I agreed to do a film shoot tomorrow for a university promotion thing, where they wanted me to set up a telescope and look "professorial". I figured as long as I was setting up a telescope anyway, I might as well do some sidewalk astronomy with a solar filter for whoever walks by.

And then they sent out a university-wide email inviting everyone to come look through a telescope at the sun. I guess I'm going to set up TWO telescopes and hope that someone else will show up who can help me...

A big issue with LLMs:

It makes some people believe they are more competent than they are.

Example: this company sent me some spam. At first I thought it was a scam. Then I dug into the company.

And I realised: oh. No. This company is not a scam per se. No, these people actually believe their own bullshit.

The CEO ran a web server as root. It got owned. He posted about it on LinkedIn. He used Claude to "fix" it. It was not fixed.

Now you see, before LLMs, if you didn't understand a domain, you'd hit a wall very fast. Documentation would use terms and concepts you didn't understand. Tools would require interpretation. You'd have to ask somebody more experienced. That friction produces a very useful insight: "Oh. I don't know this domain very well." So you'd find someone who did, or learn.

The thing about LLMs is that they can remove an enormous amount of that friction. That friction is not noise, it is signal. That friction is telling you that you need to learn the domain you are working with. Removal of this friction is a very, very bad thing.

The LLM can hold your hand, do a lot of heavy lifting for you, use a lot of tools, run a lot of commands, and make things appear to work.

And the thing is, all of the individual outputs along the way might be pretty good and it might look correct.

The issue is in aggregate. Before you know it, you've built a house out of sand.

And worse yet, people can emerge from the experience with a dangerously incorrect higher-order conclusion: "I am now a domain expert in this."

They are not a domain expert. They understand absolutely nothing about it. The LLM did everything for them and it's wrong. And worse yet, they haven't reviewed it, because they lack the competency to review it. Therefore, they assume it is fine. But it's all incorrect. A true expert would see that it is wrong, and correct it, or push back on the LLM, or do it themselves.

And this is how you wind up running web servers as root and getting owned.

RE: https://tldr.nettime.org/@tante/117109914335818507

we went from "it doesn't matter, you should judge the output on its quality" to "weeeeeeh I'm going to be found out weeeeeh" real fast.

The job of management boils down to deciding where to invest company resources to maximize returns.

But saying yes to one thing means saying no to something else. Deciding what gets cut (and getting buy-in for those decisions) is hard! So instead they started agitating for "velocity" in a foolish belief that they would never have to prioritize one idea over another ever again.

https://productpicnic.beehiiv.com/p/no-amount-of-velocity-can-shield-your-team-from-opportunity-cost

#AI #LLM #tech #softwaredevelopment #UX #UXDesign

Let's talk cooking and data-viz, 2 things I love.
I don't think I've seen a recipe presented in this sort of "ingredient to process flow" diagram. I love how it's a very intuitive way to present a cooking recipe. Also, this sauce looks delicious.

https://joshnesbitt.cooking/xo-ish-sauce/

Recipes and website by Josh Nesbitt

#Cooking #DataViz

IA gets reduced to sitemaps and nav labels. Fifty years into the discipline, many people still don’t know what it’s actually for.

https://jarango.com/2026/08/07/whats-the-purpose-of-information-architecture/

#informationarchitecture #uxresearch #designsystems (https://uxbrad.com/activity/20260817/121500/)

RE: https://functional.cafe/@joomy/115749500002317829

This is my favorite "the old way meets the new way" story about tech.

I went to school in India, and we learned the multiplication tables with a chant that involved a postfix operator called "zaar":

two one zaar two
two two zaar four
two three zaar six
two four zaar eight
two five zaar ten
two six zaar twelve
…

We just accepted such a postfix operator as the most natural thing, and freely used it when talking to each other: “what is seven six zaar?” (42)

Some of us were adults before realizing (and some probably never realizing) that these “tables” we all sang as children had probably started as:

two ones are two
two twos are four
…

Racket - the Language-Oriented Programming Language - version 9.3 is now available from https://download.racket-lang.org

See https://blog.racket-lang.org/2026/08/racket-v9-3.html for the release announcement and highlights.

It's terrifying how accusations of plagiarism have come to be used as a weapon against scholars. Yes, sometimes there really is plagiarism, but it's usually discovered by a fellow scholar in the same field. If the accusations are coming from non-scholars who have a political agenda, then the evidence they present is probably really ambiguous, and they are making the accusation in order to silence or intimidate the scholar.

🔇 🔇 🔇 My book, Deep Unlearning, is now available for preorders! I don't know how people write so many books because this one took a lot out of me. It turns out, I had a lot to say 😂, with editors cutting out 50k of my words 🫣

✴️ You can now preorder it at https://www.simonandschuster.com/books/Deep-Unlearning/Timnit-Gebru/9781668093078

✴️ Is there a local bookstore that you think should order this book? Can you let me know, or, even better, let them know?

✴️ Is there a place I should plan to come for a book tour when the book is out? Can you lmk?

What we call "experience" is "residue" of encounter too. Our heuristics (what Rechtin calls contextual sense) are distilled from (reflection on) experience. Even our imagination is in good part residue of where we have stretched ourselves, reading, listening, watching, exploring playfully... And the "scar tissue" of experience, that warns and constrains us. And so on. Experience factors.

We bring that to the design table.

https://mastodon.social/@RuthMalan/116731374436270852

Learning to Imagine by Amy Kind

https://scispace.com/pdf/learning-to-imagine-37j5pkmd.pdf

Imagination works by combining, adding, subtracting and modifying material we've previously encountered.

RE: https://mastodon.social/@dahukanna/117099273860083300

Thank you @dahukanna — this (section of an) interview with Brian Eno on the genesis of oblique strategies is wonderful!

‘You tend to get lost in the details. Details are easy to deal with. Whereas to be able to step out of the situation, and look at it in a bigger way is hard to do [..]. So I would go into the studio with a list of things I wanted to do, and the first on the list was “honor thy error as a hidden intention”’

~ Brian Eno

“Black academic hounded to death by the press for plagiarism allegations started by a known racist” versus “AI companies owned by white billionaires pushing industrial plagiarism machines down everyone’s throats whether they want them or not”

I don’t see the lawyers or academics who let an LLM produce fake citations getting the same treatment. Remember kids, plagiarism is *super cool* if you use Claude to write your code, or ChatGPT to write that essay!

"you'll never need to look at source code again" is what someone says as a prelude to making the source code entirely untrustworthy

Picard leadership tip: Particularly if you're in a high profile position, be careful what buttons you press.

When someone has performed the same move 500 times and you have performed it 10 times, their speed feels like intelligence.