Last week, Elon Musk predicted in a livestreamed interview that we would probably have artificial intelligence (AI) that is “smarter than any one human” by the end of next year.
He added that by the end of the decade, AI capabilities would probably be greater than “that of all of humanity combined”.
He did add a qualification, however, that this could depend on factors such as shortages of powerful chips and increasing demands for power.
Artificial general intelligence (AGI) has long been a goal of AI researchers and developers.
Precise definitions of AGI can vary, but in general terms, it refers to AI that is both powerful and versatile enough to match or outperform humans on a wide range of tasks.
AI that can significantly outperform humans is sometimes known as superintelligence.
Other prominent experts in the field of AI disagree with the SpaceX CEO and founder of xAI over when these thresholds are likely to be passed.
Some see it as much further over the horizon, while others believe that AGI is not too far away.
Predictions for AGI range from within this year to decades away
Dario Amodei, co-founder of AI safety and research company Anthropic, recently said that we could expect to see AI matching a “generally well-educated human” within the next two to three years.
Demis Hassabis, co-founder of Google’s DeepMind, is looking more towards the end of the decade.
On a podcast in February, he said that when DeepMind was started in 2010, it was envisaged as a 20-year project.
He added that they still seem to be on track for this and that he “wouldn’t be surprised” if AGI was achieved within the next 10 years.
Thanks to ChatGPT, OpenAI has been perhaps the most prominent player in the AI sphere over recent years, but chief executive Sam Altman is more circumspect on AGI, saying only that it might be achieved in the “reasonably close-ish future”.
Yann LeCun of Meta said that achieving AGI will be a process rather than an event, which could stretch on for decades and is “much harder than we think”.
He said that the whole history of AI was filled with the “obsession of people being overly optimistic” before realising that what they were attempting to achieve was far more difficult than they originally thought.
Today’s news was brought to you by TD SYNNEX – the UK’s number one solutions distributor.
