Alex Hanna
LightNews LightNews
banner
alexhanna.bsky.social
Alex Hanna
@alexhanna.bsky.social
21K followers 1.2K following 4.3K posts
Director of Research, @dairinstitute.bsky.social
Roller derby athlete
https://alex-hanna.com
Book: thecon.ai
Pod+newsletter: https://dair-institute.org/maiht3k
🇪🇬ⲙⲉⲧⲣⲉⲙⲛ̀ⲭⲏⲙⲓ
🏳️‍⚧️ she/هي
📸 @willtoft.bsky.social
Rep📘 @ianbonaparte.bsky.social
Posts Replies Media Videos
alexhanna.bsky.social
New workers' inquiry from the @dataworkersinquiry.bsky.social project: the story of Michael Geoffrey Asia, who tells his story of impersonating romantic and sexual partner for users who think they are speaking to a chatbot.

data-workers.org/michael/
A picture of a dark-skinned person against a dark red background, holding a chat bubble which has a ragged heart in the middle. Written in the heart are the words "The Emotional Labor behind AI Intimacy. Michael Geoffery Asia." The chat bubble is coming out of the Data Workers' Inquiry logo, which is a megaphone on a laptop trackpad. Chat moderators are hired by companies such as Texting Factory, Cloudworkers, and New Media Services to impersonate fabricated identities, often romantic or sexual, and chat with paying users who believe they’re forming genuine connections. The goal is to keep users engaged, meet message quotas, and never reveal who you really are. It’s work that demands constant emotional performance: pretending to be someone you’re not, feeling what you don’t feel, and expressing affection you don’t mean. Over time, I began to suspect that I wasn’t just chatting with lonely users. I was also helping to train AI companions, systems designed to simulate love, empathy, and intimacy. Many of us believed we were simultaneously impersonating chatbots and teaching them how to replace us. Every joke, confession, and “I love you” became data to refine the next generation of conversational AI.
December 9, 2025 at 4:16 PM
alexhanna.bsky.social
Milton Friedman's grandson poo poos the suggestion that network-state projects in Africa would be colonizing, although he says that people from their prospective sites would need to be resettled.

This is all very on the nose.
Please use the sharing tools found via the share button at the top or side of articles. Copying articles to share with others is a breach of FT.com T&Cs and Copyright Policy. Email licensing@ft.com to buy additional rights. Subscribers may share up to 10 or 20 articles per month using the gift article service. More information can be found at https://www.ft.com/tour. https://www.ft.com/content/b127ee7a-5ac4-4730-a395-c9f9619615c7 Patri Friedman initially pushes back on the notion of these projects as neocolonialist, declaring “most of our projects are greenfield”. He pauses. “Although, in Africa, we are looking at land parcels large enough that there will be people living there, in which case we will offer relocation bonuses to pay for anybody who wants to move out of the zone.”
December 8, 2025 at 1:42 AM
alexhanna.bsky.social
It is the most Google shit to literally paint an ad across the conference center where NeurIPS is being held.
An ad painted on a building, across the windows and all. Black background, white tech that says a bunch of corporate speak and says Work at Google at the bottom and gives a link.
December 6, 2025 at 4:33 PM
alexhanna.bsky.social
I dunno, maybe poking at people on the platform that don't glorify AI, Grok, crypto scammers, and ethnostates is your issue.
A BlueSky post from Armin Ronacher (@mitsuhiko.at) that reads "I'm not sure what this platform can do to make it more enjoyable, but I'm definitely not enjoying whatever pile-on is taking place right now in my notifications and I'm also not really sure how I could make it stop."
November 27, 2025 at 2:42 AM
alexhanna.bsky.social
us: Hi Google Scholar, could you stop deadnaming trans people?

GS: No, but how about AI?

us: No thanks. I just would like articles to not show my deadname.

GS: We can't do that because of reasons. But do you like synthetic text?

me: I don't want this. Let me change my deadname.

