I've found it useful in a few cases:

Writing dos batch files to do specific tasks. I've been using Dos since the 80's, and have a workable knowledge of batch files, but I have to look up more complex commands and syntax.

Pasting in something I've written for a grammar/typo/consistency check (knowing full well it's probably adding that text to its training, so nothing important.). It may not be perfect but it's light-years ahead of grammar tools I've briefly looked at in the past. Mostly it can catch sentences I've edited once or twice and accidentally left an extra word in.

Some linux/ubuntu stuff which I don't use that often, like setting up a logical volume across several old drives in a repurposed/dumpster bin PC I'm configuring for a 3rd or 4th copy of backups. I check every step of the way, but it's easier to ask chatgpt using my specific hard drive designations (e.g. sde to sdi) than copy/paste from a website and edit all the examples. Having said that, on one occasion it included my boot drive sdc in the command to wipe an old LV which I'd already said was on sdb to sdj, so yes, check every step and know what you're doing.

On that note, diagnosing linux error messages too.

But I'm certainly not going to ask it to get creative, or try and get it to do research for me, or instruct me in something I don't already have a working knowledge of. People who pick up a scalpel and ask ChatGPT how to do brain surgery are the problem, not AI itself.

Why do you say it wouldn't be expected to have an accretion disk? The fact that we haven't seen it already isn't evidence.

"We don't know whether there are any kangaroos in my house. The fact that we haven't seen any means that they are invisible. How do we know that they are invisible? If they were visible, we would have seen them already."

Not close enough to notice. Stars are very far apart. No, strike that. Very very far apart.

The way to test it is to find Nemesis.

There were multiple searches looking for Nemesis, all of which failed to turn up anything. There's a bit of a discussion in the book "Collisions", by Alec Nevala-Lee
  https://unabridgedbookstore.co...

It should be rejected if it is a false positive correlation.

People have been looking for some periodicity in mass extinctions ever since the "Nemesis" hypothesis some forty-plus years ago. The problem is that each study trying to find periodicity found a different period, depending on the details of what they counted as disasters and what they didn't. (For what it's worth, the recent biography of Luis Alverez, "Collisions", talks about the search for periodicity,)

I worry about this part: they studied "mass extinctions, volcanic eruptions, ocean crises, and other upheavals". These are completely different things. Adding different things together makes false positives much more likely... and also makes the possibility of unconscious "p-hacking" by picking what events, and what kinds of events you choose to analyze to find a set that gives you that false positive.

(the actual paper's abstract lists "aspects of global tectonics in the ocean basins and on continents, fluctuations in sea levels and resulting sedimentation patterns, pulses of active volcanism, variations in ocean/atmospheric composition and climatic changes, biodiversity, and mass-extinction events" as what they included, which doesn't make me worry less.

Also telephone sanitation engineers and hairdressers.

Given that accidents requiring evacuation on commercial jets are roughly one in a million take-offs, and accidents resulting in loss of life about one in ten million take-offs, it's unlikely that setting up practice centers would save any lives.

Also worth noting that this follows on from the earlier Northrup-Grumman "Mission Extension Vehicle", which in 2019 launched to dock and reposition Intelsat 901, a first for a telerobotically operated spacecraft.
https://en.wikipedia.org/wiki/...

Some sites:
  https://www.northropgrumman.co...
  https://news.northropgrumman.c...

Coincidentally, Northrup Grumman was also involved in the SWIFT reboost mission, launched the beginning of this month, tp extend the lifetime of the Swift orbital telescope by boosting it into a higher orbit. (In this case, though, they didn't build the spacecraft, just provided the launch service, on a Pegasus)

https://www.cbsnews.com/news/n...

The incident report's actual words were:
  The campaign was run by an autonomous agent framework ... executing many thousands of individual actions across a swarm of short-lived sandboxes,

Ooh! I love syllogisms.

1. LLMs cannot actually do things like that with proper protections in place.
2. A LLM did that.
3. Conclusion?

Conclusion: proper protections were not in place.

Bingo, I think you nailed it. Starbucks wants to sell you those fancy syrupy latte caramel drinks, and to get any coffee flavor at all after they add all the sugary sweet gunk and cream, they need a super dark roast.

If you just want coffee, not a desert-in-a-cup, go anywhere else. Your local indy coffeehouse, for example.

Simply not true. Beneficial effects of coffee have been known for decades. This work is simply confirming what previous studies have already shown.
  https://time.com/4116129/coffe...
  http://well.blogs.nytimes.com/...
  http://www.npr.org/sections/th...
  https://www.medicalnewstoday.c...

This mostly isn't new, just confirming what has been seen in studies going back decades https://time.com/4116129/coffe...

But an important question is, what do they define as a "cup"?

Years back, a "cup" of coffee was defined as a teacup, maybe five ounces. When I drink coffee, I use a coffee mug, which is perhaps twice the size of a teacup. Maybe more, for some of the extra-large coffee mugs.

I took a quick scan through the paper, searching on "cup", but didn't see an actual cup volume given.

Yeah... but interesting (and even potentially useful) spam. They are really cool vehicles; I'd want one even if they weren't solar.

Of course, the company has been promising them for years...