The Turing test, developed by Alan Turing in 1950, measures whether a machine can think like a human. It has been a north star in the field of artificial intelligence for decades. However, with the increasing sophistication of chatbots, the test may be outdated. Mustafa Suleyman, co-founder of AI lab DeepMind, believes that a new standard for measuring AI sophistication is needed.
In an interview with Bloomberg, Suleyman argued that the Turing test doesn’t tell us anything about what the system can do or understand, nor does it say anything about whether it has established complex inner monologues or can engage in planning over abstract time horizons, which is key to human intelligence.
Instead, Suleyman proposes a new gauge, which he referred to as the modern Turing test, based on entrepreneurship. This test requires AI to turn $100,000 in seed money into $1 million by devising an original product idea, creating an e-commerce business plan, finding manufacturers, and listing the item. This test would allow evaluators to measure AI’s ability to set goals, plan, and execute complex tasks.
While the original Turing test allowed chatbots to emulate human writing without any true level of understanding, Suleyman’s model would require machines to exhibit advanced strategy and reasoning skills that make human intelligence unique.
In his upcoming book, The Coming Wave: Technology, Power, and the Twenty-first Century’s Greatest Dilemma, Suleyman argues for this new gold standard for measuring AI sophistication. He believes that if a machine can pass this entrepreneurial test, then the consequences for the world economy are seismic.
Suleyman thinks that AI will pass his proposed update to the Turing test within the next two years. However, it remains to be seen how this new gauge will be received by the AI community.
