Remote
Greg Wilson
@gvwilson@mastodon.social
mastodon 4.8.0-nightly.2026-10-06I program, write, and teach. Co-founder of Software Carpentry and It Will Never Work in Theory; co-editor of The Architecture of Open Source Applications.
5661 Followers
141 Following
50 Posts
Joined October 28, 2022
Personal site:
Teaching Tech Together:
The Architecture of Open Source Applications:
I just learned that oat milk was discovered in the 1960s, when NASA found out by accident what zero-g environments do to goat milk.
Thank you, thank you, I'll see myself out.
Boosted by @trending@homestead.social
RE: https://mastodon.social/@Carl_Zimmer/117345065655681591
I think you have to be very naïve to believe that AI providers aren't (at least occasionally) skimming information from users' prompts.
Quoting
Four days ago, a scientist was surprised by the news that Anthropic discovered new virus genes for making DNA. He says he’s been studying them for years—and feeding his data to Anthropic’s AI as part of his own research. Coincidence? Here’s my story. https://nyti.ms/3VSWayc
Open quoted postBoosted by @trending@homestead.social
parton (n.): unit of measure for the extent to which a wealthy person uses that wealth to make the world a better place. Sadly, the milliparton and microparton are more commonly used. Note that negative values are frequently observed.
Boosted by @trending@homestead.social
In his essay on Salvador Dali, Orwell argued that because Dali was a repulsive human being, the right wouldn't admit that he was a great artist; conversely, because he was a great artist, the left wouldn't admit he was a repulsive human being. I'm seeing the same thing with AI: because it's unethical, one side won't acknowledge that it's useful, but because it's useful, the other side won't acknowledge that it's unethical.
Time to make another cup of tea…
Boosted by @trending@homestead.social
Happy Skynet self-awareness day, for those who celebrate: https://www.youtube.com/watch?v=qxW0yGvWRTk
Boosted by @trending@homestead.social
Looking for source of quote: "The real purpose of AI is to give the wealthy access to skill without giving the skilled access to wealth." thanks in advance (please note: not in search of commentary at present time)
Boosted by @joe@f.duriansoftware.com
Ma et al 2026: "Do Coding Agents Reuse Existing Code or Reinvent the Wheel?" https://arxiv.org/abs/2609.35357 "An audit over 3,000 turns shows that agents progressively stop exploring relevant repository code, reuse their own history less even when it is fully in the workspace, and leave duplicated logic in 50.8% of task chains by turn 5 - all while pass rates barely move."
Please help if you can: I am working with some students who are studying programmers' beliefs about software engineering folklore. If you can spare a few minutes to answer the question in https://survey.bth.se/survey/2545 we would be very grateful. We would also be grateful if you could circulate the survey link to colleagues and friends - we would like to reach as diverse a demographic as possible. Thanks in advance.
"I was only following prompts." - Skynet
No, this isn't a joke - it really happened today.
Starting Aug 17, Atlassian gives itself the right to scrape your data for AI/ML training purposes: https://letsdatascience.com/news/atlassian-enables-default-data-collection-to-train-ai-f71343d8
When exactly did setting up a dev environment on a new machine become functionally indistinguishable from solving crossword puzzles in Ancient Sumerian?
Suppose you discovered that AI transcription mistakenly ascribed points made by women to male participants in meetings 40% of the time - would you continue to use it? What if it was 4%? Or 0.4% (i.e., once every 250 meetings)? Do you and/or your employer have a threshold or SLA? Are you/they even thinking in those terms?
I had the privilege of reading a pre-release draft of @grimalkina@mastodon.social 's "The Psychology of Software Teams". Serendipitously, I followed it with Sarah Stein Lubrano's "Don't Talk About Politics", which is actually about the psychology of how people change their minds (or don't). The combination gave me a _lot_ to think about, and I highly recommend both.
How is this legal? https://www.schneier.com/blog/archives/2026/07/mit-to-become-hotbed-of-ai-video-surveillance.html
edit: how would you feel if the press release from the university was, "Once these cameras are installed, we will be able to track the identity and position of every woman on campus in real time."
Replying to
@wendynather@infosec.exchange I'm really liking people framing this (and social media) as "cognitive pollution". Don't know who said it first, but we don't pass laws to stop kids from eating paint with lead in it - we ban lead in paint.
"Most dependency strategies are attempts to avoid interpersonal negotiation. The tooling doesn’t remove the disagreement, it just picks who loses by default." https://nesbitt.io/2026/07/10/package-management-as-org-chart.html
Boosted by @joe@f.duriansoftware.com
RE: https://social.afront.org/@stylus/116991865079758024
There is an alternate timeline in which the Jupyter notebook was built in OpenOffice instead of in a browser, so that people could (for example) draw diagrams directly in the Office-y parts of the doc beside their embedded code cells. It wouldn't have played nicely with version control, but neither (in practice) does Jupyter's fastidious JSON and embedded images. I still wonder sometimes what kinds of coding environments would have evolved from that…
I remember reading a science fiction story in the late 70s by Poul Anderson in which an aging capitalist and an aging communist come to blows in a bar in a small town somewhere in the not-very-distant future, arguing about where it all went wrong. Turns out that as soon as solar became really, really cheap, both of their economic models just…went away, along with the belief systems they'd sustained. The younger townspeople don't even really understand what they're arguing about…
On December 22, 1967, Spock asked the Enterprise computer to calculate pi to the last digit in order to paralyze it (episode 2x07, "Wolf in the Fold"). Is this the first example of prompt engineering in modern fiction?
