The Machine Ghost in the Human Shell by Kawita Vatanajyankur and Pat Pataranutaporn is set to premier at the 11th Asia Pacific Triennial of Contemporary Art, Queensland Art Gallery & Gallery of Modern Art, Australia Massachusetts Institute of Technology (MIT) MIT Media Lab QAGOMA qagoma.qld.gov.au/exhibition/the… media.mit.edu/projects/the-m…

I read this study before bed last night: “Conversational AI Powered by Large Language Models Amplifies False Memories in Witness Interviews.” It scared me more than I thought it would. The study explores how chatbots powered by LLMs can influence the formation of false memories

Super-intelligence or Superstition? Exploring Psychological Factors Underlying Unwarranted Belief in AI Predictions. arxiv.org/abs/2408.06602

Thrilled and honored for the premiere of Cyber Subin: Evolving Cultural Heritage through Collaborative Choreography with Generative Virtual Characters, co-created with Pichet Klunchun, in Bali tonight media.mit.edu/projects/cyber… cybersubin.media.mit.edu

For MIT Technology Review, Melissa Heikkilä digs into the ways people are actually using AI tools—including developing relationships with them. She talks to Media Lab researchers Pat Pataranutaporn + Robert Mahari to find out what that could mean for users' emotional wellbeing. technologyreview.com/2024/08/12/109…

Trust in news organizations has never been lower. People trust individual personalities, not outlets. We will see continued fracturing of shared reality, as Americans embrace more contradictory reality tunnels, and a growing impulse towards isolation inspired by info-exhaustion.

This problem is inherent to the design of current AI (as @garymarcus repeatedly points out): Unlike human minds, it knows no facts; has no explicit conception of people, things, places, causes, and times; just statistical patterns of patterns of patterns. | Generative-AI

This morning I rode an dinosaur to a temple; it was a great experience. Recommended for everyone visiting Thailand.

"Technology has come to be a synonym for progress, but technology that robs us of the time, wisdom, and focus needed for deep reflection is a step backward for humanity." Outstanding article and important conversation by Robert Mahari and Pat Pataranutaporn technologyreview.com/2024/08/05/109…

My article with Robert Mahari is in the top 3 most popular article on MIT Technology Review today! Wow!! 🤯 I hope it is not too addictive Massachusetts Institute of Technology (MIT) MIT Media Lab : technologyreview.com/2024/08/05/109…

Give me a number of H100s big enough and a datacenter in which to place them, and I shall move the LLM leaderboard.

With more and more people accessing the web via generative AI, I investigated these products' capacity to influence people's knowledge, opinions, even memories. The issue may be more complex and pernicious than a fake image or article: theatlantic.com/technology/arc…

In our most recent study, we found that AI could significantly increase false memory formation, inducing over 3 times more immediate false memories than the control 36.4% of users were misled through the interaction. media.mit.edu/projects/ai-fa… arxiv.org/abs/2408.04681

Thank you Matteo Wong and the The Atlantic for featuring our research in the brilliant article on AI and false memories theatlantic.com/technology/arc…

In MIT Technology Review, Robert Mahari and Pat Pataranutaporn consider the potential harms of human relationships with AI agents. “AI companionship is no longer theoretical,” they write. “We are already starting to invite AIs into our lives as friends, lovers, mentors, therapists, and teachers,”

Fascinating study that examines the persuasive power of deceptive AI compared to honest AI in spreading misinformation by Valdemar Danry Pat Pataranutaporn Matt Groh Zivvy Ξpstein Pattie Maes arxiv.org/pdf/2408.00024

I am excited to deliver a keynote lecture on “Cyborg Culture: AI for the Examination, Experimentation, and Evolution of Cultural Heritage” at the Heritage Future Forum 2024, hosted by the Korean National Research Institute of Cultural Heritage. MIT Media Lab

When you see your self wearing a dinosaur suit around korea 🦖😂

Today, I’m excited to share with you all the fruit of our effort at OpenAI to create AI models capable of truly general reasoning: OpenAI's new o1 model series! (aka 🍓) Let me explain 🧵 1/

92 seconds of “thought”. What might 92 days bring?