Google says it’s close to creating human-level artificial intelligence


Loud statement

Google researchers said they are close to creating Artificial General Intelligence (AGI, Strong AI), an artificial intelligence capable of performing most of the tasks that were previously only subject to humans.

Google DeepMind lead researcher Dr. Nando de Freitas boldly stated that as we scale AI, we are getting closer to AGI.

The scientist is quite optimistic about the emergence of an artificial intelligence system called Gato.

It is a general-purpose system capable of performing 604 tasks, including creating descriptions for images, having a dialogue, laying blocks with a robotic arm, and playing an arcade game.

Gato itself, however, cannot yet be called a full-fledged AGI. It should be noted that the current AI models are now mostly highly specialized, most often they solve one problem.