NTT Research Launches New Physics of Artificial Intelligence Group at Harvard

When a parent teaches his young child to relate to the world, they teach through associations and the identification of patterns. Take, for example, the letter S. Parents show their child enough examples of the letter and soon they will be able to identify other examples in contexts where guidance is not active; School, a book, a billboard.

Much of the ever -emerging artificial intelligence (AI) technology was taught In the same way. Researchers gave the system correct examples of something they wanted it to recognize, and just like a young child, AI began to recognize patterns and extrapolate such knowledge to contexts that it had never experienced before, so that his own “neural network” was formed for categorization. Just like human intelligence, however, experts lost sight of the input that informed AI’s decision -making.

The “Black Box -problem” of AI thus emerges as the fact that we do not fully understand how or why an AI system makes connections, nor the variables that play in its decisions. This issue is especially relevant in trying to improve the reliability and safety of systems and to determine the administration of AI acceptance.

From an AI-driven vehicle that does not brake over time and hurts pedestrians, to AI-Relian health equipment that helps doctors diagnose patients, and prejudices exhibited by AI hiring screening processesThe complexity behind these systems has led to the rise of a new field: the physics of AI, which strives to further establish AI as a tool for people to achieve a higher concept.

Now a new independent study group will tackle these challenges by combining the areas of physics, psychology, philosophy and neuroscience in an interdisciplinary exploration of AI’s mysteries.

The newly announced Physics of artificial intelligence group is a spin-off of the NTT Research’s Physics & Informatics lab (Phi) and was unveiled last week at the upgrade 2025 conference of NTT in San Francisco, California. It will continue to promote the physics of artificial intelligence approach to understand AI that the team has investigated in the last five years.

Dr. Hidenori Tanaka, who is a PhD in Applied Physics & Computer Science and Engineering of Harvard University, will lead the new research group, building on its earlier experience in NTT’s Intelligent Systems Group and the AI ​​research program of CBS-NTT in the physics of intelligence at Harvard.

“As a physicist, I am enthusiastic about the subject of intelligence, because, mathematically, how can you think of the concept of creativity? How can you even think about kindness? These concepts would have remained abstract if it is not for AI. It is easy to speculate, this is my definition of kindness, but now we want a kind of math kindness isFor example, “said Dr. Tanaka me last week on the sidelines of the upgrade conference.

Early in their research, the Phi -Lab acknowledged the importance of understanding the ‘black box’ of AI and Machine Learning to develop new systems with an improved energy efficiency for calculation. However, AI’s progress in the past mid -decade has always called for important considerations of safety and reliability, which have therefore become crucial for industrial applications and management decisions on AI acceptance.

Through the new research group, NTT research will tackle the similarities between biological and artificial intelligences, so that the complexity of AI mechanisms is unraveled and more harmonious merger of cooperation between Mensai.

Although new in the integration of AI, this approach is not new. Physicists have been trying to reveal the precise details of technological and human relationships for centuries, from the Studies of Galileo Galilei about how objects move and his contribution to how the steam engine informed the insights of thermodynamics during the industrial revolution. In the 21st century, however, scientists try to understand how AI works in terms of trained, collecting knowledge and making decisions so that more coherent, safe and reliable AI technologies can be designed in the future.

“AI is a neuronetwork, the way in which it is structured is very similar to how a human brain works; neurons connected by Synapsen, all of whom are represented by figures in a computer. And then we believe that there can be physics … Physics is about taking something from the universe, formulating mathematical hypotheses about their inner work, and tested, Dr. said Dr.” Dr. Dr. Hanaka.

The new group will continue to collaborate with the Harvard University Center for Brain Science (CBS) and is planning to collaborate with Stanford University University teacher Suya Ganguli, with whom Dr. Tanaka has authorized different articles together.

Dr. However, Tanaka emphasizes that an approach to natural sciences and cross-industry will be fundamental. In 2017, when he was a PhD student at Harvard, the researcher realized that he wanted to do more than traditional physics, and to follow in the footsteps of his predecessors, from Galilei to Newton and Einstein, to open new conceptual worlds in physics.

“Currently, AI is the only subject I can talk to everyone about. As a researcher it is great because everyone is always aware of talking about AI, and I also learn from every conversation because I realize how people see and use ai differently than academic contexts. I see the mission of NTT as the catalyst to arouse these conversations,” Dr. Tanaka handled.