Alix Dunn
LightNews LightNews
alixdunn.com
Alix Dunn
@alixdunn.com
2.5K followers 240 following 230 posts
I bloop about the politics of technology and host the Computer Says Maybe podcast: https://www.themaybe.org/podcast
Posts Replies Media Videos
Reposted by Alix Dunn
abeba.bsky.social
are you displeased with today’s AI safety evaluation landscape and curious about what greater conceptual clarity, methodological soundness, and rigour in AI evaluation could look like? if so, consider coming to Dublin to pursue a PhD with me

apply here: aial.ie/hiring/phd-a...

pls repost
About the PhD: Audits and evaluation of AI systems — and the broader context that AI systems operate in — have become central to conceptualising, quantifying, measuring and understanding the operations, failures, limitations, underlying assumptions, and downstream societal implications of AI systems. Existing AI audit and evaluation efforts are fractured, done in a siloed and ad-hoc manner, and with little deliberation and reflection around conceptual rigour and methodological validity. This PhD is for a candidate that is passionate about exploring what a conceptually cogent, methodologically sound, and well-founded AI evaluation and safety research might look like. This requires grappling with questions such as: What does it mean to represent “ground truth” in proxies, synthetic data, or computational simulation? How do we reliably measure abstract and complex phenomena? What are the epistemological or methodological implications of quantification and measurement approaches we choose to employ? Particularly, what underlying presuppositions, values, or perspectives do they entail? How do we ensure the lived experiences of impacted communities play a critical role in the development and justification of measurement metrics and proxies? Through exploration of these questions, the candidate is expected to engage with core concepts in the philosophy of science, history of science, Black feminist epistemologies, and similar schools of thought to develop an in-depth understanding of existing practices with the aim of applying it to advance shared standards and best practice in AI evaluation. The candidate is expected to integrate empirical (for example, through analysis or evaluation of existing benchmarks) or practical (for example, by executing evaluation of AI systems) components into the overall work.
January 15, 2026 at 11:55 AM
alixdunn.com
Random ask: I regularly work with teams and partners to name things. Episodes, series, projects, you...name it.

Doing it creatively, collaboratively and in a controlled process is often kind of hard.

Do you have a process you use to brainstorm & name things (that you like)?
January 14, 2026 at 6:56 PM
alixdunn.com
Wild juxtaposition for Iran to order the execution of protestors the same week a federal officer of the US executes someone for protesting.
January 14, 2026 at 5:34 PM
Reposted by Alix Dunn
marietjeschaake.bsky.social
The decisions we make now about the true value of sovereignty will determine whether countries retain meaningful autonomy or become clients of foreign imperialism.

A citizen is not a customer and sovereignty is not a service ↘️

giftarticle.ft.com/giftarticle/...
Tech sovereignty should not be a subscription model
Concern about the loss of jurisdiction and agency to US companies is becoming more pressing for Europeans
giftarticle.ft.com
January 14, 2026 at 4:58 PM
alixdunn.com
Excited to be at Sundance in a few weeks moderating a conversation about AI hype, histories, and politics!
January 11, 2026 at 8:14 PM
Reposted by Alix Dunn
abeba.bsky.social
are you disgruntled by the current safety evaluation landscape? curious about what conceptual clarity, methodological soundness and rigour in AI evaluation might look like? if so, consider coming to dublin and doing a phd with me

