This is the opinion of Boris Katz, a researcher at MIT's Laboratory of Computer Science and Artificial Intelligence.
Boris Katz is dedicated to the study and development of computer language skills for more than 40 years. It is known for the design and implementation of START, an online question and answer system that aims to transmit to the user the information he seeks by presenting a natural language.
However, in an interview with Technology Review, the researcher expressed his skepticism about the direction that language has been following in Artificial Intelligence.
The MIT expert said that while he was proud of the popularity of virtual assistants such as Siri or Alexa, they also felt embarrassed that they were "incredibly stupid."
For Boris Katz, the path will have to go through a greater understanding of human intelligence, calling on researchers to "leave their offices and speak to people from other areas."
According to Katz, "Artificial Intelligence needs to assimilate ideas from fields such as developmental psychology, cognitive science or neuroscience, to reflect what is already known about how humans learn and understand the world."
The researcher believes that "all progress in the field of machine learning is based on ideas from 20 or 25 years ago" and that now will require another type of approach for this technology to continue to evolve.
Boris Katz admits that "the challenge of creating truly intelligent machines is very difficult", but in his opinion this "is also one of the most important that we face."