At PopTech 2025:Enigma in Washington DC, STEM comedian and roboticist Sarah Rose Siskind unveiled FetusGPT— https://fetus-gpt.netlify.app/ a world-first experiment where her unborn baby and an artificial intelligence were both trained on the same life data: the sounds of her world. AI created by Coding Educator, Daniel Shiffman. @thecodingtrain With help from Eleven Labs sound engineer, Jordan Conragan. 🤖 About the Project FetusGPT compares natural and artificial intelligence on the same sensory diet. Two learners: baby & bot. Same input: one mother’s life. Outputs compared: first sound / first word. 🎬 What You’ll See • The conception of FetusGPT—how an AI was “born” alongside a human baby • Real audio from a @PlaudAI recorder capturing every sound in Sarah’s life • How fetuses hear through amniotic fluid vs. how AI “hears” via neural nets • Daniel Shiffman’s coding of the model (NYU Tisch School of the Arts) • FetusGPT’s first babble generated by a GAN and first words from a mini-LLM • What the experiment reveals about learning, language, and consciousness • Reflections on control, motherhood, fear, and awe 📍 Recorded Live at PopTech 2025: Enigma in Washington DC This talk explores what happens when comedy, neuroscience, and machine learning meet parenthood - routine stuff for https://poptech.org/ Subscribe: https://secure.everyaction.com/NrEjNdZeJUmtPHfFdG7T2w2 Grab a PopTech 2026 membership: https://secure.everyaction.com/o3lbl2vFnEmd0yBL33sNWg2 Explore timeless content on the PopTech YouTube Channel https://www.youtube.com/user/poptech 💬 Engage with Sarah & FetusGPT Comment → Are you #TeamBaby or #TeamBot? Like & Subscribe for more science-comedy experiments that mix AI and emotion. Follow @SarahRoseSiskind and @HelloSciCom for behind-the-scenes updates on FetusGPT and future projects like TeenGPT.

YouTube video by PopTech