No, regular GPS can't get you in the centimeter ballpark. One meter resolution is the best you'll get, infrequently, and requires DGPS augmentation (IMO, DGPS augmentation is common enough to count as "regular GPS", even though it is more than just GPS).

You can get centimeter precision with GPS + RTK, or with Static GPS, but neither of those are what people would consider "regular old GPS". GPS + RTK uses very precise ranging measurements to a ground station with a precisely-known location in addition to GPS. It's kind of like DGPS, but way better, except it only works fairly close to the ground station. Static GPS gathers GPS location samples over a long period of time -- days, even weeks -- and combines them statistically to get a very precise location. Static GPS can achieve sub-millimeter resolution.

Regular mobile GPS receivers get about 3-5 meters accuracy under ideal conditions. This improves to 1-3 meters with the addition of a common differential GPS signal, such as WAAS. You'll pretty frequently see 2 meter resolution when outdoors with a good view of the sky and DGPS.

We should look into genetically engineering strains of rice to absorb less arsenic.

Obviously we should also stop emitting CO2, but even if we were net zero today, CO2 levels would take a long time to decrease, and temperatures would continue rising for a while. And obviously it's going to take some time to get to net zero, so we need to think about mitigating the damage, not focus solely on emissions reduction.

It's really not. The party is large and complicated, and there is no group that is really in control of it.

What you could say with accuracy, I think, is that the Democrats in leadership positions failed to push back against a small but vocal extremist minority. They didn't exactly enact that minority's policies, though they did take small steps in that direction, but by failing to push back against the minority's extremist rhetoric they allowed the GOP to paint the whole party as supportive of the extremist position.

Why? Lots of strategic and tactical reasons, but ideologically the major factor was that Democrats as a whole agree with the foundational beliefs that drive the extremists, just not their extreme conclusions. For example, Democrats as a whole believe that all people should be accepted for who they are and not arbitrarily and unnecessarily denied opportunity.

This means that trans people should not be discriminated against. The extremists take this to mean "trans women are women, full stop", which poses some potential risks to cis women. But the mainstream Democrats didn't want to push back too hard because they didn't want to appear to be denigrating trans people, and besides the potential risks to cis women appear to mostly be just that: potential, as in, not actual. So, mainstream Democrats accepted but didn't really fight for trans women in women's sports, at least provisionally, largely on the basis that they couldn't come up with any good arguments against it that didn't threaten the basic principle of equal treatment, and based on the fact that there just aren't that many trans athletes and the ones that exist mostly aren't that good.

Meanwhile, the GOP made tremendous political hay, broadcasting far and wide the most extreme positions, claiming that Democrats wanted trans women in women's locker rooms and that trans women would push cis women right out of sports.

Lather, rinse, repeat over a lot of hot-button issues where mainstream Democrats didn't go along with the extremist positions but didn't openly fight against them, while the Republicans convinced their base that the extreme position was the Democrats' position.

This isn't correct. LLMs are language models, yes, but that description understates them, primarily, I think, by underestimating how much of the world is encoded in language. They could not generate reasonable-seeming output without also containing sophisticated models of the world, models that are almost certainly far broader and deeper than any expert system we've ever developed. And the newer LLMs aren't just LLMs, either. They have a reasoning overlay that enables them to reason about what they "know". This is actually extremely similar to how our brains work. The similarity is not accidental.

Maybe, but I think the your description misstates the LLM's role in such a hybrid system. Rather than the LLM being "just" an interface, I think you would ask the LLM to apply its own knowledge and reasoning to use the expert system in order to answer your question. That is, I think the LLM ends up being more like a research assistant operating the expert system than a mere interface.

However, if you're going to do that, do you really even need the expert system? Its role is to provide authoritative information, but curating and compiling that sort of authoritative data is hard, and error-prone. You probably don't want the LLM to trust absolutely in the expert system, but to weigh what it says against other information. And if you're going to do that, why bother building the expert system? Just allow the LLM to search whatever information you'd have used to curate the data for constructing the expert system, and to compare that to knowledge implicit in its model and perhaps elsewhere.

Many of our current-generation systems have access to the web. I've been using Claude and found it extremely good at formulating search queries, analyzing the content of large numbers of relevant pages and synthesizing a response from what it found. It annotates its output with links to the sources it used, too, enabling me to check up on its conclusions (I've yet to find a significant mistake, though it does miss important bits on occasion). It could be better, could analyze a little more, but it's already shockingly good and I'm sure it will get rapidly better.

This seems like a much more sensible way to make LLMs better than by backing them with exhaustively-curated expert systems. Yes, they will make mistakes, similar to how a human research assistant would. But this approach will ultimately be easier to build, and more flexible.

