Willem Huiskamp

Normal people: "Why are we wasting all these resources for this slop?"
Google: "Drain that lake, we need to Guarrmocation scrretary development"

Yeah, it's long been a problem that once you leave somewhere you immediately lose IT access. Would be nice to have such a solution, but often they don't really care - you're gone and no longer their problem.

Food for thought to those out there using/paying for these services.

Don't know tbh. When I first started here as a postdoc I sent a few things via my gmail and was told quite clearly that that wasn't OK. Guess it's one of those things they don't make much effort in enforcing.

That definitely would not fly here, given our data security requirements.

I just don't get it. I feel like I'm losing my mind watching so many others around me embrace this crap.

Quite literally the official chatbot of the war in Iran

"Cancel ChatGPT" movement goes big after OpenAI's latest move After Anthropic refused flat out to agree to apply Claude AI to autonomous weapons and mass surveillance of American citizens, OpenAI jumps right into bed with the United States Department of War.

You love to see it

Useless

One I stumbled across the other day from a colleague of mine back at UNSW; CounterCurrent AI - uses machine learning and forecasting to make shipping more efficient (no idea re: the effectiveness, from what I understand they haven't been around long, but a neat idea).

🌊 New paper in @natclimate.nature.com: 'Mapping tipping risks from Antarctic ice basins under global warming', by @pik-potsdam.bsky.social

"Marine-based sectors in East Antarctica, representing ~5 m of potential sea-level rise, are at risk of losing stability at 2–5 °C."

▶️ rdcu.be/e5bQS

A screenshot of prompting chatgpt with "how many new clusters were detected by hunt and reffert". It invents a first author for a paper that doesn't exist and returns a wrong answer (five) instead of ~2300.

I'm starting to find that papers mis-cite my own papers in really, really perplexing ways. It's quite obvious why if you just go to that one hyped up tool and ask it about your own work...

ChatGPT got my supervisor's name right but changed me to be a "William" (lol) and gave the wrong number

YouTube video by LastWeekTonight Twitter: Last Week Tonight with John Oliver (HBO)

Great timing. John Oliver says the same thing over half an hour:

Reason #29407 to stop legitimising X....

The utterance of "A.I.'s inevitability" is one of the most stark pure performatives I've seen in my time working in higher ed. Every time it is uttered, it is clearly not reporting a fact about the world but instead actively trying to create the reality it narrates. We can and must refuse.

Unflinching refusal to use or allow the use of A.I. in reading & writing should be the uncontroversial, default position of teachers from K-12 to colleges & universities. If our administrators tell us we "have to" we tell them to eff off. We will not kill the planet's water for systemic plagiarism.

ICYMI,
THIS is a great example of why supporting independent journalism is so crucial.

Mainstream media is on a unity ticket in enabling - or at the very best, failing to question - authoritarian creep*.

[*like Minns and Albanese] #nswpol #auspol

This is the kind of thing a rich CEO says at the beginning of a science fiction movie to establish for the audience that he is a sociopath and the villain.

The vehicular equivalent to a cringy neoliberal half-baked "solution"

I'm always surprised, as this was already being beaten into us 10+ years ago back when we were doing our PhDs...

Is it bad that my first thought was: "How did the rainbow colour palette get published in nature in 2026?"

This is completely correct other than the infrastructure not existing. I think people need to realize WHY this is happening is because, for example, NVIDIA sells GB200 NVL72s (racks of 72 GPUs) with *17TB of RAM* and likely over 500TB of storage.

YouTube video by Channel 4 News Collapse of this Atlantic current could change Europe's climate forever

Great Channel 4 News report on the latest #AMOC monitoring.
"Trouble is: there is now growing scientific consensus that the AMOC is weakening."
@noc.ac.uk
youtu.be/JpOpsMOBsL4?...

My friend Jon has died | First Dog on the Moon Jon Kudelka loved his beautiful family and was fiercely intolerant of greed and hypocrisy

Superb by @firstdogonthemoon.bsky.social

www.theguardian.com/commentisfre...

💯

Life comes at you fast

Keep your eyes on the ball. An ineffective opposition means Labor can get away with such brazen climate vandalism.

R.I.P Jon Kudelka @kudelka.bsky.social

“If I did a sequel it would probably involve a scientist swearing a great deal”

allouryesterdays.info/2025/04/28/i... @allouryesterdays.bsky.social

OpenAI crashing *this* year?! Don't make me hope...

A study at U Mass Amherst finds that Elon Musk manipulated Twitter to bolster the voice of the Neo Nazi Alternative für Deutschland (AfD) before the German 2025 parliamentary elections. No surprises here. groups.cs.umass.edu/equate-ml/20...