hot tip to students and junior devs: skip generative AI slop and hone your problem solving skills + programming ability on human written code. Guaranfuckingteed when the slop house roof caves in, it’ll be an appreciated skill*

* despite what people repeating the marketing propaganda of a company run by a eugenicist and financed by genocidal blood money tell you, it’s still an appreciated skill. Right now, in fact. It’s like doing your homework when all the other students are cheating: they might have a little more time now but mannnnn will they be fucked come exam time.

Well it has been 12 years since I have experienced this but I got laid off last week. Thankfully I have 10 months of employment still but it sucks as I loved this job and worked with one of the best teams ever. If you know of anyone looking for 20+ years experience with high tech, web dev, web design or training/project management or ethics compliance let me know. Feel free to boost. #FediHire #pdx #remotework

@anildash @iris_meredith @jalcine “Step right up! STEP RIGHT UP! To Machina the All-Knowing, Machina the All-Seeing! Machina the All-Understanding! Tell her your problems! Maybe she’ll answer them!! All for only ten cents a question!”

@kevinriggle there's an potential paradox here, the better LLM code gets, the more closely you have to look at it, because the bugs and insanity get harder to find.

It’s wild growing up under constant ridicule from the neurotypicals for being at the computer all the time instead of playing outside, to then two decades later witness them all anthropomorphize statistical models and act like they’re “friends”, not a vehicle of capitalisms need for control over the consumer.

I think that's the whole set.

@xgranade every industry once they get to market saturation does some combination of planned obsolescence, turning buying into renting and becoming an bank

when the social-economic maelstrom that is the current LLM hype wave fizzles or collapses there's going to be so much "how could so many smart people be so wrong?" and it's imperative you do not accept their framing of themselves as "smart people" - they are, in judgment and values and intellect, deeply compromised from living inside a machine that harms humanity for profit.

@futurebird In a sense, I suppose achieving long-distance space flight when you're trying to optimize for getting queens and drones to fly far and wide is more plausible than when you're trying to optimize for building better flint handaxes. ;-)

Tariffs on parts but not new devices? That’s a double standard. More people will choose replacement over repair. That means more e-waste, fewer domestic repair jobs, and more dependence on imported tech.

We need policies that fix the system, not punish fixers.

(Notwithstanding the rumored retaliatory tariffs, which are TBD.)

Full article: https://www.ifixit.com/News/110121/tariff-exemptions-are-terrible-news-for-electronics-repairability
—
#iFixit #RightToRepair

https://mltshp.com/p/1R225 “Everyone will not just”

"I've spent 20 years developing a pedagogy that's about wrestling with big ideas through writing and discussion, and that whole project has been evaporated by for-profit corporations who built their systems on stolen work. It's demoralizing."

Teachers are not OK: https://www.404media.co/teachers-are-not-ok-ai-chatgpt/

#AI #teaching #education #ChatGPT #LLM

If you were (wisely) offline over the weekend, I wrote about the ideology of AI and why we need to shift our stance in relation to it: https://aworkinglibrary.com/writing/toolmen

i'm looking for work as of today!

if you need a staff level backend/database person, i could very well be that person: https://tef.computer

if you know your company is hiring, drop me a link to the careers page.

nb: due to having a sleep cycle resembling a melatonin free jazz solo, i usually work in Eastern Time, remotely.

as a result i generally work for east/west coast firms in the us, or very flexible employers in the uk

New blog post: Pattern machines that we don’t understand

https://surfingcomplexity.blog/2025/06/01/pattern-machines-that-we-dont-understand/

@trochee @jalefkowit someone fixes the bug, and McCartney lets out of gasp of thanks before crumbling into dust (as a motif based on "Yesterday" plays in the background.)

This is the story of a fifteen minute mini-lesson, where I had students probe ChatGPT on topics where they are experts.

It exposed to them the tendency of the AI to just make s**t up on topics it hasn't been trained.

https://buttondown.com/natebowling/archive/oooh-these-ais-be-lying/

Exams and homework are kinda bullshit, but I think it's an interesting intellectual exercise to consider what kind of exam question you could ask to hypothetically assess prompt engineering knowledge. What kind of question would be fair to ask in that setting (well, as fair as exams ever are)? How would you know if a student understands the difference between a good and a bad prompt, and *why* one is better than the other for some task?

https://wandering.shop/@xgranade/114607571719781514