New: The case for AGI.

Our mission

Our mission is to do basic research aimed at human-level AI, and to ensure that human-level AI is ultimately beneficial to all.

What makes us different

Many other organizations work on AI. But most are concerned with "narrow AI"—building AIs for highly specific purposes. We are focused on human-level AI: the problem of building an AI that can handle whatever tasks we throw at it. This is a long-term, basic research project, and it needs dedicated institutions to ensure short-term goals don't crowd out long-term ones.

In addition, BasicAI is approaching human-level AI from a direction that no one else is taking. For one, we believe in taking human thought as a serious object of study. We study how humans perform demanding cognitive tasks, and use that as a basis for our models. We also place an emphasis on symbolic methods, at a time when many groups are over-committed to neural networks. These and other strategic decisions—which you can read more about below—set us apart from any other research project out there.

Strategy

Here are some of the principles that drive our research.

  • Seek to understand how humans think—don't spend too much time building systems that are nothing like human minds.
  • Focus on the necessary ingredients: if a human-level AI could learn something for itself, through experience, don't try to bake it in.
  • Focus on the aspects of cognition we don't understand very well, especially reasoning, knowledge representation, and the acquisition of knowledge through experience.
  • Study these issues in context of concrete tasks of appropriate difficulty.

You can learn more about our strategy and progress on our Publications page, especially from our report From humans to human-level AI.

Contact: info@basicai.org