When it comes to creating artificial general intelligence (AGI), or human-level intellect, Dr. Nando de Freitas, a lead researcher at Google's DeepMind, has proclaimed "the game is finished."


Following the release of DeepMind's Gato, Freitas turned to Twitter to respond to an opinion post by Tristan Greene, in which Greene says that humans may never achieve AGI, and that it "appears like AGI won't be occurring in our lifetimes."


DeepMind's Gato artificial intelligence (AI) is dubbed "A Generalist Agent." Freitas directed the Gato project and is one of the authors of a paper that was just published on the arXiv pre-print service.


Achieving AGI, according to Freitas, is a foregone conclusion for humanity and simply a matter of time. "What will bring AGI," according to Freitas, is overcoming the hurdles of scaling up AI models and their memory, efficiency, safety, and other aspects.


In a follow-up Twitter, Freitas added that OpenAI Chief Scientist Ilya Sutskever is correct, presumably alluding to a tweet from a few months ago in which Sutskever remarked "it may be that today's big neural networks are partly aware."