A thought experiment occurred to me today.
Suppose you gave claude a dataset and asked it make a function modeling the outputs; let's say, GMO use in corn grown in Iowa. (I'm tweaking something I'll credit in this thread)
And you get a result like this. Hey, that looks pretty good! p < 0.01
Vibe coding broke people's brains because they had a bad understanding of the software design process.
The pop culture model goes something like this: start out with a sketch and then render that same idea in progressively finer detail.
So when tools could "skip" to high detail, people went WOW.
And BOOM goes the dynamite. To all the AI Bros who have slid into my mentions to tell me that you CAN in fact copyright AI materials, the Supreme Court has told you to go suck rocks. A prompt is not authorship. And if there is no author, there is no valid copyright.
www.engadget.com/ai/the-supre...
AI doesn't actually take jobs; it is just used as an excuse to avoid backfilling them.
In addition to demolishing organizational knowledge, this also kills off opportunities for junior ICs *and* junior managers to get experience. Instead, all roles are reduced to mindless efficiency hunting.
Cisco said there are no workarounds for the vulnerability and urged customers to apply available patches immediately. The company also recommended reviewing system logs, validating controller integrity, and implementing additional hardening measures where possible.
www.csoonline.com/article/4137...
Users hate AI. So tech made it mandatory. Even if you don't use it, it pollutes what you read and how systems make decisions. The computer's hallucinated word is final.
And it needs all of your data to do it.
In the age of no consent, UX exists to normalized complete acquiescence to surveillance:
"15+ years later, Microsoft morged my diagram"
nvie.com/posts/15-yea...
this becomes less funny when you remember every Microsoft employee is mandated to morge 🫠
“As some universities and journals adopt AI-text detectors, they risk creating a feedback loop that constrains how authors express themselves. I have seen colleagues intentionally simplify their grammar or break down complex rhetorical structures to avoid arousing algorithmic suspicion.”
The fact that these sentences always end with "... which we will cram with more work for them" and not "... so we moved to a four-day workweek" really illustrate the hollowness of some AI evangelists' promises that AI will deliver us to some kind of post-work utopia.
skeet from Simon Willison On the @oxide.computer and friends podcast last month we (primary credit @ahl.bsky.social coined the term "Deep Blue" for the sense of psychological ennui leading into existential dread that many software developers are feeling thanks to llms right now
AI enthusiast podcasters recreating labor alienation from first principles while repeatedly pulling the lever on the labor alienation machine
Instead of regurgitating the bromide that LLMs are just "autocomplete on steroids" (even by people who know better), maybe we can actually engage in some public education. The problem with genAI is better expressed through a classic computer science concept, known as SYMBOL GROUNDING. 🧵
Preliminary results show that the current framework of "AI" makes ppl less likely to help or seek help from other humans, or to seek to soothe conflict, and that people actively prefer that framework to any others, literally serving to make them more dependent on it.
The rise of Gen AI use in professions like teaching and librarianship is connected to the devaluing of those fields and the expectation to do more with less every year. AI offers people a fast and cheap (or free) solution to a problem whose real solution is more funding and resources.
Right, another key thing that historians do is working with important records that have not been digitized.
And the vast majority of the world’s records *have not been digitized* and thus do not exist in any format that LLMs/AI can work with.
I’m a big fan of sandwiches. I find myself turning to them a lot when I’m hungry.
This is what I deal with at work too
Here is one of the dumbest things about vibe coding that I don't see a lot of people talking about
Imagine you're an open source maintainer and someone opens up a pull request against your repository that they vibe coded. They may or may not tell you they vibe coded it; it doesn't really matter… 🧵
Sorry just jumping in mid thread, but my local game store owner talked about
agency + person = game
But I believe they are a critical part of the relationships that we develop among our colleagues. If we use LLMs, we pollute that channel, and I think that matters.
We aren't robots, and work is not just about tasks.
I read a great piece by Erik Johannes Husom (erikjohannes.no/posts/202601...). I resonated with it.
One part that I think is under-appreciated is what we lose in business communication when all of it gets laundered through an AI. All of the small talk and emojis and back and forth may seem trivial
My pull-string Sheriff Woody doll said “there’s a snake in my boot” when there was not in fact a snake in his boot
“The affordances of AI systems have the effect of eroding expertise, short-circuiting decision-making, + isolating people from each other. These systems are anathema to the kind of evolution, transparency, cooperation, + accountability that give vital institutions their purpose and sustainability.”
What? The AI companies have misrepresented what they're doing and how it was done? That's unpossible!
I keep catching strays from "technology is neutral" people, so let's set things straight.
1) Technology is not neutral, especially the subset of technology called "algorithms"
2) The form-factor that makes a technology into a tool is not neutral
3) How tools are productized is not neutral
[rant]
Yeah it's super easy to knock out sprints worth of code in a day if you don't care if anyone else understands it or can maintain it sure
Really enjoyed this fantastic piece by @sonjadrimmer.bsky.social and @cnygren.bsky.social
Here are few favorite pull quotes of mine + one little quibble at the end:
There’s this thing I like to call LinkedIn writing where people describe some normal ass daily experience as if they’re telling a grand parable that always has to end with how this experience is actually good for the business grindset mindset
this is true across all fields now and it didn’t used to be *so* prevalent
i’ve been calling it “Deliverable-itis”
So interesting, former colleague drew this same analogy:)
medium.com/venturehq/wi...
