|
Neuroscience & artificial intelligence: mechanisms, perspectives and consequences| title | Neuroscience & artificial intelligence: mechanisms, perspectives and consequences |
|---|
| start_date | 2024/05/23 |
|---|
| schedule | 09h-19h |
|---|
| online | no |
|---|
| location_info | / |
|---|
| summary | Artificial intelligence (AI), which is based on artificial neural networks with learning capabilities initially inspired by the functioning of the nervous system, has made significant progress in recent decades, and is playing an increasingly important role in technological development in a wide range of fields (human-machine interaction in many industries: telephony and office automation, home automation, automobile, defence, health, etc.).
Neuroscience has long been a key driver of advances in artificial intelligence (AI). Conversely, the models used in AI are used to understand how the brain works. However, the structure of the networks and the learning mechanisms currently used remain far from the principles observed experimentally. Developments of artificial systems towards forms more inspired by biology could improve the capacity of algorithms. On the other hand, a better understanding of the network mechanisms involved in AI could benefit both their development but also a better understanding of the functioning of the human brain. Neuroscience can therefore play an important role in accelerating the progress of AI and realizing its vast potential. However, the development of AI-type capabilities is also likely to raise serious ethical and societal issues, impacting on human creativity, economic growth, employment, the place of machines in society as well as their potential autonomy.
The first day of this meeting will be dedicated to the study of the network mechanisms involved in artificial networks and the underlying learning mechanisms. The second day will be dedicated to the societal consequences and ethical perspectives of the development of AI in our societies. |
|---|
| responsibles | Leblois, Procyk, Rougier |
|---|
Workflow history| from state (1) | to state | comment | date |
|---|
| submitted | published | | 2024/02/02 14:01 UTC |
| |
|