As AI enters the trough of disillusionment, experts seek to consider its next phase. Chris Howard, Gartner’s global chief of research, discusses neuro-symbolic AI, questioning the pursuit of 100% accuracy and emphasising the power of fluency and creativity in problem-solving. He posits that combining neural and symbolic approaches can lead to innovation.
According to Howard, neuro-symbolic artificial intelligence is simply a fusion of styles of artificial intelligence.
“Neural being things like large language models, and the symbolic being things like knowledge management, knowledge graphs, that kind of technology, and how those two things really should work together in the pursuit of better results,” Howard said.
Howard believes that expecting large language models to be 100% precise is an unreachable goal.
“Look for mistakes. Be creative. Use the tools you have at your disposal for the things they're best aimed at,” he said.
Full video here
