

Reading Brain Waves, AI Goliaths, & more on COVID-19
Apr 11, 2020
Dive into the world where AI meets neuroscience, revealing a system that decodes brain activity into written words. Explore the challenges of predictive sociology and the limitations of using data to forecast complex human behaviors. Discover how AI optimizes operations for large firms while highlighting barriers for smaller players. Plus, hear about the drive for diversity in AI development and the innovative ways AI has been applied to tackle the challenges of the pandemic.
AI Snips
Chapters
Transcript
Episode notes
Brainwaves to Text
- Scientists turned brain activity into text by tracking neural data from volunteers with implanted electrode arrays.
- The system learned to map brainwaves to spoken words, showing potential for aiding communication in patients unable to speak.
System Limitations
- The system is limited, performing well on trained sentences but struggling with new ones.
- It likely relies on a mix of memorization, word identification, and general pattern recognition.
Predicting Life Outcomes
- Researchers tried predicting children's life outcomes using 13,000 data points from 4,000 families over 15 years.
- No model, simple or complex, achieved reasonable accuracy, highlighting the complexity of such predictions.