Few things drive home the fact that I probably don't belong in tech any more is the fact that whenever some of the smartest people (like, actually smart) chat online I basically end up disagreeing with everything from the underlying premise up to the high-level observations.

NEW – Factcheck: How nuclear, gas, wind and solar power are affected during heatwaves | @mollylempriere.carbonbrief.org w/ comment from @iain-staffell.bsky.social @henrypreston.bsky.social Read here: buff.ly/yfvQnUS

Today is the day the story There Will Come Soft Rain is set. An amazing and heartbreaking work of fiction for our modern age. Read it here:

a wooly man without a face@jojororo.bsky.social · yesterday

archive.org/details/brad...

"A Revised #Haskell 2010 Language Report" by David Binder blog.haskell.org/revised-hask...

A Revised Haskell 2010 Language Report | The Haskell Programming Language's blog

blog.haskell.org

It's a bad time to be named Al.

Congrats Austria for the highest-ever measured temperature there! (105.8°F/41ºC in Vienna today). Here in the DESERT (latitude 32ºN btw, Vienna is 48ºN) it will reach 104ºF today.

To be a leader in Tech these days, one must have the vision, the bravery, the raw intelligence to look at something that works and think, "How can we make this both worse and more expensive?"

Plenty of libraries hand you low-level GPU APIs. A high-level, functional one is a different story, and that's the gap Cyfra was built to fill. Join Szymon Rodziewicz at #ScalaDays 2026 for a look at how it works. Read more: scaladays.org/sessio... 12-13 October. Berlin.

The edge of one's ability is where pathways to new skills are built.

ffs

"Pandoc currently supports 51 input formats and 76 output formats, thus 3876 distinct conversions (not counting the variants that are possible by adjusting extensions)." ← wild

Haskell programming language@haskell.org · 2d ago

"Twenty Years of Pandoc" by John MacFarlane pandoc.org/twenty-years... #Haskell

This is fun: BoxSquad. Limbs & borders are all draggable. [spacetypegenerator.com]

STG v.boxSquad

spacetypegenerator.com

the allegory i use is like: • imagine you're an artist in ancient greece/rome/whatever • say you want to paint a painting • so you mine the pigments (paint) • grow a field of flax (canvas) • chop down a tree (frame) • kill a wild boar (brush bristles) now: what was it were you gonna paint?

Tomasz Godzik believes tooling can be great again, and he can prove it. Join him at #ScalaDays 2026 for a look at Metals 2: indexing rebuilt to stay fast, and keep working even when part of the project won't compile. Read more: scaladays.org/sessio... 12-13 October. Berlin.

Shadowrun Anarchy 2.0 was not on my radar (and I missed the campaign) but I couldn't be more excited. One of the best TTRPG worlds ever created. Here's hoping that this rules variant / iteration gets the popularity the setting deserves.

Periodic reminder: The integration of generative AI models into every sector of society is the installation of ideological and epistemic choke points, regardless of who made the model.

World Politics Review@wpr.bsky.social · 3d ago

The spread of Chinese AI models with built-in censorship mechanisms could quietly shape the global information environment for the worse.

In response to the unprecedented number of game developers laid off this year, we have created a bundle on itch to benefit a Game Industry Hardship Fund, managed by @videogameworkers.bsky.social. More than 100 games for 10+ bucks! Please share widely and donate generously! itch.io/b/3802/game-...

It looks all cute and beautiful now... but wait until the bunny is done eating. 🐰

daily bunnies 🐇✨@dailybunnies.bsky.social · 3d ago

Going on vacation or a conference? Russian state cyberattacks are targeting travelers through the hotel, airport, and event Wi-Fi networks. The actor has been compromising captive portals, the login pages that appear before a device can access guest Wi-Fi. www.microsoft.com/en-us/securi...

An added wrinkle to this debate and one reason why so many of us are super-sceptical about managers and executives who seem to mostly be saying the right thing is we all remember how reasonable the 37 Signals guys sounded before they decided that fascism was a better strategy

Baldur Bjarnason@baldurbjarnason.com · 3d ago

Anyway, since I’m always annoying my own brain by trying to come up with clearer and more specific wording for my own opinions: I reject the notion that LLMs can be deployed “responsibly” at scale and believe that anybody claiming otherwise is fooling themselves, conning you, or possibly both

When we stop doing things for rich people and start doing them for each other it’ll be mind blowing

By Cory Doctorow (GPG 0xBF3D9110957E5F4C)@doctorow.pluralistic.net · 4d ago

They demonstrate the adage that worker-driven automation improves quality, while capital-driven automation improves *throughput*: pluralistic.net/2026/07/28/h... 70/

To underscore Suresh’s point, I’ve had discussions around the shortcomings of LLMs with senior leaders in my company who fully agree, but are afraid to say so internally because they will get shitcanned. This is not how any functioning business should work.

By Cory Doctorow (GPG 0xBF3D9110957E5F4C)@doctorow.pluralistic.net · 4d ago

Suresh reports on his experience talking with hundreds of executives (and, more importantly, their subordinates) about what (if anything) AI is doing for business in an essay entitled "AI Mania Is Eviscerating Global Decisionmaking": hermit-tech.com/blog/ai-mani... 11/

It is funny that as more people get duped by AI, the response more frequently seems to be to insist that the models are developing godlike intelligence and we all need to be sooo careful when the real lesson is clearly that the human psyche can get absolutely one-shotted by personalized flattery

Anyway, since I’m always annoying my own brain by trying to come up with clearer and more specific wording for my own opinions: I reject the notion that LLMs can be deployed “responsibly” at scale and believe that anybody claiming otherwise is fooling themselves, conning you, or possibly both

"I really don’t understand why somebody would want to make me say things that I don’t want to say as they please and then sell it as my own voice. It’s heartbreaking. Why can’t Japan, which calls itself a content superpower, establish rules for content use or tell the world,..."

The voice of Shinji in Evangelion says AI voice cloning is 'heartbreaking,' calls for Japanese government to intervene

It would be nice if somebody did something, for once.

pcgamer.com

i keep getting surprised by just how little the average programmer actually knows about programming

is there anyone that actually understands how rice's theorem works omfg rice's theorem does not say that you can't prevent bad things from happening in a turing complete programming language.

If not for anything else, Logic Pro's session "players" are a handy tool for practicing your instruments. Punch in some chords, select a couple of session players and you have a usable backing track to practice along with.

The fact that you're "in charge" is hard to reconcile with the fact that the people you're nominally in charge *of* tell you that all your ideas are impossible, illegal, immoral, or lethal: pluralistic.net/2026/01/05/f... 62/