GS: but... AI!
The Google Scholar logo and search bar. Under, it says "New! Scholar Labs: An AI Powered Scholar Search" And has a gradient box with text that says "A new way to search. Try Scholar Labs"
November 19, 2025 at 12:11 AM
alexhanna.bsky.social
The Information's AI Agenda newsletter reporting on ad placement in "AI" coding assistants. This is a screenshot of SourceGraph.

"So far, around 25 companies have placed ads, including AI cloud-server provider Baseten, compliance startup Vanta, and database startup Turbopuffer..."
A coding assistant interface. There is an ad on top that reads "Add Google, Microsoft, and GitHub login to your app with WorkOS AuthKit. Add social login"
November 17, 2025 at 3:11 PM
alexhanna.bsky.social
An unholy arrangement is being forged between struggling crypto miners and AI companies who are desperate for data centers.

Crypto firms that were doing poorly are now reaping memecoin rewards because of massive deals with Anthropic and Google.

www.theinformation.com/articles/goo...
Shares in TeraWulf and competing crypto miners have soared in recent months after they struck deals with tech companies desperate for data centers and energy resources to power their AI bets. Anthropic, maker of the Claude chatbot, added fuel to the trade this week by partnering with cloud startup Fluidstack, which has entered a series of byzantine arrangements with TeraWulf and Google to secure data center space. Fluidstack and Google have signed a similarly structured deal with Cipher Mining Technologies for a data center it’s building in Texas, stoking interest in that company’s stock. Crypto miners have the hard assets like land and power that have become bottlenecks to the growth in AI.
November 14, 2025 at 3:12 PM
alexhanna.bsky.social
Dont blame me, I voted for
"KODOS for SPRINGFIELD" in the Mamdani font
November 8, 2025 at 3:39 PM
alexhanna.bsky.social
Holy shit. Noam Shazeer, one of the original authors on the "Attention is All You Need" paper and Character.AI founder, came out as major transphobe. Like Trumpian levels of "this is child mutilation" of transphobia.

(via The Information)
This spring, Google AI leader Noam Shazeer posted a comment in one of the company’s internal discussion forums, in response to a post about how Google employees could support their transgender and nonbinary colleagues on International Transgender Day of Visibility. “I do not believe that humans have an attribute called gender,” Shazeer wrote, according to a copy of the post viewed by The Information. “I do not believe that G-d puts people in the wrong bodies. I do not believe that it is okay to sterilize children. You have the right to your beliefs. I do not share them.”
November 7, 2025 at 2:17 PM
alexhanna.bsky.social
Opening up #MozFest, @ruha9.bsky.social asks us "who owns the future?" but also problematizing the notion of ownership. Need to shift the question, expanding the vision of how we imagine new futures.
A projected screen. A picture of Nichelle Nichols, the late actress who played Uhura, sitting in the Captain's seat on the Starship Enterprise. Superimposed on the picture are the words in multiple colors: "who owns the future?"
November 7, 2025 at 11:30 AM
alexhanna.bsky.social
Gramsci mentioned, followed by Lil Wayne in opening from Nabiha Syed #MozFest
"The crisis consists precisely in the fact that the old is dying and the new cannot be born." - Antonio Gramsci What's a goon to a goblin? - Lil' Wayne
November 7, 2025 at 10:55 AM
alexhanna.bsky.social
A linguist at UMN shared this with me today, which I thought was a really helpful discussion of genAI in the classroom as a conversation around consent.

