I have both airPods Pro Gen2 and a recent version of medically fitted hearing aids.

  TL;DR: Airpods are better than nothing as pseudo-hearing-aides and cheaper than my prescription hearing aids

  The Pro's were A LOT cheaper (more than a factor of 10) and the pros have the additional nicety of active noise suppression but:

          My hearing aids work better for speech: which can be literally life or death (or at least employed vs fired) when mishearing something important
          I regard that improved speech for hearing aids as worth the money

          I chose a more open air flow for my hearing aids in order to better deal with playing instruments and situational awareness

          The hearing aids are smaller and more comfortable (although bluetooth is a little iffy from time to time) and have longer lasting batteries

The pros are very nice in computer machine rooms (they handle RF interference and noise suppression) and are better than nothing for speech.

The balance between them may change in the immediate future with recent changes to hearing aid prices and a new tech . Your mileage may vary.

In the 1980s there was a project hosted at ISI called "MOSIS", which supported researchers around the country who wanted to create their own experimental chips according to their own specifications. They would send chip descriptor files to MOSIS, and MOSIS would grind out a run of chips according to those specs.

Unfortunately, some of those researchers were connected via CSNET, the world's first ISP, where I worked at the time. CSNET did provide reasonably high-speed IP access to customers who could afford it, but these were broke-ass researchers we're talking about here, and they could only afford dial-up email. Now, MOSIS did support accepting chip description files by email, but these were three megabyte monsters and dialup did not like them AT ALL. They took 3-4 hours to transfer over the phone, and something would go wrong and the phone call would drop nearly every time. We had to rip those monster files out of the queue by hand to get communications with the researcher's site working again. I remember those days not so fondly, but silver lining: this sort of faux pas was responsible for the "Ask Mr. Protocol" column in CSNET's newsletter, which in turn gave rise to a column of the same name that ran in the trade press for twelve years thereafter.

I wonder if they're going to use the SCIPS interplanetary internet protocol Vint Cerf has been pushing, or regular TCP/IP? If the latter, they're going to have to modify a few timers, I bet...

I momentarily read the article title as, "Humans Having Sex With Neanderthals Gave Us Protection Against Ancient Academics". This is plausible on so many levels that I rather think someone should write that article.

We have a type of surge protector/powerboard sold commonly in California called a âoeSquidâ. It resembles a squid: it has an oval body and five short cords of varying lengths, each one ending in a female power socket. It can power five wall warts at once regardless of their egregious size.

Apple rather slickly has each update of each recent iOS be specific to a phone. ONE physical phone. Probably to prevent the skeleton key scenario.

  Each "copy" (not really an appropriate word here) of the update is unique (I don't know the details) which makes it hard to just use the same binary to on every phone. Each "copy" only works on one phone.

Just a few thoughts:

There is a real advantage to back in the day: parametric statistics needs calculus but a lot of modern statistics
are more simulation-based so that could be stressed. Easier to understand (or at least less difficult) and usually
more accurate. Parametric statistics can wait.

Some appropriate subset of "How to Lie with Statistics" might be apropos early on and throughout the time spent.
It's practical information in life, gives a deeper understanding and is relatively fun. Care needs to be taken that this
isn't taken as "All Statistics Lie".

Consider bringing in language teachers for help. The words in statistics often have a subtle (or huge) difference
from common usage and they may be able to help with that. I had a mathematics background when I started statistics
and wasted a lot of time in early days because "variable" meant something different than what I was used to.

I've been doing UNIX since about 1974. I started out on research version 5 on a PDP-11, because that's the only architecture UNIX ran on in those days. v6 was the version that was much more widely distributed to academics, and v7 was the even more widely distributed update that led to the BSD derivatives.

v5 was pretty damn raw. There were no shell variables. "ed" was still written in assembler. Etc. Uphill through the snow both ways. Still, it was FAR better than any of the vendor OSes, no matter what people say about RSX-11. So I founded the first UNIX User's Group Software Distribution Center, purely so I could get my hands on all that goody-poo software. I also produced the very first T-shirt with a UNIX demon on it, for the Urbana, Ill. UNIX meeting - the first national meeting of UNIX users. I gave one to Ken Thompson, one to Dennis Ritchie, and kept two for myself. I still have them. If you've ever seen early USENIX T-shirts with a PDP-11 with pipes, demons, pitchforks, and a barrel labeled NULL, well, that was me (art by Phil Foglio to my design).

RealNetworks has a service, "RealTime", that works across most platforms for doing just this kind of thing. The use case explained to me was a family sharing video where much of the family is Windows but also there is a Unix person, a few Apple-or-nothing, and another few "Android All the Way".

Since this indeed describes my extended family, it was helpful.

> No company would want to give north korea or isis a reason to actively target undersea cables.

  It might give them MORE reasons but, given my understanding of their agenda, those two (at least) already have reasons
  to want a capability of targeting undersea cables.

  A bit more in the sensor department might at least give warning that something is about to be damaged
  and info about the thing doing the damaging.

Letters of Marque and Reprisal, as I've heard it. And "reprisal" is certainly closer to the mark (no pun intended).

It might be worth looking at INGRES (not INGRESS) and their business history.

They've managed to stay in business as long as Unix has had commercial databases (albeit with a close call every decade or so) and currently are in a similar position to yours but are still standing and still innovating. Many lessons in tech company survival, some of them of the "don't do that" variety to be learned there.

Sadly, the people that know the name tend to think of Ingres as what it was way back in the 70's-80's, not what it is, so the company is called "Actian" these days.

  I wonder if sleep apnea is considered dementia in this context?

  Sleep apnea is highly correlated with obesity at that age and it can give the sufferer a disturbingly
  similar experience to senile dementia when severe and untreated.

  A competent epidemiologist would control for the "They die before they develop dementia" effect.

  Given this is a peer reviewed study I think it hugely likely they controlled for that.

  Having gone through some hand nerve damage over the years I found it useful to have more than 1 keyboard;
  having my hands in different positions during the day has proved useful to preventing further problems.

  Most of my typing is done on a Unicomp Model M, which is very much an IBM Model M. I type most
  reliably there although the I can type longer at the Kinesis models and probably would be almost as fast if there
  if I really devoted the effort to it.

I find the clickety-clack of Model M type switches cheerful and I need all the happy thoughts possible when
debugging Ansible scripts.

I probably wouldn't inflict that on people in an open office.