Google DeepMind CEO Predicts Superintelligence Could Arrive in Just Four Years

Forecast on Artificial General Intelligence (AGI)

According to ХВИЛЯ: At the Google I/O developer conference, Demis Hassabis, CEO of Google DeepMind, made a significant announcement regarding the imminent arrival of artificial general intelligence (AGI). He projects that AGI could emerge as early as 2029, with a more likely timeframe around 2030. Hassabis urges governments and society to begin preparing for this pivotal technological milestone. This prediction comes amid rapid advancements in AI capabilities, sparking both excitement and concern across industries.

Hassabis described the current phase of AI development as a period of gentle self-improvement, distinguishing it from recursive self-improvement, which he views as a separate benchmark. During his address, he stated,

“You can imagine that the era of agents next year is something like a trial run,”

and noted that

“we are already seeing agents actually working,”

adding that he can envision how much more capable and useful they will become within a year.

Safety Concerns and Preparing for Change

Hassabis pointed to the power of Anthropic's Mythos system as an example of unpreparedness, highlighting the urgent need for safety measures. He has been discussing these measures with leaders from other AI labs, emphasizing that

“safety needs to be accelerated,”

and that

“now is the right time to strike while the iron is hot.”

He also noted that discussions about AI's potential to reshape society remain largely confined to tech circles, and that

“this needs to be taken seriously.”

Expressing concern over the lack of seriousness with which some economists approach this issue, Hassabis added that

“all leading labs are quite focused on this.”

According to him,

“there will be obvious gains in research speed,”

but also risks associated with developing such systems. The upcoming wave of AI agents should be viewed as a societal stress test, and a potential AI executive order could mandate testing new models before their release.

Hassabis's remarks underscore the importance of preparing for the challenges that AGI development may bring. Experts believe that successful integration of new technologies into society will require active discussions on ethics and safety. Given the rapid pace of AI progress, it is crucial for governments and society at large to be ready for changes that could impact the economy, jobs, and everyday life.

As the landscape of artificial intelligence evolves, concerns regarding its implications on the workforce are becoming increasingly prominent. Recently, the CEO of OpenAI acknowledged a miscalculation regarding the impact of AI on entry-level office jobs, highlighting the urgent need for a broader dialogue on the societal changes driven by these technologies. This discourse aligns with Hassabis's call for proactive measures to address the challenges posed by advancing AI capabilities.