refusinggenai.wordpress.com/2025/08/29/a...
On the other hand, I felt disappointed because I missed out on an opportunity to learn from my professor. I lost a learning moment, the possibility to add a new skill to my repertoire, and the opportunity to feel proud of my work. Most importantly, I watched what I valued in my education slip away. What could have been an insightful and reciprocal conversation turned one-sided and transactional. What could have been a moment of open-ended intellectual inquiry turned into a multiple-choice question. I have had other classes with this professor and appreciated the rich conversations we had in the past. I was hoping for something similar again. It reminded me of Paulo Freire’s theories in Pedagogy of the Oppressed, a work that has been engrained in my mind from my writing tutor work. I realized that this interaction followed Freire’s banking concept in some ways: I was deposited an answer to my question without the time to work through it with my professor and truly learn the process of answering this question for myself.
November 4, 2025 at 1:08 AM
alexhanna.bsky.social
Here I made this
A scribble on a notepad from an app. It says "DEATHLY TAMARA" and "die" and "murder" in the corners
November 3, 2025 at 11:41 PM
alexhanna.bsky.social
Just got this email from Instagram.

The drive to adtech was going to be the only way they were actually monetizing any of this.
On December 16, 2025, we’re making changes to our Privacy Policy. Here are some details. Personalizing your experiences We’ll start using your interactions with AIs to personalize your experiences and ads.
November 3, 2025 at 1:45 PM
alexhanna.bsky.social
I know no brand is immune to this bullshit, but it pains me to say that my beloved Casio has developed an "AI-powered" tribble that is apparently a big hit in Japan.

www.theinformation.com/articles/wan...
Moflins rely on what Casio calls “emotional AI” to learn and respond to their environments, developing different personalities based on their owners’ interactions. The companion app, MofLife, allows users to track a Moflin’s mood to see how affectionate and energetic it feels.
November 2, 2025 at 4:23 PM
alexhanna.bsky.social
It seems like the only way tech companies are able to compel AI usage is by coercion in performance review processes?

(via The Information "AI Agenda" newsletter)
3. AI adoption is easier said than done. Even as executives pressure workers to use AI, getting people to do that throughout an organization is easier said than done. Rukmini Reddy, an engineering executive at incident management software maker PagerDuty, does so by making AI usage a part of her employees’ annual performance reviews. This strategy seems to be working, as she said that 98% of her engineers use coding tools like Anthropic’s Claude Code or Microsoft’s GitHub Copilot on a day-to-day basis now.
October 30, 2025 at 2:30 PM
alexhanna.bsky.social
TOMORROW -- join @timnitgebru.bsky.social and I at this event sponsored by Respond Crisis Translation.

From ICE Abductions to Genocide in Gaza, Language Workers Resist: A Language Justice Forum
🗓️ Oct 30, 2025 - 5:30 PM
📍Mission San Francisco
RSVP for location = bit.ly/rct-rsvp
A black and yellow flyer for an event. The title reads: From ICE abductions to Genocide in Gaza, Language Workers Resist. A Language Justice Forum with Respond Crisis Translation. Community food + drink reception, multilingual cultural performances, and interactive sessions: Language rights Abuses in ICE Custody: Survivors Speak Out On the ground in Gaza: Palestinian Translators Surviving and Resisting Lost in Machine Translation: AI-Fueled Language Violence against migrants Featured Speakers: Dr. Timnit Gebru Gheed Anna S
October 29, 2025 at 11:20 PM
alexhanna.bsky.social
"If I had a nickel for every time Altman referenced heroin on the livestream—in reference to the idea that OpenAI wants adult users of ChatGPT to have more freedom with the chatbot as long as they’re not doing something as dangerous as selling heroin—I’d have at least two nickels."
AI Agenda newsletter. The first two grafs read: Hello! OpenAI flooded the zone yesterday with a long-awaited announcement regarding its revised agreement with Microsoft and corporate restructuring, and an hour-long livestream in which CEO Sam Altman and Chief Scientist Jakub Pachocki answered questions from the public. (For more on the restructuring, check out this story from my colleagues.) On the livestream, the two men covered a range of topics, from OpenAI’s decision to allow more NSFW content like erotica on ChatGPT to the deprecation of popular old models such as GPT-4o to the threat of job loss from AI. (If I had a nickel for every time Altman referenced heroin on the livestream—in reference to the idea that OpenAI wants adult users of ChatGPT to have more freedom with the chatbot as long as they’re not doing something as dangerous as selling heroin—I’d have at least two nickels.)
October 29, 2025 at 8:29 PM
alexhanna.bsky.social
You don't say, LinkedIn news.

