tomo,

New one on me too. Like you say, a bit odd:

The colors in the image represent deviations in height (–100 m to +100 m) from an ideal geoid

Doesn't make much sense to me. The definition of the geoid is an equipotential surface — equal gravity everywhere — but obviously the earth's actual surface differs from this by a lot more than +/- 100m. What exactly is "deviating" in height here? Perhaps it's the true geoid as identified by the satellite vs. one of the theoretical geoids.

Well, whatever that all is, one assumption they seem to be making is that the big shifts are icebergs and the like on the surface. Just how much is known about flows in the mantle? Much denser material down there and (I'm going to guess) it's not "evenly mixed" and averaging it away would be a mistake.

.,
even after watching it, I still don't know. He did mention that it was portable; at the very least it was lightweight.

Forgot to mention in yesterday's that the neurosurgeon used my favourite technical term, "N=1", when describing how brain surgery has to be conducted: every brain is unique. We just need to extend that principle to the whole person, and get that message through to the medicos and, more importantly, to the authorities who have made guidelines into laws.

did you know?

https://www.youtube.com/watch?v=VvDsH8Fkxdo

Ross Lea,

That was a good report and the author had a nice level of dry humour in the introductory part.

I'd have liked it to have treated el Nino rather differently. I think it should be pointed out as one of the Emperor's private parts when we've been invited to admire his new clothes.

All through the global warming scare, people talk about el Nino/la Nina/neutral as inputs, as if they were elementary things like insolation or albedo. To me they are and always will be *results* of weather systems.

Climate modellers routinely use el Nino (et al) as an excuse for why their models have diverged from reality: how were we to know that would be an el Nino year? That their super computer models can't predict these things just demonstrates the utter inadequacy of their models.

Some facts for legislators to consider when trying to mandate battery powered trucks. Another emperor's willy on show.

I think the long-haul trucks are nearly as tough a problem as planes. Unlike your typical car, the trucks run nearly all the time at quite a high percentage of their engine's rated power. A truck does a long slog hauling a heavy load; the car might just have a quick zip away from the lights and then easy cruising from then on. In power consumption the car might use 70kW for a few seconds before settling down to to 15kW at cruising speed. Comparable figures for the truck might be 300kW for accelerating, but still needing 200kW to hold speed.

Still, it's not as if we need them, and won't the highways be so much nicer without them...

Enjoyed this EconTalk were Russ interviews a neuro-surgeon and a neuro-scientist on the mysteries of the brain. The neuroscientist had fun way of expressing himself. Couple of highlights.

At one point Russ started to contrast the surgeon's work with lobotomy (also covered recently on EconTalk) but the surgeon didn't let him finish his point being so offended at any comparison. The scientist defused it nicely by saying that the surgeon had been doing Formula One surgery, where lobotomy was more of a demolition derby.

The other remark was on a conversation he had with a robotics researcher. She had been working on robotic soccer players (apparently a robotics parallel for AI's chess challenge). She reckoned that high quality soccer players were well achievable, but what would really impress her was if somebody managed to come up with a quality soccer *supporter*.

I agree. If ChatGPT ever gets to the level of Golden Gordon I'll concede that it really is intelligent.

A very comprehensive review of climate based on empirical measurements. Very informative and relevant graphs.

https://www.thegwpf.org/content/uploads/2023/04/State-Climate-2022.pdf

Ross Lea,
Thanks for the link. Only skimmed once so far, and grabbed some of the LT data just for a look. Kind of reassuring that they stuck with FORTRAN for the rewrite. For what they're doing, the only reason to move to something more "modern" would be fashion.

Strange thing this global warming; it seems to have turned the blowtorch on the arctic. Something must be up with that OCO2 satellite 'cause it didn't show a super concentration of CO2 up there...

But the science is settled (especially by Mailman's mates Connolley and Rabett).

Today, Jo Nova carries the happy news that Ivermectin is no longer just a horse de-wormer. I hadn't heard that a class action was on the way against the (former) head of Australia's TGA. I hope he feels great discomfort. A nice shot across the bows for power-besotted bureaucrats everywhere.

Ross,

I had forgotten all about The Rabbit and Willy Boy until I saw them in the comments section!! Those two really should hang their heads in shame for what they have done to set climate science back by decades!

Impressive paper by Dr. Roy Spencer; a true scientist very transparent with some interesting comments.

Version 6.0 of the UAH Temperature Dataset Released: New LT Trend = +0.11 C/decade.

https://www.drroyspencer.com/2015/04/version-6-0-of-the-uah-temperature-dataset-released-new-lt-trend-0-11-cdecade/

tomo,
It's a drum worth banging. In a way, their false claims of "well mixed" also merit my repeated complaint about averaging away complexity.

OTOH, downstream from that tweet, I don't feel very compelled about gravity's influence on CO2 concentrations. In any case, the horzontal differences are much more interesting than the vertical.

DaveS,

The key thing to remember about AI is the A that comes before the I.

Afraid I don't agree. If it truly was intelligence, that it was man-made would make it all the more remarkable. Perhaps the right term for what has been developed is EI: ersatz intelligence.

From reports on the net it seems that some AI researchers believe that their programs really have become intelligent. I suspect this is because, unlike old-fashioned procedural programming, the neural net stuff has no obvious path starting from input and leading to output. It makes good results seem miraculous: I didn't do that, the machine worked it out for itself. In the early days of neural nets, the word training was used metaphorically. Now it appears that some researchers think of it literally.

So thirty odd years of advances in silicon and research into artificial intelligence and, lo and behold, we have achieved real stupidity.

The key thing to remember about AI is the A that comes before the I.