostly run in cities, where tall buildings would block the signals, so again, the cars would crash into things constantly.

Maybe if you had a dedicated cellular network running in parallel with the public one on a private frequency band AND dedicated trunk lines to a control center, you might be able to get something that worked most of the time, but at a cost that would be mind-blowing.

I finally dropped service a couple of years ago because the service had crept up to something like $70 a month for a basic landline phone with Caller ID. I kind of feel like they're laughing all the way to the bank either way.

Caffeine is a vasoconstrictor. It could well be the case that narrowing blood vessels causes blood to flow more rapidly through them, increasing its ability to wear away plaques that are attached to the walls of arteries, or that the flexing associated with the constriction and later relaxation helps these plaques break away in microscopic amounts rather than building up where they become a clog risk.

AliExpress and Amazon are both cesspools, where thousands of sellers sell identical or nearly identical products under different brand names with different markup. And most of the products are trash.

Case in point, I bought a bunch of metric bolts and nuts at Amazon, and they wouldn't screw onto each other. They would get to some random point and the nuts would stop turning. Sometimes it was half an inch up, sometimes it was two inches up, but the failure rate was something like 40% and I budgeted for 20%. I ended up buying a tap and die set and rethreading all the bolts. It actually would have been cheaper to buy the bolts in 5-packs from Lowe's. (On the other hand, if it happens again, I don't have to buy the tap and die set again, so I guess that's a "win".)

I've reached a point where about one in five Amazon products results in a return, because the quality is so poor, the descriptions are inadequate, etc. I buy enough stuff that they tolerate my return rate, but it's a huge hassle that I wish I didn't have to deal with. If I want a high risk of fixer-upper hardware, I'd rather buy from AliExpress for half the price and wait a week. :-/

Oh, thanks. I misread per hour as per day. But at 5 miles per hour, you won't get 40 miles per day. IIRC, the full-sun-equivalent hours in the U.S. varies from about 2.5 to about 6 or so. You won't get the equivalent of eight hours of sun unless you're smack in the middle of the Mojave desert with no mountains nearby on a dry day. So depending on where you are, you will get anywhere from 12.5 miles per day to 30 miles per day.

Here's a little more detail.

Researchers a few years back ran tests. ADHDers and neurotypicals were asked to harvest berries and edible plants.

Neurotypicals tended to over-harvest and damage bushes to the point where recovery would be slow if possible at all, and tended to pick areas clean, wiping out local populations of edible plants.

ADHDers tended to move relatively quickly from area to area, harvesting a decent amount from each area but not excessively. The quantities tended to be at the sort of level that have long been established as a good balance, such that plants and populations would tend to actually do better the following year. Not only that, but they tended to gather more.

https://www.psychologytoday.co...

Crofts and small farms rapidly went to three-stage and four-stage crop rotation, but different fields would be with different crops. This meant that the failure of a single crop would be much less of a survival thing.

You'd see fields with a mutitude of diverse crops, because diversity was critical. Failure was highly likely, so you localised it. Failure in one area had no meaningful impact overall. Things like the Irish potato famine and the parsnip failures across France in the 1700s were relatively rare in history. Likewise, when industrialisation wiped out cottage industries, farms that had been very successful suddenly went bankrupt and mass starvation was common. The ability to task-switch and have a large number of small projects going was essential to life at the edges.

You're absolutely right, more ADHD would have indeed curbed the excesses. The industrial revolution caused huge problems by trying to do too much with too little diversity. Which eventually destroyed the economies built around the early mills, because the markets were too easy to undermine. There was no capacity to switch. But during the IR, the obsession with coal meant nobody really looked at alternative fuels or alternative activities. Economies became tightly constricted, which was devastating when it came to global warming, but equally devastating when it came to distributing metal-devouring bacteria (which did enormous damage to infrastructure) and radioactive particles (old mill areas are still unusually radioactive).

More ADHD wouldn't have eliminated the problem, but it would have diversified the solutions and would have diversified the markets, resulting in a much more robust, much healthier marketplace and a much more robust, much healthier population.

Uh, no. As in, 100% of gatherers would be ADHD. IF they weren't, you'd be extinct within the year. 20% of the hunters would also be ADHD, because over-harvesting was also going to wipe out your food source. 100% of your explorers would be ADHD, 100% of your experimenters would be AuDHD.

ADHD was essential to survival. If you didn't want to die out, you stuck the neurotypical outside so that the wolves and other nasties actually hunting (tigers wouldn't be present) would get them. Neurotypicals were the main threat to survival.