Profile avatar
ulyssespascal.bsky.social
Geography Postdoc in Finance and Big Tech at VUB, Brussels. PhD in Information Studies from UCLA. Writing “Circuits of Finance: NASDAQ and the making of the global market.” Platforms | Information Infrastructure | AI | Sustainability | Uneven Geographies
86 posts 305 followers 1,083 following
Regular Contributor
Active Commenter

A free chat interface is not open ai’s product! You are criticizing the hype. Not the product. Am i the only one who finds these takes naive and dangerous?? We are being categorized and surveilled by ai. It’s built into everything on the backend. The chat interface is just the 1% you see

Is it unethical to get your Al to work better by telling it that it needs to do o good job or else its mom will die of cancer?

If you submitted an article to a journal and they didn’t get back to you for 2+ years what do you do?

In Rome they offered bread and circuses, in the US they offer chatbot arenas

I work with AI in social science and it takes months of iterative improvement to engineer and evaluate prompts to ensure they don’t cause numerous bugs and unexpected behavior. Feeding government data into AI is dangerous www.washingtonpost.com/nation/2025/...

How did they ship this?

Quantum gravity is my Roman Empire

Under DOGE, Musk may be violating the Federal Information Security Modernization Act (FISMA), NIST Cybersecurity Framework, Federal Risk and Authorization Management Program (FedRAMP), Gramm-Leach-Bliley Act (GLBA), Sarbanes-Oxley Act (SOX), OCC Guidelines on Third-Party Relationships

🚨BREAKING. From a program officer at the National Science Foundation, a list of keywords that can cause a grant to be pulled. I will be sharing screenshots of these keywords along with a decision tree. Please share widely. This is a crisis for academic freedom & science.

Did elon buy the pee tape from putin? I need a better explanation of what is actually going on

Is anyone doing a 4s panel on political economy of AI?

If you don’t code or haven’t done math in 10+ years, AI feels like just a bad writer. It’s a very different visceral experience to watch it spit out code that works than text that falls flat.

Tell your students if they use AI to do their HW you will use AI to write their recommendation letters.

I’m am making a google sheet add-on that integrates with LLM providers (like open ai and Deepseek). It is currently bare bones but works great! Now I’m trying to turn it into a Google workspace add-on and it is a NIGHTMARE! Help! How do mortals do this???

We are so monopoly pilled that we can’t conceive of competition between two companies as anything other than “loss”

Sometimes I wish surveillance capitalism was just a little bit better.

You can circumvent Deepseek censorship with pig latin! Unfortunately, Deepseek is not very good at pig Latin so I had to use ChatGPT to translate it back into English

Proof of human-ness is about it get more relevant: www.youtube.com/live/CSE77wA...

$500 Billion to build AI without safety regulations or oversight in Abilene, Texas where it recorded its hottest day ever last year at 113°F (45°C)

“Trump will save TikTok” is crazy. It was Trump’s Executive Order in 2020 that began the ban on TikTok

An under reported reason for why people turn to AI output instead of finding information online is there are no ads (yet)

OpenAI preemptively asking for exemptions from state regulations on AI

I have work on AI and platform capitalism that Id love to share at a conference. I also have developed several hands on AI digital methods and LLM forensics workshops (please invite me to your classrooms). Manifesting more scholarly community 2025

This one simple trick will maximize unregretted user-seconds on twitter/x… delete your account!

We need bot proof social media platforms. Im so sick of bots.