apply here: aial.ie/hiring/phd-a...
About the PhD Audits and evaluation of AI systems — and the broader context that AI systems operate in — have become central to conceptualising, quantifying, measuring and understanding the operations, failures, limitations, underlying assumptions, and downstream societal implications of AI systems. Existing AI audit and evaluation efforts are fractured, done in a siloed and ad-hoc manner, and with little deliberation and reflection around conceptual rigour and methodological validity. This PhD is for a candidate that is passionate about exploring what a conceptually cogent, methodologically sound, and well-founded AI evaluation and safety research might look like. This requires grappling with questions such as: What does it mean to represent “ground truth” in proxies, synthetic data, or computational simulation? How do we reliably measure abstract and complex phenomena? What are the epistemological or methodological implications of quantification and measurement approaches we choose to employ? Particularly, what underlying presuppositions, values, or perspectives do they entail? How do we ensure the lived experiences of impacted communities play a critical role in the development and justification of measurement metrics and proxies? Through exploration of these questions, the candidate is expected to engage with core concepts in the philosophy of science, history of science, Black feminist epistemologies, and similar schools of thought to develop an in-depth understanding of existing practices with the aim of applying it to advance shared standards and best practice in AI evaluation. The candidate is expected to integrate empirical (for example, through analysis or evaluation of existing benchmarks) or practical (for example, by executing evaluation of AI systems) components into the overall work.
December 17, 2025 at 7:33 PM
Reposted by Alix Dunn
hrdag.org
This year, HRDAG took an unprecedented step: speaking publicly against rising authoritarianism in the United States.

Our executive director joined Computer Says Maybe's @alixdunn.com to talk about why this moment is different and how we are responding. www.themaybe.org/podcast/who-...
The Maybe
www.themaybe.org
December 16, 2025 at 5:58 PM
alixdunn.com
A little advent calendar of #ComputerSaysMaybe episodes coming your way.

The first is an episode beautifully cut by our team to give you an overview of my time at @mozilla.org Festival in Barcelona, including some interviews with amazing folks.

www.themaybe.org/podcast/comp...
The Maybe
www.themaybe.org
December 15, 2025 at 4:46 AM
alixdunn.com
Has anyone proposed a maximum age for social media use?
December 12, 2025 at 12:25 AM
Reposted by Alix Dunn
nkalamb.bsky.social
"AI is the asbestos in the walls of our technological society, stuffed there with wild abandon by a finance sector and tech monopolists run amok. We will be excavating it for a generation or more."

This piece on AI by Cory Doctorow is spectactular.

pluralistic.net/2025/12/05/p...
The Reverse Centaur’s Guide to Criticizing AI (permalink) Last night, I gave a speech for the University of Washington's "Neuroscience, AI and Society" lecture series, through the university's Computational Neuroscience Center. It was called "The Reverse Centaur’s Guide to Criticizing AI," and it's based on the manuscript for my next book, "The Reverse Centaur’s Guide to Life After AI," which will be out from Farrar, Straus and Giroux next June:
December 9, 2025 at 2:58 PM
alixdunn.com
📻 New episode of #ComputerSaysMaybe

Who Knows? Fact-Finding in a Failing State w/ HRDAG and Data & Society

www.themaybe.org/podcast/who-...
The Maybe
www.themaybe.org
December 5, 2025 at 6:33 PM
alixdunn.com
📻 New episode of #ComputerSaysMaybe

Who Knows? Independent Researchers in a Platform Era w/ @geurkink.bsky.social

www.themaybe.org/podcast/who-...
The Maybe
Imagine doing tech research… but from outside the tech industry? What an idea…
www.themaybe.org
December 2, 2025 at 11:25 PM
Reposted by Alix Dunn
gilduran.com
Trump's pardon of Honduras' drug trafficker president is raising awareness of Prospera, the weird tech-ruled city he helped to create.

Prospera is #NetworkState, a scary cult of tech billionaires who plan to replace nation-states with privatized dystopias.

