In the field of artificial intelligence and so-called deep machine learning there is already a lot of groundwork. However, all the currently available self-learning neural networks have one important point: they are not completely independent. That is, they need to initially set certain parameters by which they will work. At the same time to learn from scratch, as does any person in your life, no electronic brain can not. And that’s a learning AI and plan to create experts from MIT and Honda.
As reported by the editors of Engadget, the new AI will be called Curious Minded Machine, and the project should take about three years. The ultimate goal of the developers, according to one of the authors of the project Soshi IBA research Department of Honda, is to create new types of AI that will be able to learn continuously, regardless of the original data.
“We are creating algorithms that will allow not only to study but also to interact with the world around. We want to develop intelligent machines that use elements such as curiosity and thirst for knowledge to serve the common good, understanding the needs of the people, expanding human capabilities and, ultimately problem solving.”
As an example, the authors cite the fact that the system can observe the work of people in a particular area. And then, analyzing these data and examining this area in more detail, the AI will be able to provide better and more efficient ways to perform these tasks.
In this project, Honda will act as an investor and project Manager, providing the data to their practices in the field of AI. MIT also will create a system responsible for forecasting different events and the consequences of their actions, as well as information processing.
In addition to MIT and Honda in the project will involve experts from the University of Washington and the University of Pennsylvania. The first will be to develop an algorithm of “curiosity” (curiosity), based on data about how learning about the world children. And their colleagues from Pennsylvania will create a “simulated biological interactions” — an analog of the learning processes that are inherent in people, adapted for use with AI.
This and other news you can discuss in our chat in Telegram.