That's a rule I could have seen applied to me -- especially after I tried to plug something in on which the tines were bent outward, so naturally *I squeezed them inward while inserting the plug*.

Sometimes amazed I survived childhood.

We're all astronauts now.

Watched the splashdown with my boys and explained to them the great things the United States could accomplish and how science and engineering made this feat possible.

Back in blog days I called it shorter syndrome. ‘Shorter’ was shorthand for a bad faith paraphrase. Sometimes they were funny and arch and on point, a lot of the time it just gave people a lazy way to be self righteous. Micro-blogging (ie the Twitter format) really encourages it.

I figure it's either just a law of statistics (larger follower count == posts more likely to be seen by more people) or clout-seeking. Probably a bit of both. But either way, it seems more prevalent than I remember in times and venues past.

I've seen many misunderstandings in online discussions. But the amount of posts from people who loudly and with great certainty attack others while blatantly misinterpreting them has become pretty remarkable.

It seems to happen more to accounts with large follower counts.

This is the face of a lunar scientist who has just been told that the #Artemis II crew saw SEVERAL impact flashes (the flashes when meteors hit the lunar surface) in real time 😃 🌓💥

LIFTOFF #artemis

Heard it too - multiple aircraft. Flightradar24 didn't show much at the time, which seems to indicate it was a convoy of fighter jets or other military aircraft with transponders off.

LOUD jets in DC area you heard it here first folks

Flightradar24 doesn't seem to show them, which probably means they don't have transponders on. Also odd.

OK the jets flying low and loud in the Marylard burbs right now at 12:21 am is concerning. Anyone?

Ok so why are there fighter jets zooming around over DC at 12:22 am? Bc that’s what it sounds like

You would think, with the various bases near DC, that this kind of thing is typical, but it's really not, especially at midnight.

Huh. Multiple loud jets heard in the MoCo area near DC.

Interesting thought. Apart from the mountains, the distance depicted is roughly three times greater than the length of the Panama Canal - which took 10 years to dig/build.

Kind of reminds me of the DC-area snowfall forecasts for last weekend's storm.

It's that time again!

Unfortunately, I think that's kind of the point. A lot of people think it's patently obvious that LLMs/AI are utterly useless and terrible, and that is actually NOT patently obvious, and I agree with @chrislhayes.bsky.social that there is blindness to what is coming.

You can list a way in which it's useful (I mean, is that really surprising for any technology?) and all the caveats and ways in which it isn't, and you'll still get piled upon for supporting AI.

I think one can be plenty critical of AI/LLMs on the merits.

I find Google generates links in response to queries all the time. It's kind of the point.

Regardless, a correction: I didn't say, in my original post, that I was getting "horribly incorrect" results. In fact, the results in question were due to an incomplete query.

I mean, it's kind of a description of technology in general?

I wouldn't take the output of a Google Search and just insert the links as references in a paper without reviewing them. But people tend to think Google is pretty useful.

There are issues. Needing human intervention isn't one I'd cite.

But more generally, it's envisioned as being a useful tool for humans.

For all its well-known flaws, it's not going away.

...the sometimes-intuitive leaps-of-logic that so often figure into real breakthroughs just aren't there yet.

Moral of the story: AI in general (not just LLMs) have been a holy grail for decades for very good reasons. For some, yes, it's been because they want to replace workers.

There is a reason scientists complaining about nonsense but interesting-sounding LLM-generated hypotheses submitted by non-experts has become so routine. It's because it happens all the time now. LLMs have gotten pretty good at applying existing knowledge. But...

It REALLY does depend upon the use case. LLMs are pretty good at interpreting well-stated plain language requests for e.g. data analysis, BUT if you aren't also an expert in the field, you won't be able to apply a sanity check to the results.

"We can't just change channels. We have to go where the physics leads."

It was this one - (now previous) U.S. model run for total snow over the next 2-1/2 weeks.

The second snowfall over the weekend into Feb 2 was up to 30" in spots.

And now, the latest U.S. model (GFS) is trending back southward for the snow across the DC area. This will continue over the next few days.

Comparison between previous run and current: