📢 The code for CTAG is now available! You can easily generate sound from prompts by using a modular synthesizer (SynthAX ⚡️) github.com/PapayaResearch…

At #ICML2024, Media Lab researchers Manuel Cherep, Nikhil Singh, and Jessica Shand will present “Creative Text-to-Audio Generation via Synthesizer Programming,” a text-to-audio generation method using a virtual modular synth. Explore the code! ctag.media.mit.edu

Congratulations to Media Lab student Manuel Cherep on being selected for the 2024–2026 "la Caixa" Fellowship! media.mit.edu/articles/manue…

✨New Preprint ✨ How are shifting norms on the web impacting AI? We find: 📉 A rapid decline in the consenting data commons (the web) ⚖️ Differing access to data by company, due to crawling restrictions (e.g.🔻26% OpenAI, 🔻13% Anthropic) ⛔️ Robots.txt preference protocols

Excited to be (jet lagged) at ICML presenting "Creative Text-to-Audio Generation via Synthesizer Programming" Let's talk ML for audio 🤖, synthesizers 🎹🎶, etc. (work ICML Conference w/ Nikhil Singh Jessica Shand) ctag.media.mit.edu

Thrilled to join Dartmouth CS as Assistant Professor in Jan 2025! I’m seeking 1-2 PhD students to join in Fall 2025. Application is by December 15th; please feel free to reach out with any questions. More details here: dartgo.org/ns-phd-apps-20…

A great opportunity to work on your PhD with Nikhil Singh, one of the most incredible people I've ever met!

I'm recruiting a PhD student to join the Human AI Collaboration lab at Kellogg School Northwestern University Computer Science NICO If you're excited about computational social science, LLMs, digital experiments, real-world problem solving, this could be a great fit Please reshare! Deets 👇

✨New Report✨ Our data ecosystem audit across text, speech, and video (✏️,📢,📽️) finds: 📈 Rising reliance on web, synthetic, and YouTube data. 🛑 80%+ datasets carry hidden restrictions. 🌍 Relative representation in languages and creators has not improved for 10+ yrs.

Are we sure AGI isn't available through Google Cloud, but we just can't find it?

Thrilled our global data ecosystem audit was accepted to #ICLR2025! Empirically, we find: 1⃣ Soaring synthetic text data: ~10M tokens (pre-2018) to 100B+ (2024). 2⃣ YouTube is now 70%+ of speech/video data but could block third-party collection. 3⃣ <0.2% of data from

Congratulations to @fluidinterfaces PhD student Manuel Cherep and group head Prof. Pattie Maes on receiving a 2024 Amazon Research Award for their project "Understanding How LLM Agents Deviate from Human Choices"! media.mit.edu/posts/manuel-c…

Can an AI model predict perfectly and still have a terrible world model? What would that even mean? Our new ICML paper formalizes these questions One result tells the story: A transformer trained on 10M solar systems nails planetary orbits. But it botches gravitational laws 🧵