Benjamin Geer

Software developer and ex-academic (sociology, conceptual history, Arabic culture, etc.), sometimes publishes classical sheet music, lives in Paris.

Développeur de logiciels et ancien universitaire (sociologie, histoire conceptuelle, culture arabe, etc.), publie parfois des partitions de musique classique, vit à Paris.

en/fr/ar/de/it

Blog: https://benjamingeer.name/en/
ORCID: https://orcid.org/0000-0002-2449-8558

Public Key

npub19jscweanp4p20a3fjes3tsl5crnzwaxdywvjs8m63pqm0rzqswss4gsvpy

NIP-05 Address

benjamingeer_at_piaille.fr@momostr.pink

Profile Code

nprofile1qqszegv8v7es6s487c5evcg4c06vpe38wnxj8xfgraagssdh33qg8ggpz3mhxue69uhhyetvv9ujuerpd46hxtnfduqs6amnwvaz7tmwdaejumr0ds5grtsj

Publishing to

relay.nostr.band

relay.damus.io

nos.lol

Author Public Key

npub19jscweanp4p20a3fjes3tsl5crnzwaxdywvjs8m63pqm0rzqswss4gsvpy

Show more details

Last Notes

2026-08-11 12:04:56 UTC

- reply

So the Bob Dylan song refers to the earlier meaning!

2026-07-05 20:00:25 UTC

The Moroccan geographer Mohamed Berriane just told me an amazing story:
About ten years ago, in Corsica, he talked to some Corsicans who had been hiking in the mountains in Morocco. On a high mountain slope, they encountered some bad weather and got lost. So they did what Corsicans do in that situation: they sang a specific song, which is known in Corsica as a call for help. And right away, some local Amazighs heard them and came to help them, because the Amazighs have the same song, which they use in the same way.
Mohamed thought this was fascinating and should be studied, but hasn't found anyone who's interested.
[@lameensouag.bsky.social](https://bsky.brid.gy/r/https://bsky.app/profile/lameensouag.bsky.social )

2026-06-27 12:47:16 UTC

- reply

This is how I'm motivating myself to do more hiking and cycling 😄

2026-06-27 11:56:45 UTC

I know this is totally banal, but I got a little solar panel, I’m charging my phone with it, and it makes me happy ☀️😊
#solar
https://static.piaille.fr/media_attachments/files/116/821/944/274/677/901/original/6beef6ce7311dd21.jpeg

2026-06-20 13:20:01 UTC

Open source work is like a musician doing a gig
'A musician doesn't agonise over "how do I monetise this gig?" The gig is the work. You play because you're a musician.... That's not a business model with a hole in it. It's a craft. "People gotta earn a living" is true, but it sits beside the gig, not inside it. Musicians earn a living through all of it together: the gigs, the teaching, the session work, the reputation, the relationships. You don't solve "how do I monetise this one gig?" You build a life where doing the work makes sense.'
https://www.linkedin.com/pulse/open-source-livelihoods-gig-cristovao-verstraeten-iyhwe/
via @npub1m49…yl0h

2026-05-08 17:03:05 UTC

- reply

And then there’s this: https://nooneshappy.com/article/appearing-productive-in-the-workplace/

2026-05-08 17:00:58 UTC

- reply

I wouldn’t want to have to maintain those applications, or get woken up in the middle of the night because one of them has a bug.

2026-05-08 16:52:28 UTC

- reply

I often think of this: [@ramin_hal9001](https://fe.disroot.org/users/ramin_hal9001 ) https://eigenmagic.net/@daedalus/115533552668217311

2026-05-07 11:53:35 UTC

- reply

If you look at the categories at https://lib.rs, or even the shorter list at https://blessed.rs/crates, you can ask: would it be practical for the developers of the language itself to have to maintain all that functionality? It might be, if they had unlimited resources and expertise, but in practice, all language maintainers have to draw the line somewhere between what goes in the standard library and what has to be maintained by others. Different languages draw the line in different places, for different (and perhaps equally valid) reasons.
[@ramin_hal9001](https://fe.disroot.org/users/ramin_hal9001 ) [@khinsen](https://scholar.social/@khinsen )

2026-05-07 11:39:28 UTC

- reply

I think it depends very much on what you're doing. For a lot of uses of extension languages, you might not need third-party libraries or even a large standard library. But if you need, say, cryptographic algorithms, you probably don't want to implement them yourself, you'd rather use battle-tested (and perhaps externally validated) implementations maintained by security experts.
Plus, the standards and protocols that your code might need to deal with are going to change over time. Yesterday it was Unicode, today it's gRPC and Kafka, tomorrow it will be something else.
[@ramin_hal9001](https://fe.disroot.org/users/ramin_hal9001 ) [@khinsen](https://scholar.social/@khinsen )

