Get your free personalized podcast brief
We scan new podcasts and send you the top 5 insights daily.
Even as AI models surpass technical AGI benchmarks, the host argues people will keep moving the goalposts. The true, socially accepted definition of AGI will be its "feel"—its ability to generalize and execute complex, nuanced tasks with minimal instruction, like a human.
Related Insights
Today's AI models have surpassed the definition of Artificial General Intelligence (AGI) that was commonly accepted by AI researchers just over a decade ago. The debate continues because the goalposts for what constitutes "true" AGI have been moved.
A consortium including leaders from Google and DeepMind has defined AGI as matching the cognitive versatility of a "well-educated adult" across 10 domains. This new framework moves beyond abstract debate, showing a concrete 30-point leap in AGI score from GPT-4 (27%) to a projected GPT-5 (57%).
Hassabis argues AGI isn't just about solving existing problems. True AGI must demonstrate the capacity for breakthrough creativity, like Einstein developing a new theory of physics or Picasso creating a new art genre. This sets a much higher bar than current systems.
OpenAI's CEO believes the term "AGI" is ill-defined and its milestone may have passed without fanfare. He proposes focusing on "superintelligence" instead, defining it as an AI that can outperform the best human at complex roles like CEO or president, creating a clearer, more impactful threshold.
The pursuit of AGI may mirror the history of the Turing Test. Once ChatGPT clearly passed the test, the milestone was dismissed as unimportant. Similarly, as AI achieves what we now call AGI, society will likely move the goalposts and decide our original definition was never the true measure of intelligence.
The definition of AGI is a moving goalpost. Scott Wu argues that today's AI meets the standards that would have been considered AGI a decade ago. As technology automates tasks, human work simply moves to a higher level of abstraction, making percentage-based definitions of AGI flawed.
Sundar Pichai shares his working definition of Artificial General Intelligence (AGI), developed with DeepMind CEO Demis Hassabis. He describes it as a system that can comprehensively perform a wide range of tasks, including cognitive ones, in a way that is comparable to human ability.
Shane Legg proposes "Minimal AGI" is achieved when an AI can perform the cognitive tasks a typical person can. It's not about matching Einstein, but about no longer failing at tasks we'd expect an average human to complete. This sets a more concrete and achievable initial benchmark for the field.
A practical definition of AGI is its capacity to function as a 'drop-in remote worker,' fully substituting for a human on long-horizon tasks. Today's AI, despite genius-level abilities in narrow domains, fails this test because it cannot reliably string together multiple tasks over extended periods, highlighting the 'jagged frontier' of its abilities.
The race to manage AGI is hampered by a philosophical problem: there's no consensus definition for what it is. We might dismiss true AGI's outputs as "hallucinations" because they don't fit our current framework, making it impossible to know when the threshold from advanced AI to true general intelligence has actually been crossed.
