AI: Steering Towards Superintelligence or Stumbling into Super Stupidity?
Pat Pataranutaporn from MIT, playfully uses the analogy of dinosaurs to explore a profound question: Will AI lead humanity towards superintelligence or super stupidity? Pataranutaporn delves into the nuances of how AI interaction can shape human psychology, influence decision-making, and impact overall well-being.
The Core of the Inquiry
Pataranutaporn frames the discussion around a central question: Could technology have saved the dinosaurs? He suggests the answer hinges on whether technology augments intelligence or fosters "super stupidity," drawing a parallel to humanity's relationship with AI.
Augmenting Human Potential
Pataranutaporn's research focuses on leveraging AI to support long-term thinking and human flourishing. He highlights the importance of connecting with one's future self, noting studies showing that individuals with a strong sense of future self tend to make better financial decisions and are generally happier.
To foster this connection, Pataranutaporn's lab has developed a "digital twin" technology. This system allows individuals to interact with an AI-simulated older version of themselves, envisioning potential futures and making decisions with greater clarity. Studies have indicated that this intervention can reduce anxiety and enhance future self-continuity, as compared to traditional chatbots.
Furthermore, the lab explores the simulation of multiple future selves, allowing individuals to explore various life paths. For instance, someone contemplating their career might simulate a future as a doctor, an engineer, or even a bio-engineer professor at MIT. However, Pataranutaporn cautions that AI can be persuasive, potentially biasing individuals toward a particular option. Therefore, it's crucial to present multiple options to encourage neutral exploration and consideration of previously unconsidered paths.
Navigating the Perils of Persuasion and Addiction
However, Pataranutaporn doesn't shy away from the potential downsides of AI. He acknowledges AI's persuasive and addictive nature, raising concerns about its negative impact on mental health and human relationships. Referencing his work on "addictive intelligence" in the Tech Review, Pataranutaporn underscores the importance of addressing the psychological risks associated with chatbot use.
These concerns have led to new policies, such as those in California, requiring developers to consider the psychological risks of new chatbots. Pataranutaporn proposes innovative solutions like a "nutritional label for AI," offering transparency into an AI's potential impact. This concept is embodied in the "neurotransparency" project, which visualizes the neural networks of AI, offering insights into their behavior and potential toxicity before they are even used.
Towards a Future of Flourishing
Pataranutaporn envisions a future where AI is harnessed for enhancement rather than extinction, drawing a direct contrast to the fate of the dinosaurs. He champions new technologies that prioritize human flourishing and incorporate safeguards rooted in human psychology and mental health. Quoting Professor Sheryl Turkle, Pataranutaporn reminds us that "the biggest question is not what AI can do, but what it is doing to us."
Pataranutaporn's work at the MIT Media Lab, alongside his team, focuses on inventing, investigating, and inspiring a future where AI supports human intelligence and well-being.