Theoreti.ca – Research notes taken on subjects around multimedia, electronic texts, and computer games.

Anthropic co-founder remarks on “Magnifica humanitas”

Anthropic co-founder Chris Olah was one of the respondents on Pope Leo XIV’s encyclical Magnifica humanitas. (You can see my blog post here.) His remarks have been posted on the Anthropic site. Unlike some accelerationists who don’t want to listen to anyone outside the magic circle, he is humble and invites dialogue. Some of the points he makes:

He starts by recognizing how those who work in the AI industry operate “inside a set of incentives and constraints that can sometimes conflict with doing the right thing.”
Which is why he welcomes “Magnifica humanitas” and calls for dialogue.
He hopes that “if this technology is coming, it must go well—for our common home, and for the children to come.” Note that he doesn’t assume it is coming.
He describes it in almost human terms. “They are grown, on a structure roughly modeled after the brain, on an enormous inheritance of human thought and speech.” He says they are not the cold beings we expected but developing them is, “a little like bringing a fictional character to life.” Note that he doesn’t say it is like bringing a person to life, but making close as fiction. (Remember that “fiction” comes etymologically from “fashioning.”)

He names three questions where the Church’s discernment is needed. For that matter, he is clear that these discussion are not just for the engineers or Church, but for all.

The first is our duty to the global poor.
The second is the need for moral imagination and ambition regarding human flourishing.
The third is the need for discernment on the nature of AI models.

I would like to ask how a dialogue that goes beyond those in and with power could take place? Many just wait and hope they don’t lose their jobs or end up managed by a machine.

Pope Leo’s ‘Magnifica humanitas’: AI must serve humanity not concentrate power

Pope Leo the XIV just released an encyclical on artificial intelligence, Magnifica humanitas: On Safeguarding the Human Person in the Time of Artificial Intelligence. The encyclical was released on the 135th anniversary of the encyclical Rerum Novarum: On Capital and Human Labor by his namesake pope, Leo XIII.

The two are clearly meant to be connected. Magnifica humanitas, or “magnificent humanity” doesn’t paint AI as either good or bad, or for that matter, neutral. It is about humans making choices that safeguard (or not) our dignity (and magnificence.) I rather like the emphasis on our magnificence and the suggestion and AI could be one of our great deeds.

The encyclical identifies 5 principles that are not the usual principles for AI as they are principles of Social Doctrine with application to AI:

Common good
Universal destination of good (that not all goods be concentrated in the hands of the few)
Subsidiarity (that have shared responsibility, not paternalistic welfare)
Solidarity (that we care for others and future generations)
Social justice

The encyclical is long and deals with a number of issues including the need for dialogue. You can read a summary on the Vatican news web site here. The New York Times has an article on Main Takeaways From Pope Leo’s Encyclical on A.I. and one on how At the Epicenter of A.I., Pope Leo’s Warnings Are Dismissed. This last article captures the attitude of some that “no one outside of Silicon Valley understands AI and therefore we don’t need to listen to them.” No need for dialogue with others when you think you are magnificent.

Edmond de Belamy by Obvious

The Portrait of Edward de Belamy is a Generative Adversarial Network (GAN) ink print created by the art group Obvious.

The name “Belamy” is a joke on the name of the developer of GAN, Ian Goodfellow.

The print was bought at auction for $432,500 (USD), a first for AI generated art.

I learned about this from a talk on “Alternative Histories of genAI” by Lauren Tilton. She talked about the institutions involve in authorizing art. Is it art because it was bought for a lot of money?

This talk was part of a conference in Nice and Cannes, Colloque “Créativités artificielles: Approches critiques de l’IA” @ Nice that I also presented at. My talk was on “Thinking-Through Trust in AI”.

The Coming Tsunami of Transnational Repression

Professor Ronald Deibert, Director of the Citizen Lab at U of Toronto, appeared before the House of Commons’ Subcommittee on International Human Rights of the Standing Committee on Foreign Affairs and International Development (SDIR) on March 23rd, 2026, to testify on transnational repression. His brief included a discussion of artificial intelligence and its potential use for repression. He also had a clear recommendation that the government should regulate AI. Here is the recommendation from the text of his testimony:

Regulate AI. The government must squarely and soberly address the huge potential for widespread harm associated with LLMs and AI systems, as well as social media
platforms which are connected to them. Although there are many potential economic and other benefits associated with these systems, the current political and economic context all but assures there will also be major harms emanating from their use and abuse. The government should cease any cooperation with governments on AI that are known perpetrators of TNR and DTR. It should engage in meaningful public consultation with affected communities on how these systems have begun to negatively affect peoples’lives, as called for by the People’s Consultation on AI. And it should find ways to regulateAI uses, particularly among public agencies, to mitigate harms and ensure equitable outcomes. Part of this regulation must include independent due diligence audits of all tech platforms in a transparent and accountable manner consistent with Charter of Human Rights protections on freedom of speech and access to information.

