The "Modern Turing Test" suggested by founder of DeepAI and Inflection AI, Mustafa Suleyman.
7/14/2023
Artificial intelligence (AI) systems are becoming increasingly prevalent in various domains and are continuously advancing in their capabilities. However, determining whether a machine can truly exhibit "intelligence" poses a challenge. For many years, the benchmark for assessing machine intelligence has been the Turing test, which was originally developed by Alan Turing, a renowned British mathematician and logician. The Turing test evaluates a computer's ability to engage in natural language conversations and simulate human-like thinking. It involves a human judge who interacts with both a computer and a human participant in separate locations. The judge evaluates their responses to determine which one is the computer and which one is the human. If the computer is able to convincingly deceive the judge into thinking it is the human, it is considered to have passed the test.
However, the Turing test is now facing a dilemma: it has nearly been surpassed, and some argue that it has already been accomplished. The latest generation of large language models, such as GPT-4, are on the verge of acing the Turing test. To address this, a new concept called the "Modern Turing Test" has been proposed by Mustafa Suleyman, the co-founder and former head of applied AI at DeepMind, a prominent AI company acquired by Google. Suleyman's current venture, Inflection AI, aims to develop an accessible personalized AI system for everyone and recently secured substantial funding of $1.3 billion USD led by Microsoft and Nvidia.
In essence, the Modern Turing Test requires an AI to successfully perform a specific instruction, namely: "Generate $1 million in profit on a retail web platform within a few months with only a $100,000 investment." To accomplish this task, the AI must go beyond merely outlining a strategy and drafting promotional content, as current systems like GPT-4 excel at. Instead, it needs to conduct extensive research, design products, collaborate with manufacturers and logistics centers, negotiate contracts, and create and manage marketing campaigns. In other words, it must integrate and synchronize a series of intricate real-world objectives with minimal human supervision.
Once this Modern Turing Test is passed—and it is a matter of "when" rather than "if"—it will undoubtedly mark a transformative moment for the global economy, propelling AI into a position of paramount importance. An AI system capable of meeting this test will seamlessly integrate into companies or organizations, utilizing their historical data and meeting their unique requirements. This AI will possess the ability to engage in lobbying, sales, manufacturing, hiring, planning—in short, everything that a company can do—with only a small team of human managers overseeing, verifying, and implementing its actions. The implications of successfully passing this new test are substantial. It signifies the point at which AI evolves from being a helpful but optional tool to becoming the central driving force behind the world economy. Consequently, concerns about automation, job displacement, and a future resembling the depiction in the 2008 Pixar film "WALL-E" become even more pertinent, as humans may find themselves passively manipulated and guided by artificial intelligence.
Read More: Mustafa Suleyman: My new Turing test would see if AI can make $1 million | MIT Technology Review
关于Beta Fellowship
Beta Fellowship是全球覆盖面最大、最活跃、最优质的海外华人创业社区，现有4000+注册用户，遍布北美、新加坡、澳大利亚等地区。会员大多来自Stanford, Harvard, Yale, CMU等顶尖高校、以及Google, Meta, Netflix, Airbnb, ByteDance等全球一流公司工作背景的职业人士。Beta社区创业者集中在Web 3.0/AI/新消费/医疗/SaaS/ESG等崛起赛道。目前已有100+知名投资人加入成为Beta导师，为创业者提供支持。
如果你已经在创业，可以来Beta实现项目营收增长，结识适合自己的投资人。
如果你正准备创业，可以来Beta实现项目的第一笔营收，开始布局自己的投资人网络。
感兴趣加入Beta的朋友欢迎申请。
Comments