As an aside, I had a very interesting experience with Claude the other day. I needed to analyze a bit of code to see whether or not it made a security vulnerability that I had already identified unreachable, or whether there was some input that could be crafted to provide the output the attacker needs in order to exploit the vuln. Claude did not immediately give me the right answer, but it pointed out important and relevant characteristics of part of the code and analyzed the result almost correctly. I pointed out some errors in its conclusions and it corrected its mistakes while pointing out an oversight I made. I pointed out some oversights and mistakes it made, and so on. Over the course of about 10 minutes, we jointly arrived at a full and completely correct answer (bugs in the bit of code did indeed block exploitation of the vuln) and a concise but rigorous proof of why our answer was correct.

In sum: The interaction was pretty much exactly like working with a smart human colleague on a hard problem. Neither of us were right in the beginning, we both drew (different) partially incorrect conclusions, and were fuzzy about different parts of the problem. In discussion we each pointed out flaws in the other's reasoning, and in the process both came to understand the problem better until we finally arrived at the correct conclusion. I contributed more than Claude did. I think I can safely say that at least in the area of code analysis, I'm smarter than Claude (though Claude is faster). But this really wasn't a case of "rubber ducking". Claude also contributed significant insights.

LLMs today are not just schochastic word generators (assuming that phrase ever had any real meaning). If you think they are, you haven't used them much.

I think this particular form of hallucination -- inventing an explanation for an event -- goes even deeper and may be intrinsic to intelligence, period, or at minimum it's a characteristic that humans share.

There are a number of fascinating and clever psychological experiments that demonstrate this. Just one example: Researchers have asked people to answer a set of questions, and then a few weeks later asked the people to provide explanations of why they answered the way they did. But, in random cases, the researchers changed the answers and asked the people to explain why they gave those answers, even though they didn't. It was done subtly enough that few people realized the answers had been changed, even though they were often inverted. The really interesting part is that the people were just as good at explaining the answers they didn't give as explaining the answers they did.

Those and many other experiments seem to indicate that the primary job and ability of the reasoning layer of our mind is not to figure out what's right or wrong, or even what we do or don't want, but instead to invent explanations justifying whatever it is that we already think, from some deeper, non-verbal layer. There doesn't seem to be any evidence that the reasoning layer gets any hints from the deeper layer, either, it just finds something that makes sense. And we're extremely good at this.

The explanations we invent often don't hold up to scrutiny, but it's clear that our reasoning layer doesn't apply much scrutiny, at least not by default. We can vastly improve the accuracy of our reasoning just by making ourselves think through the process of explaining and defending our reasoning to another person, even without involving another person. This appears to work because while our reasoning ability is very good at inventing explanations, it's perhaps even better at identifying logical deficiencies in other peoples' explanations. So just going through the mental exercise of pretending to explain to someone else engages our reasoning to poke holes. And of course it's even better to actually engage with another person. The evolutionary advantages for a species that lives cooperatively but with internal competition are obvious. The person who is better at generating good explanations and poking holes in others' explanations will get their way more often, enabling them to reproduce and ensure the survival of their offspring. But because the rules of logic we apply work, as in help us to come to objectively correct decisions about the world, tribes arguing with each other will make better decisions, improving the odds of their offspring's survival.

Anyway, back to AI, it appears that's exactly what SAM did in this case. The backend system made a decision (to reject access) and while Sam didn't know the actual reason for the rejection, it invented a plausible one.

As a security engineer, it really makes me laugh that the AI chose an explanation that attributed it to security. It's so common for humans to do this. A huge percentage of the time when I see such security-related policy statements about systems whose security I think I understand, there is no actual security justification for the policy. Sometimes the speakers think there is, but they don't really understand security, and they're wrong. Probably sometimes the actual policy really is based on misunderstood security concerns, but I suspect that a lot of the time it's based on completely different concerns, and security gets invoked either as an intentional deception or, like Sam, because the entity generating the explanation doesn't know and invents something reasonable.

If anyone is interested in what I think is our best understanding of how human reasoning appears to work, I highly recommend "The Enigma of Reason", by Sperber and Mercier.

They're not claiming they've achieved AGI. No one has made that claim, AFAICT.

That said, at the current pace of development, it seems likely that -- unless we hit some unexpected wall -- we will achieve AGI within three years. Specifically, we'll get to the point where our AI models are as capable of improving themselves as our human AI researchers, but orders of magnitude faster. Unless it turns out that there is some sort of fundamental limit on intelligence, this will almost certainly mean that AGI will quickly become ASI (artificial superintelligence).

It's a very, very good time to start thinking hard about what that near-future world might look like.

That might have allowed the US to stockpile. As long as China gets the licensing system running quickly, and it's clear that this was really just a response to US nuttiness, the impact on their trustworthiness as a trading partner should be minimal to nonexistent.