You mean the plagiarism machine that spits out wrong things isn't a reliable replacement for actual human labor?
A headline from LinkedIn news which reads "AI may not have been behind tens of thousands of job cuts" by Cate Chapman, Editor at LinkedIn News
October 29, 2025 at 8:14 PM
alexhanna.bsky.social
Never thought I'd be someone to be able to get a writing residency, but I am doing one early next year at the Virginia Center for the Creative Arts. Lemme know if you're around there??

www.vcca.com
The VCCA logo, with the text "Virginia Center for the Creative Arts", and below "Writer Going to VCCA"
October 23, 2025 at 6:00 PM
alexhanna.bsky.social
DAIR’s (@dairinstitute.bsky.social) latest report offers a chilling glimpse into how Amazon is using technology to ramp up surveillance of delivery drivers, violating existing workers' rights and creating unsafe working conditions.

www.dair-institute.org/projects/dri...
The text reads: “Amazon is Building a Surveillance Empire on the Backs of Delivery Drivers. A new report from the Distributed AI Research Institute (DAIR) exposes the dark underbelly of technology-driven workplace surveillance.” Above the text is an old school television with an antenna featuring a surveillance snapshot from a camera inside of an Amazon delivery van. The Amazon logo is upside down and is a frown. Around the television are disembodied eyes and ears.
October 22, 2025 at 2:41 PM
alexhanna.bsky.social
Appreciate that Maggie points out that these chucklefucks don't care about all the other externalities of AI slop.
It’s also worth pointing out that the letter doesn’t clarify the reality that AI doesn’t have to reach superintelligence to cause chaos: as it stands, generative AI tools like chatbots and image and video-creation tools — primitive technologies when compared against imagined future superintelligent AI systems — are upending education, transforming the web into an increasingly misinformation-prone and unreal environment, expediting the creation and dissemination of nonconsensual and illegal pornography, and sending users of all ages spinning into mental health crises and reality breaks that have resulted in outcomes like divorce, homelessness, jail, involuntary commitments, self-harm, and death.
October 22, 2025 at 1:03 PM
alexhanna.bsky.social
People of means can afford actual information about the administrative state. When they mess up, they get a light pat on the wrist.

For everyone else, they get cheap facsimiles and dire consequences.
As beta testers for the city’s first public-facing use of AI, a group of multilingual, stressed-out, marginally housed people living below the poverty line and — in some cases — struggling with mental health and substance abuse issues, is an unusually vulnerable group of users. That particularly goes for the kind of tech that, if it fails, could mean the difference between relative stability and chaos.
October 21, 2025 at 8:52 PM
alexhanna.bsky.social
Mystery AI Hype Theater 3000 Episode 64
www.buzzsprout.com/2126417/epis...

@reveleth.com joins @emilymbender.bsky.social and I to talk about how we don’t need tech bros to read sci-fi better, they need to understand it better.
October 15, 2025 at 2:40 PM
alexhanna.bsky.social
The Information confirming what most of us already seemed to get a handle on -- "AI" has been profitable for only the chip manufacturers, and the circular network of cloud providers.

www.theinformation.com/articles/ai-...
Meanwhile, we’ve seen precious little evidence that AI is a profitable business for those using the servers to sell AI apps or those developing AI models. The cost of AI models weighs on all AI-powered applications, although those costs have come down over time. Coding assistants, for instance, aren’t exactly high margin. We’d bet that’s true of other AI apps, whether they’re run by big companies or startups, although we don’t really know. Few of the big public companies have said much about the revenue they’re getting from AI apps, let alone the profits.
October 13, 2025 at 9:07 PM