Matthew Chalmers
@matthewchalmers.bsky.social
Computer scientist into Ubicomp, HCI, theory and (a long time ago) data visualisation. Also kind of keen on mountain things, fine food things, and fine food in the mountains.
“People said we were using an unacceptable amount of power, and with these upgraded facilities that require so much more of it, we’ve shown that we took those criticisms to heart,” said Sam Altman
Technological Advancements Allow Data Centers To Run On Even More Electricity
SAN FRANCISCO—Hailing rapid developments in the infrastructure needed to support artificial intelligence, OpenAI CEO Sam Altman announced Tuesday that technological advancements were now allowing data...
theonion.com
Someone had just emailed me on a Friday night to ask if I have any thoughts to share on AI in digital govt and I think my answer is that we should focus on making basic services work for everyone and stop worrying about AI
I can see how the team at Anthropic might be getting off on feeling all Renaissance Man in making this kind of funding investment, but - without wishing to sound very unimaginative and old-school here - probably the last thing we need is for tech companies to enclose/colonise humanities research.
The reason I am particularly incensed by this is because Google's biggest "avoided emissions" claim (and they're all flimsy, but this is the worst) is from Google Earth specifically, where Google take 100% of the emissions reductions credit for any wind/solar project ketanjoshi.co/2026/07/01/g...
I am assuming literally zero people will read this footnote, so just in case you were interested in an incomplete list of people who literally cannot stop themselves from doing the thing that they have repeatedly raised public objections to.
Data centres are the horrific hyper-stimulated version of something once relatively harmless (like a golden retriever with rabies), but humanoid robots have always been and will always be a dangerous, expensive and resource-sucking hypey scam, and I will never stop enjoying when they fail
No amount of anecdotal “I used it for this one thing and it was useful” changes the huge sums that AI requires to exist, fixes the unprofitability of AI labs, or creates the $1.5tn+ in demand needed to saturate the 190GW of planned capacity. www.wheresyoured.at/the-more-you-buy-the-more-you-lose/
Such an interesting article. Hats off to the author, Lindsay Ems.
MIT Press@mitpress.bsky.social · last wk."We should strive — perhaps not unlike the Amish — to accept only the advances that meaningfully improve our lives."
Watching the world burn as all the money flows into a multitrillion-dollar Magic 8 Ball
Watching the world burn as all the money flows into a multitrillion-dollar Magic 8 Ball
On this week's episode of The Kettle, we welcome a special guest to discuss how the AI datacenter buildout is setting the world on fire – literally
theregister.com
Just think of all the pedagogical innovations and research on best practices for teaching over the decades that have gotten zero support and funding for implementation, but AI has no proven pedagogical value and is literally sucking up all the funding available in higher ed. These are facts.
How about a jet-lagged live posting while I read? Trying to read this because I keep getting tagged in posts about it, but the first paragraph already contains synthetic text (a quote from Claude). I will skip that bit and persevere... www.theguardian.com/technology/2...
Could AI be conscious?
Experts believe it’s at least possible. We urgently need a plan to navigate the ethical implications
theguardian.com
The anatomy of an AI-supported academic article is becoming more apparent. I desk reviewed ~70 manuscripts this week for our journal. Reckon 20ish had significant AI assistance. On top of hundreds of others recently I'm starting to see patterns. Are other editors are seeing the same or similar? 1/
Hacked source code reveals how Suno scraped decades worth of music and podcasts from the internet to train its AI tool.
Hack Reveals Suno AI Music Generator Scraped YouTube, Deezer, and Genius
Hacked source code reveals how Suno scraped decades worth of music and podcasts from the internet to train its AI tool.
404media.co
Irish datacenters now guzzle 23% of the country's electricity
Irish datacenters now guzzle 23% of the country's electricity
Consumption rose another 10% while restrictions on most new grid connections remained around Dublin
theregister.com
This paper keeps showing up on my feed. I'm just a guy who studies science but I don't think the message here is that LLMs are crazy powerful. We already know that LLMs aggregate opinions. And we know social scientists don't tend to test hypotheses thought to be 50-50 to go in either direction.
Large language models can predict the results of social science experiments - Nature
Large language models can be used to estimate the results of social science experiments about as accurately as a group of human forecasters—even for experiments published after the...
nature.com
So good!
M.J. Crockett@mjcrockett.bsky.social · 4w agoThis gorgeous essay from the social-media-less Lily Hu is everything: on what it means to "do" philosophy amidst pressure to automate it (with lessons that travel much more broadly). Rhymes with brilliant insights from @biblioracle.bsky.social @janerosenzweig.bsky.social Ted Chiang & others
China tells devs to ditch Claude Code over 'backdoor code' fears
China tells devs to ditch Claude Code over 'backdoor code' fears
National vulnerability database claims monitoring mechanism can forward Chinese users' data to remote servers
theregister.com
Childish, I know… but, also, that is admirable commitment to the dramatic role.
fesshole 🧻@fesshole.bsky.social · 4w agoI cheated on my boyfriend at a Halloween party, which is a confession in itself. However the guy was dressed as Shaggy and said "zoinks" when I took my top off, but I still went through with it.
Quick tutorial on how to opt out of Meta's new feature which allows anyone to tag your Instagram handle to generate AI images of you. @evystadium.bsky.social on Muse Image in 40 seconds:
The problem is not too little growth, but the wrong kind. What matters is not how much we grow, but what we grow — and for whom. Our response in @financialtimes.com to @jburnmurdoch.ft.com Bottom line: economic and climate justice require rethinking growth, not necessarily abandoning it.
"We document how tech companies are transforming established user interfaces to impose AI use and show how and to what extent these strategies fit within established deceptive pattern categories" arxiv.org/abs/2508.08672
Imposing AI: Deceptive design patterns against sustainability
Generative AI is being massively deployed in digital services, at a scale that will result in significant environmental harm. We document how tech companies are transforming established user interface...
arxiv.org
“As the public overestimate the costs of net zero by up to 14,000%, the idea that decarbonisation is an unaffordable luxury needs to be put to bed” www.theguardian.com/environment/...
A bargain price for reaching net zero in Britain | Letters
Letters: Dr Charlie Gardner says the public are vastly overestimating the cost of net zero. Plus letters from Moira Gommon and Diana Francis
theguardian.com
![Comic. [Back of car with bumper sticker reading “I forth [red heart emoji]”]](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:cz73r7iyiqn26upot4jtjdhk/bafkreie63wpgnvtc2bwmewc5o7vjndx7okspndmqn73ifodbzfzvs5nnyi)