Exactly. But also, you need to figure out how to be more competitive economically. Doing all of the rest but continuing to be a bad place to do business won't be enough. There are some bright spots in the EU in this regard, but none of them are the powerhouses needed to compete with China, or the self-sabotaging and perhaps Russia-aligned[*]US.

([*] I don't think we'll actually move into alignment with Russia on any kind of a long-term basis, but I never thought we'd elect such an obvious Kremlin stooge, twice.)

Europe should be focused on getting its shit together, strengthening the EU so it can override petty internal disputes and rationalizing regulations to become more economically competitive. The world needs y'all to step up and lead the western world, to become the new economic hegemon and guarantor of security, because Pax Americana has just ended, and the other option is a world order led by China. If we want liberal democracy, the rule of law, human rights, etc., to continue improving human existence, it's your turn to make that happen. The US is going to spend the next couple of decades, maybe more, just recovering from our self-inflicted wounds and (I hope) figuring out that we really do believe in our own founding ideals.

Oh, and you're going to have to do this while contending with climate change and the resulting series of refugee crises. Sorry, but we're going to be of limited assistance for a while.

Indeed, and his stupidity is deeper than that, because he's focused on bilateral balances of trade. Even if trade overall needed to be balanced (it doesn't, though there are some concerns about long-term global trade deficit), trying to balance it with each country individually is just... utterly clueless.

Consider a world in which there are three countries. Each of them has a surplus of a different resource, and needs a different resource. A needs what B has, B needs what C has and C needs what A has. So A sends money to B for stuff, meaning A has a trade deficit with B. B takes that money and sends it to C for stuff, meaning that B has a trade deficit with C. C takes that money and sends it to A for stuff, meaning that C has a trade deficit with A.

Add this all up and each country has a trade deficit with one country and a surplus with another country, yet each country's balance of trade nets out to zero... global trade in this simplified economy is perfectly balanced, yet each country has a large deficit with another country.

Real-world trade is far more complex than this, of course, with nearly 200 nations in the world and a very large number of different commodities and manufactured goods changing hands, but the principle is that same. You expect pairs of countries to have imbalanced trade, even if every country has a net perfect balance of trade with the rest of the world. So trying to "correct" pairwise trade imbalances is utter foolishness; if successful it would just break international trade and make everyone poorer.

What will really happen, of course, is that the rest of the world will reorganize its trade to exclude the US, so they'll all get wealthier and the US will get poorer.

Nonsense.

There are three problems with this argument. The first is that the relationship of American political camps to the rest of the world isn't really relevant. When you're discussing US politics, "left" and "right" are relative to the US range.

Second, if you want to include the rest of the world, you have to include all of it, not just Europe. That means you have to include China, Russia, India, etc. The result is a much broader spectrum with a center that is well to the right of Europe.

Third, you're looking only at economic policy, on which the US left is centrist relative to European ideas, sure. But if you look at immigration, race relations, abortion rights[*], etc., the US left is leading the world.

[*] Abortion rights in the US are complicated post-Dobbs. There are states with a near-total ban, there are states with hardly any limitations at all. The average is probably somewhere around the pre-Dobbs, Roe state, which was more permissive than the European norm.

Carter did that, not Reagan, and it was a good thing. It was what was needed to end the stagflation. Reagan's only real contribution was deciding not to replace Paul Volcker in 1983, though the work was mostly done by then.

That makes no sense in Trump's case, unless the puppet masters in question are Putin and Xi. Certainly it doesn't make any sense if you believe that US oligarchs are pulling the strings, because the Mad King has ended American economic hegemony, and that's only going to hurt them.

The undecideds are shifting, so his net approval rating is dropping. Mar 1 his net approval was +1, now it's -6, even though his approval rating has only dropped by about 3. It can't drop much further without the approval decreasing, though.

It's worse than that. Depressions end. The Mad King has ensured that America is no longer going to have a privileged place in world trade. After Trump 1.0 the rest of the world began quietly decoupling from the US, building trade coalitions, but Trump 2.0 has accelerated that process and made it irreversible. If the EU can get their shit together, they will become the new center of the western economic world, to counterbalance China. If they can't, then the Mad King has just handed the world's economy over to China. Either way, within a decade the dollar will no longer be the world's reserve currency and the US will not be the world's richest country. If the EU can capitalize, our GDP per capita will fall below theirs.

The Mad King thinks that America has been getting ripped off by trade. The fact is that we've benefited tremendously, but he has "fixed" that now.

Thus ends the American Empire. Not with a bang but with a toddler's tantrum.

I must have missed the announcement of the ChatGPT time machine that makes features launched today available a few days ago.