More:

www.youtube.com/watch?v=6iHl...
1. Tech Billionaires vs. Democracy: Elon Musk and the Rise of "Network States"
YouTube video by The Nerd Reich with Gil Duran
www.youtube.com
December 2, 2025 at 10:48 PM
Reposted by Alix Dunn
eryk.bsky.social
What was ChatGPT? Now nearly three years old, we can look at OpenAI's LLM as a product of its time, optimized ever since to its earliest uses. While this period of deep disorientation and social isolation has been obscured from public memory, it remains embedded within the interface.
What Was ChatGPT?
A Chatbot Optimized for Social Distance Three years after the launch of ChatGPT, we can finally speak in hindsight about what it was and how it came to be. Its meteoric rise shocked the world, gather...
mail.cyberneticforests.com
November 23, 2025 at 12:37 PM
alixdunn.com
The chaos monkey on my shoulder really wants him to veto it.
November 19, 2025 at 1:03 AM
Reposted by Alix Dunn
karenmonroy.bsky.social
Dirty Money
#Econsky
November 17, 2025 at 12:56 PM
Reposted by Alix Dunn
alicetiara.bsky.social
New in @theverge.com - half of US scams originate on a Meta property. The company must do more. @lanalanalana.bsky.social and I weigh in www.theverge.com/tech/820906/...
Meta must rein in scammers — or face consequences
Scams are ruinous to users — but, reportedly, big business to Meta.
www.theverge.com
November 14, 2025 at 7:52 PM
Reposted by Alix Dunn
smw.bsky.social
This week Open AI walked back a call for the govt to backstop financing for its trillion dollar investments in data centers. This was only the tip of the iceberg; a slow bailout for AI firms is already underway. Read more from @ambakak.bsky.social and I in @wsj.com: www.wsj.com/opinion/you-...
Opinion | You May Already Be Bailing Out the AI Business
Washington is treating the industry as if it’s too big to fail, even as the market sends lukewarm signals.
www.wsj.com
November 12, 2025 at 10:56 PM
Reposted by Alix Dunn
heidykhlaaf.bsky.social
New Report: Fission for Algorithms. We draw on our nuclear expertise to dissect the risky fast-tracking initiatives hastening nuclear development in service of AI. This includes proposals to use Gen AI for nuclear licensing, whilst lowering well-established nuclear thresholds.
Fission for Algorithms: The Undermining of Nuclear Regulation in Service of AI - AI Now Institute
A report examining nuclear “fast-tracking” initiatives on their feasibility and their impact on nuclear safety, security, and safeguards.
ainowinstitute.org
November 12, 2025 at 11:05 AM
Reposted by Alix Dunn
nicdawes.bsky.social
Many news orgs are "all-in" on AI, fearful of repeating the errors of the early digital era, but what if the real risk lies in handing chunks of our business, our civic function, and our values to a handful of hyperscalers? @alixdunn.com prodded me to think that through with her for @themaybe.org
The Maybe
www.themaybe.org
November 7, 2025 at 9:34 PM
alixdunn.com
This is so good!
rebeccawilliams.info Rebecca Williams @rebeccawilliams.info · Nov 5
For @techpolicypress.bsky.social I laid out a tech agenda for Mayor-elect zohrankmamdani.bsky.social that resists surveillance and extraction while advancing his goals for affordability, dignity, and justice.
Zohran Kwame Mamdani (@zohrankmamdani.bsky.social)
Mayor-Elect of New York City
zohrankmamdani.bsky.social
November 6, 2025 at 2:30 AM
Reposted by Alix Dunn
themaybe.org
Computer Says Maybe is taking the stage at @mozilla.org Festival! This weekend, @alixdunn.com will be leading discussions on reclaiming scale in tech, tackling the role of AI in war, exposing the global footprint of data centers, and more.
November 5, 2025 at 7:02 PM
alixdunn.com
My 'o' key is making an oo all the time and now all my typoos look Dutch.
October 29, 2025 at 6:59 PM
Reposted by Alix Dunn
themaybe.org
Can a chatbot therapist really replace a human one? The simple answer is, no of course not. But as always, it’s a little more complicated than that.

Access to (real, human-to-human) therapy is limited. You need money, and probably have to jump through a lot of hoops.
October 28, 2025 at 5:21 PM
alixdunn.com
This week on the show, I talk to Stevie Chancellor, a co-author on a banger of a paper introducing new evaluation frameworks and some shocking findings on the use of chat bots for mental health support.
October 25, 2025 at 2:15 AM