'Few years' away from human-like AI, Google Chief makes startling revelation

Hassabis believes the AGI advancement may accelerate: "I think we could be just a few years, maybe within a decade away."

However, experts are still concerned that AGI could seriously harm society if not controlled and guided by human values.

The development of generative AI, which focuses on producing text or images based on specific prompts, results from the boom in AI funding in recent years. 

Meanwhile, scientists are already focusing on AGI in addition to generative artificial intelligence (AI).

Stanford study and the letter to pause AI

In a study done by Stanford University last month, 56% of computer scientists and AI researchers claimed that AGI is now a reality. 

In contrast to generative AI, AGI is aware of what it says and does, reducing the likelihood of making incorrect or ambiguous statements.

Despite the advantages, some experts are worried about accelerating its development. 

AGI was deemed an "important concern" by almost 58% of AI professionals in the Stanford study, and 36% said it may result in a "nuclear-level catastrophe." 

Thousands of technologists and AI specialists signed an open letter in March calling for a six-month moratorium on developing advanced AI to emphasize ethics research. 

Elon Musk and Steve Wozniak, co-founders of Apple, were among those who signed the letter. 

Concerns about the dangers of unrestrained superhuman AI were expressed in the letter, which aimed to "pause AI" more potent than ChatGPT-4 for six months. 

DeepMind AI and Google's Ambitions

The recent merger of Google's primary AI research team with DeepMind and the appointment of Hassabis as CEO are both intended to hasten the company's progress toward AGI. 

Hassabis underlined that DeepMind's AGI research would not soon endanger society. 

The company's current goal is to incorporate AI into more goods, which would, in his opinion, be just as revolutionary as the first iPhone. 

Additionally, Hassabis stated that Google would responsibly create AGI and apply the "scientific method to "do very careful controlled experiments to understand what the underlying system does."

In a free-to-access database, DeepMind, which Google purchased in 2014, released over 200 million predictions after analyzing the structure of nearly every protein known at the time.

The company created technologies that reduced Google's energy costs by 40% and AI capable of identifying complicated eye disorders. 

AlphaFold, an AI system developed by DeepMind that predicted the structure of every protein known in the human genome, was another significant development with far-reaching implications for studying illness and therapeutics.