After the CEO of Anthropic claimed "AI could surpass almost all humans at almost everything" by not much later than 2027, and I stopped laughing, I asked Claude what it thought.
Claude clearly can't overhype like a CEO can.
www.theverge.com/2025/1/21/24...
File under "what could possibly go wrong" given the government's track record in this kind of thing.
excellent reading: "Falsehoods programmers believe about time, in a single list" [apologies if I shared it already. don't remember]
gist.github.com/timvisee/fcd...
So far, the sabbatical has involved finishing off a whole bunch of admin tasks from pre-sabbatical, and starting work on three grant proposals. And freezing my butt off.
I'm on sabbatical in calendar year 2025. I've been a full-time academic since 1997 and have never really had a proper sabbatical. This will be weird.
It's accreditation visit day, as our integrated biomedical engineering programs are being reviewed. The coffee is plentiful and, sadly, appalling.
The signal-to-noise ratio on Twitter is dropping, so I'm largely moving here. Expect some rants about software, safety, cats and HR.