Перейти к содержанию сайта
“Human-compatible AI” by Stuart Russell (Data Driven Paris)
Video

“Human-compatible AI” by Stuart Russell (Data Driven Paris)

Firstmark, 2018

автоматическое преобразование текста в аудио
автоматическое преобразование текста в аудио

Editorial Rating

8

Qualities

  • Scientific
  • Eye Opening
  • Hot Topic

Recommendation

Some people view the upcoming age of artificial intelligence (AI) as a threat to humankind. How long will it take for machines to take over the world? And will they care about ordinary, mortal human beings? During a meeting of the tech community Data Driven, AI expert Stuart Russell claims that people don’t need to worry: Developers can make sure that AI will always be beneficial for humans. AI systems will need to learn about individual preferences by observing humans and then assist them in living the life they want. Stay tuned if you’d like to know how this is going to be done.

Take-Aways

  • Artificial intelligence is advancing but has yet to learn how to make decisions based on “multiple levels of abstraction,” like humans.
  • The potential benefit for economic, military and political applications serves as an incentive for people to develop human-level AI.
  • Robots need to observe the behavior of individual humans to learn about their preferences and then support them.

About the Speaker

Stuart Jonathan Russell is Professor of Computer Science and holds the Smith-Zadeh Chair in Engineering at the University of California, Berkeley. He is also the founder and leader of the Center for Human-Compatible Artificial Intelligence (CHAI) at Berkeley.


Comment on this summary or Начать обсуждение