2026-05-07 09:18:13 UTC

Some thoughts on what makes a good extension language: https://benjamingeer.name/en/post/extension-languages/
Many thanks to all who responded to my question!
@npub1f7r…6d93 @npub19r2…cahj @npub103y…pnkq [@eeide](https://scholar.social/@eeide ) @npub1fp0…54p5 @npub1j9u…0szu @npub1wva…3etp

2026-03-01 09:48:45 UTC

- reply

The original question was whether LLM coding assistants would make scientists more productive. It sounded like you were arguing that they would, since LLMs are not just hype, as evidenced by their efficiency in producing fake course work, etc. Were you being ironic? [@Ruth_Mottram](https://fediscience.org/@Ruth_Mottram ) @npub1smx…sjc0

2026-03-01 09:40:53 UTC

- reply

It would surprise me if anything I said was new to you. What surprised me was that you described the production of counterfeit goods as productivity. [@Ruth_Mottram](https://fediscience.org/@Ruth_Mottram ) @npub1smx…sjc0

2026-03-01 09:30:11 UTC

- reply

What is the "good" that you want your students to produce? The thing that has real value? Is it essays or learning? Perhaps students are using LLMs to write essays because they mistakenly believe that the essay is an end in itself, rather than a means to an end. As somebody said, sometimes it makes sense to have someone cook your meal for you, but it never makes sense to have someone eat your meal for you. [@Ruth_Mottram](https://fediscience.org/@Ruth_Mottram ) @npub1smx…sjc0

2026-03-01 09:08:09 UTC

- reply

The real productivity that you're asking your students for is their own thinking and learning, right? LLMs aren't producing that, they're producing fake evidence for it, by parroting sentences that were written by people who had actually done the thinking and learning. The problem for higher education is now to figure out how to measure thinking and learning in other ways. [@Ruth_Mottram](https://fediscience.org/@Ruth_Mottram ) @npub1smx…sjc0

2026-03-01 08:45:36 UTC

- reply

Is slop productivity? LLMs are good at producing fake course work, fake scientific papers, fake political debates, etc., which can look plausible and often pass for the real thing if you don't look too closely. [@Ruth_Mottram](https://fediscience.org/@Ruth_Mottram ) @npub1smx…sjc0

2026-03-01 08:22:01 UTC

- reply

But so far nobody has found evidence of productivity gains in a controlled experiment, not even Anthropic? https://www.anthropic.com/research/AI-assistance-coding-skills

2026-02-27 22:34:50 UTC

- reply

If there are vectors and a parallelogram, I suppose there has to be some connection with the determinant in linear algebra?

2026-01-29 19:24:01 UTC

Florida Introduces ‘Sanitized’ Sociology Textbook
"The volume was created after a state review found that existing course materials violated a law prohibiting general education courses from teaching about systemic inequality. So far, at least two universities have adopted it."
https://www.insidehighered.com/news/faculty-issues/academic-freedom/2026/01/29/florida-introduces-sanitized-sociology-textbook

2025-12-29 12:00:20 UTC

- reply

Also if I remember correctly, in some cases, English got both versions, like ‘warranty’ (borrowed earlier) and ‘guarantee’ (borrowed later).

2025-12-23 21:39:58 UTC

RE: https://eupolicy.social/@bert_hubert/115768423791708472
“Industry appears to regard Azure and Microsoft as if they are water companies or electricity providers. As if the cloud comes from highly regulated utilities which will always be there. But this is not the case of course.”
#note1dxs…gl7f

2025-12-07 18:14:20 UTC

@npub1kqd…7640 [@ratatui_rs](https://fosstodon.org/@ratatui_rs ) I'm trying to figure out how to get #Ratatui to display a little TUI that doesn't fill the whole terminal window, but is just a few lines high and appears under the command you typed in the shell, like curl does when it's downloading a file. Is there a simple example of something like this that I could look at? I tried reading the Atuin code but didn't understand what it was doing.

