{"version":"1.0","type":"rich","provider_name":"Acast","provider_url":"https://acast.com","height":250,"width":700,"html":"<iframe src=\"https://embed.acast.com/$/683f0798c966cde736234a29/69ef887eca8040fd318355d1?\" frameBorder=\"0\" width=\"700\" height=\"250\"></iframe>","title":"Teaching AI empathy using brain signals","description":"<p>AIs could get much better at understanding what we truly value if we gave them access to our brain signals. And doing that is becoming easier than ever before.</p><p><br></p><p>In this episode, we talk with Thorsten Zander, professor at Brandenburg University of Technology and co-founder of Zander Labs. He coined the concept of <em>passive brain-computer interfaces</em>: devices that read brain signals to decode a user's mental state, non-invasively and without any effort on their part.&nbsp;</p><p><br></p><p>We cover:</p><ul><li>What non-invasive brain-computer interfaces (BCIs) can actually pick up from brain signals, and why that's very different from reading your thoughts or internal monologue</li><li>The hardware and software breakthroughs that are finally making passive BCIs wearable and affordable</li><li>How continuous neural feedback could dramatically improve AI training compared to current methods based on human ratings</li><li>Why Thorsten believes passive BCIs may offer the most concrete path to solving the AI alignment problem</li><li>The risk of social networks exploiting unconscious brain reactions to manipulate people, and why regulation alone is unlikely to be enough</li></ul><p><br></p><p><br></p><p>0:00 Cold open</p><p>0:56 What are passive brain-computer interfaces, and how are they different from Neuralink?</p><p>3:23 What are the applications of passive brain-computer interfaces?</p><p>4:33 What people get wrong about BCIs: reading thoughts vs. mental states</p><p>6:14 How passive BCIs could transform AI training and help AI understand you better</p><p>11:40 The misuse risk: how social networks could exploit unconscious brain reactions to manipulate political opinions</p><p>16:00 How close is mass adoption? The hardware and software breakthroughs making BCIs wearable</p><p>20:08 Why Germany's cybersecurity agency invested €30M in passive BCI research</p><p>24:22 Invasive vs non-invasive: how Europe and the US are taking different approaches to brain-computer interfaces</p><p>28:52 Should AI act on your first instinct?&nbsp;</p><p>32:56 How passive BCIs could solve the AI alignment problem (and why previous approaches have fallen short)</p><p>35:26 From professor to startup founder: what Thorsten learned making the leap</p><p>41:27 Best case scenario: what the world looks like when AI truly understands human values</p><p>46:03 How to get started in neuroadaptive AI and passive BCIs</p><p>48:18 The best advice Thorsten ever received</p><p><br></p><p><br></p>","author_name":"Foresight Institute"}