Teaching AI empathy using brain signals

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.

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 passive brain-computer interfaces: devices that read brain signals to decode a user's mental state, non-invasively and without any effort on their part. 

We cover:

• 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
• The hardware and software breakthroughs that are finally making passive BCIs wearable and affordable
• How continuous neural feedback could dramatically improve AI training compared to current methods based on human ratings
• Why Thorsten believes passive BCIs may offer the most concrete path to solving the AI alignment problem
• The risk of social networks exploiting unconscious brain reactions to manipulate people, and why regulation alone is unlikely to be enough

0:00 Cold open

0:56 What are passive brain-computer interfaces, and how are they different from Neuralink?

3:23 What are the applications of passive brain-computer interfaces?

4:33 What people get wrong about BCIs: reading thoughts vs. mental states

6:14 How passive BCIs could transform AI training and help AI understand you better

11:40 The misuse risk: how social networks could exploit unconscious brain reactions to manipulate political opinions

16:00 How close is mass adoption? The hardware and software breakthroughs making BCIs wearable

20:08 Why Germany's cybersecurity agency invested €30M in passive BCI research

24:22 Invasive vs non-invasive: how Europe and the US are taking different approaches to brain-computer interfaces

28:52 Should AI act on your first instinct? 

32:56 How passive BCIs could solve the AI alignment problem (and why previous approaches have fallen short)

35:26 From professor to startup founder: what Thorsten learned making the leap

41:27 Best case scenario: what the world looks like when AI truly understands human values

46:03 How to get started in neuroadaptive AI and passive BCIs

48:18 The best advice Thorsten ever received

On the Existential Hope Podcast hosts Allison Duettmann and Beatrice Erkers from the Foresight Institute invite scientists, founders, and philosophers for in-depth conversations on positive, high-tech futures.

Full transcript, listed resources, and more: https://www.existentialhope.com/podcasts

Follow on X.

Hosted on Acast. See acast.com/privacy for more information.