I was moderately useful to other people today.
"Passwords, GPS coordinates and private conversations have all been found in the metadata and source files of millions of preprint submissions." https://www.nature.com/articles/d41586-026-02057-8
"People dramatically underestimate what thousands of organized humans can accomplish when they are adequately fed, aggressively supervised, and denied alternative career paths." https://www.mcsweeneys.net/articles/notes-from-a-tired-egyptian-guy-whose-job-is-explaining-that-humans-built-the-pyramids
Boosted by @joe@f.duriansoftware.com
Mori & Hashimoto 2026: "On the Correctness of Software Merge" https://arxiv.org/abs/2607.07987
Enjoyed Postrel's "The Fabric of Civilization" https://isbnsearch.org/isbn/9781541617612 - good popular science / history, and some thought-provoking discussion of (for example) how the needs of weaving may have inspired development of number theory in classical times. ☆☆☆
"This is not a software business that happens to own servers. It is a real estate business that happens to compute." https://www.groundbrkr.com/p/the-second-derivative-why-no-one
Replying to
@jvschrag@hachyderm.io my daughter's reply to this was, "You can do this yourself with reddit." I'm so proud.
Replying to
Today, LLMs are spitting out shell commands that use all the available tools, because that's what their training data does. I wonder how long it will be until we see a new, more cohesive set of command-line tools that truly follow the Unix philosophy: each does exactly one thing, RISC-style, and the power comes from ever-greater composition. That, plus the "everything is structured data" philosophy of Nushell, would be an interesting future (albeit unlikely). 2/
edit: OK, so it's a way to hand government money to their more radical supporters. That (unfortunately) makes sense.
Replying to
@MarcParadise@hachyderm.io results and analysis will be posted as soon as we have 'em - thank you.
Zięba-Kozarzewski 2026: "No Edges, No Verdict: A Large-Scale Empirical Study of Declared Dependency Graphs in 78K SBOMs in the Wild" https://arxiv.org/abs/2607.22140 "52.9% of SBOMs declare no edges at all, 8.8% declare a dependency block yet leave the majority of components isolated, and 38.3% form well-connected graphs."
Canada's government recently announced billions of dollars to support development of a new pipeline to move the dirtiest crude oil in the world to market: https://www.cbc.ca/news/science/wildfire-climate-emissions-smoke-9.7274862
RE: https://mastodon.social/@gvwilson/117043030559498832
Thanks to everyone who has responded / is responding / responds in future. Several of you have had questions or comments about the survey's content - I can't answer them now for fear of biasing responses, but I'm taking notes and will do so once the survey closes. The results and analysis will be published (open access) as soon as we have them. Thanks again.
Quoting
Please help if you can: I am working with some students who are studying programmers' beliefs about software engineering folklore. If you can spare a few minutes to answer the question in https://survey.bth.se/survey/2545 we would be very grateful. We would also be grateful if you could circulate the survey link to colleagues and friends - we would like to reach as diverse a demographic as possible. Thanks in advance.
Open quoted postA: "Every time Claude writes an email, a butterfly dies somewhere in the Amazon."
B: "Luckily, I'm using Copilot."
I wish I knew enough to write a compare-and-contrast looking at how educators responded to the sudden onset of "work from home" during COVID and how they're responding to AI.
"Nothing for us, without us" is the best and also most often ignored advice in, well, everything: https://www.cbc.ca/news/canada/toronto/finch-west-lrt-accessibility-concerns-9.7276886
Looking for the source/data for the claim that approximately 1/3 of backups fail to restore - is it just folklore or does someone have a pointer to something more substantial? Thanks in advance.
"A cocaine skeptic starts an unwise habit" https://blog.djnavarro.net/posts/2026-07-26_starting-a-cocaine-habit/ is a long, thoughtful piece about AI and programming - worth reading carefully. cc @djnavarro@hachyderm.io
TIL: second-hand child seats for cars are safe - we think otherwise because of a PR campaign by manufacturers who wanted to kill the resale market. https://isbnsearch.org/isbn/9781635570113
Just learned that "slub" is actually a word.
"Disclosure Day" is an incoherent, self-indulgent mess. There are some good performances - in particular, Emily Blunt manages to create some moments - but bleah…
manufacturing preferences: https://third-bit.com/2026/05/14/manufacturing-preferences/
Replying to
@wendynather@infosec.exchange years ago, @mhoye@cosocial.ca suggested a feature for Linux that would retire certain usernames like 'dmr', the same way that teams retire jersey numbers.
Would you like to have a significant impact on the tech industry? Do you have $100K to spend? Read on… https://third-bit.com/2026/06/29/a-modest-proposal/
Replying to