2025-12-03 16:14:17 UTC

Dans les années 1960 et 1970, la rébellion marxiste-léniniste de la région du Dhofar du Sultanat d'Oman a comporté un programme éducatif ambitieux et féministe. Voici mon compte rendu du livre que l'historienne omanaise Mona Salem Said Ja‘bub a consacré à ce sujet : https://benjamingeer.name/fr/post/jabub-dhofar/ #Oman #éducation #féminisme #histoire #racisme #esclavage

2025-09-13 20:17:25 UTC

- reply

For pictures of notable people, I’ve always gone straight to Wikipedia https://en.wikipedia.org/wiki/John_Dewey

2025-08-11 19:33:52 UTC

- reply

I wish I could work with people who, instead of just picking one of the latest bandwagons to jump on, could look back over the past 50 years of computer history and think, of all those languages and technologies, which ones would make sense for the things we want to build now?

2025-08-06 17:12:27 UTC

- reply

Is there a word for when a domain name expires and someone buys it, feeds all the previous owner’s web pages into an LLM, and generates a new site containing a weird mixture of dodgy-looking adverts and LLM paraphrases of the old content (to trick people who search for it)?

2025-07-20 08:49:36 UTC

“In principle, the Max Planck Society gives institutes complete freedom to reward scientists for doing slow, reliable research.… We don’t have to play the game of chasing big-if-true Nature papers.
But most institutes chase Nature papers anyway. Partly this is because PhD students and post-docs cannot make careers at a Max Planck Institute. German employment law makes it impossible for most.”
Which Kind of Science Reform https://elevanth.org/blog/2025/07/09/which-kind-of-science-reform/

2025-07-09 15:09:15 UTC

- reply

So true. I’ve always tried to argue for less code, so there’s less to maintain. Productivity can mean deleting code.

2025-06-29 09:40:36 UTC

- reply

Do you mean having too much sympathy for people from one’s own country and not enough for people from other countries?

2025-04-05 08:09:46 UTC

- reply

The joke is even older than that, there’s a very similar one written down by a 12th-century Persian poet https://archive.org/details/AnvarisDivanAPocketBookforAkbar
https://static.piaille.fr/media_attachments/files/114/284/336/681/669/131/original/ecddedf676a8a16c.png

2025-02-08 22:56:58 UTC

“Anyone who has studied science under the Nazis, as I did for my book Serving the Reich, cannot fail to be dismayed and alarmed at the parallels in the response so far of scientific institutions and academies to the purges and abuses of power following Trump’s executive orders“ https://www.chemistryworld.com/opinion/scientific-institutions-have-a-long-history-of-anticipatory-obedience/4020931.article #science #academia #USPol

2025-02-06 21:18:57 UTC

If anyone is looking for a good way to move a static site off GitHub Pages and have it hosted in the EU, I’ve just moved one such site to https://framagit.org/ (by @npub12k0…e2sz) and https://www.statichost.eu/, and it was very easy and works well.

2024-12-09 22:09:56 UTC

“To dodge sanctions, oil companies blend Russian oil between ships — offshore, out of sight. This can lead to dirty fuel. The kind thought ‘good enough for Africa’.”
in issue 185 of @npub1umk…2hjm
#pollution #UkraineRussiaWar #Nigeria #FossilFuels
https://static.piaille.fr/media_attachments/files/113/625/150/042/623/836/original/b0b4cf6dc8304089.jpeg

2024-12-06 22:21:50 UTC

Government hires software developer to develop spyware and propaganda software, then doesn’t pay the developer #surveillance #Kenya
in issue 184 of @npub1umk…2hjm
https://static.piaille.fr/media_attachments/files/113/608/209/852/557/115/original/5a98001b5dc92649.png

2024-11-23 12:20:14 UTC

Published in an Elsevier journal.
https://doi.org/10.1016/j.ijscr.2024.110139
#science #PeerReview #AcademicPublishing
https://static.piaille.fr/media_attachments/files/113/532/224/204/059/090/original/9cfcfb0f95feba67.png

2024-11-12 19:57:28 UTC

- reply

I still think “As Slow As Possible” is pretty good though https://en.wikipedia.org/wiki/As_Slow_as_Possible

2024-10-16 11:33:20 UTC

- reply

Est-ce que la CAF serait en train de répéter la toeslagenaffaire des Pays-Bas ? https://www.mediapart.fr/journal/international/111122/aux-pays-bas-un-algorithme-discriminatoire-ruine-des-milliers-de-familles

2024-10-02 07:16:33 UTC

- reply

I don't know what the best solution is, but I would consider how it will work for blind people with screen readers.

2024-10-01 11:35:54 UTC

- reply

Donald Knuth’s _Concrete Mathematics_ makes me think that maybe computer scientists are generally better at writing maths textbooks than mathematicians are.