I read about this in a story in the Globe and Mail.

Enshittification

The Norwegian Consumer Council has released a punchy video about enshittification. If you go to the web site at the end you get to a page about Breaking Free. This has a link to a report on Breaking Free: Pathways to a Fair Technological Future (PDF) which argues that generative AI is the next frontier to enshittification. They point to how AI can generate large quantities of slop now sloshing around the internet.

The neologism enshittification was coined by Cory Doctorow. His web site has links to his book on it and videos of him talking about it.

The good news is, as Doctorow puts it in his book on the subject, “A new, good internet is possible. More than that, it is essential.” The final section 5 of the Norwegian report offers advice on how we can break free.

For me, it is essential that we resist the network effect, and just drop services that become unacceptably shitty. When they change the privacy settings, just drop it. It may be painful and it may feel as if your social life won’t recover, but that is what they want you to believe.

AI Isn’t Coming for Everyone’s Job

The Atlantic has a thoughtful article titled, AI Isn’t Coming for Everyone’s Job. The article points out that player pianos automated the playing of pianos in the early 1900s and could even play things humans didn’t have enough fingers for, but that didn’t put piano players out of work.

How could humans possibly compete? Yet today you are more likely to encounter a piano player than a player piano, despite the job being successfully automated a very long time ago. The automatons have been relegated to museums and the rare curiosity. Pianists can be found any night of the week in hotel lobbies, Italian restaurants, and concert halls.

The article goes on to talk about how live music is still appreciated even though many musicians can’t play as well as what you can get on recorded (or automated) music. People like to see, hear, and interact with other people.

It also mentions how people fought back. Above you see an image from an ad in 1930. Earlier John Philip Sousa coined the phrase “canned music” in 1906 to mock the automated sound. (At the time the cylindrical records came in can shaped containers.) According to the Wikipedia article, he testified to Congress,

These talking machines are going to ruin the artistic development of music in this country. When I was a boy… in front of every house in the summer evenings, you would find young people together singing the songs of the day or old songs. Today you hear these infernal machines going night and day. We will not have a vocal cord left. The vocal cord will be eliminated by a process of evolution, as was the tail of man when he came from the ape.

Sounds like some of the concerns we have about AI today, but again, I suspect live music will survive.

The problem is more likely to be arts where there isn’t a live person performing or interacting with you. Does it really matter if illustrations in magazines are made by humans, AIs or hybrids as long as they catch the eye and illustrate the topic? Perhaps the visual arts will shift to live performance art or those online performances like those for YouTube by Bob Ross.

Anthropic is standing up to the Department of War (DoW) on what we might call ethics issues. This story has some interesting angles.

Originally Anthropic had a contract with the DoW to provide AI services across the government. They had two red lines:

Their AIs couldn’t be used for fully autonomous lethal weapons.
Their AIs couldn’t be used for mass surveillance of US citizens.

The government pushed back and eventually cancelled the contract. Then they designated Anthropic a Supply Chain Risk which could make it hard for any government agency to contract with them. So … they are suing now. Here are some interesting links on the story:

Dario Amodei, founder and CEO of Anthropic, has this to say about Anthropic’s position.
An Air Force general had this to say about the spat.

Both are short and worth reading.

Indigitization

At the Spokenweb conference last summer I heard Gerry Lawson talk about the Indigitization.ca project for which he is the Technical Lead. This is a neat project at UBC that has kits, guides, and small grants for indigenous communities to “facilitate capacity building in Indigenous information management.” Communities can get a kit that lets them video their elders to create a digital archive of their cultural information. They do it them selves with help from the Indigitization projet.

They have a great Toolkit with guides on all sorts of subjects. The image above is from the Digital Storage guide. These guides are useful to anyone doing digitization projects!

Calculating Empires

The creative team of Kate Crawford and Vladan Joler who brought us the Anatomy of an AI System have created a much more ambitious long wall sized infographic called Calculating Empires: https://calculatingempires.net.

I saw this at the Jeau de Paume exhibit on The World Through AI. I feel it is the sort of thing I would like a large poster of so I could carefully read it, but … no luck … no posters.

Anyway, it is a fascinating map of communications technology.