Traditional benchmarks like the Turing Test are being challenged as outdated. Mustafa Suleyman, a prominent figure in the AI community and co-founder of DeepMind, has proposed a novel approach to gauge the intelligence of AI: its ability to generate wealth.
The Turing Test, proposed by the British mathematician and computer scientist Alan Turing in 1950, is a measure of a machine's ability to exhibit intelligent behavior indistinguishable from that of a human. In the test, a human evaluator interacts with an unseen interlocutor, which can be either a human or a machine. If the evaluator cannot reliably distinguish between the machine and the human based on their responses, the machine is said to have passed the test and demonstrated human-like intelligence.
A Modern Measure of AI Capability
Mustafa Suleyman, in his upcoming book titled The Coming Wave: Technology, Power, and the Twenty-first Century's Greatest Dilemma, introduces the concept of “artificial capable intelligence.” He suggests that the true test of an AI's intelligence could be its ability to turn an initial investment of $100,000 into $1 million. This modern Turing Test would require the AI to research e-commerce opportunities, conceptualize a product, and then devise a strategy for its production and sale. Suleyman's proposal, while innovative, has drawn criticism from some quarters, with skeptics likening it to the get-rich-quick schemes promoted by certain online influencers.
AI's Role in Society: Beyond Financial Metrics
While Suleyman's financial Turing Test offers a fresh perspective on AI evaluation, it also raises questions about the broader role of AI in society. Should AI's worth be solely determined by its financial prowess, or should its capacity for human-like interaction also be considered? Suleyman's own startup, Inflection AI, has developed a chatbot named Pi, which, unlike other chatbots, is designed to be empathetic but has limitations in its capabilities. It cannot browse the internet, generate code, or replace human companionship or therapy.
Despite the advancements in AI, there remains a significant gap between the capabilities of current AI models and the vision early computer scientists had for machines that could truly “think” like humans. The debate continues on whether a computer will ever achieve genuine human-like intelligence.