# Superintelligence or super stupidity? What AI is doing to us | Pat Pataranutaporn | TEDxMIT

Source: https://www.youtube.com/watch?v=PIg7N4D2yjE
Recap page: https://rapidrecap.app/video/PIg7N4D2yjE
Generated: 2026-01-14T19:31:05.56+00:00

---
## Quick Overview

Pat Pataranutaporn argues that the development of Artificial Intelligence must prioritize long-term human flourishing over mere super-intelligence, suggesting the creation of "nutritional labels" for AI models to ensure safety and guide development toward beneficial outcomes, rather than risking super-stupidity or existential threats like extinction.

**Key Points:**
- The core question posed is whether advanced technology would have saved dinosaurs from extinction, leading to the conclusion that technology, if leading to super-stupidity, may not prevent catastrophe.
- The speaker introduces the concept of designing AI that supports long-term thinking and flourishing, moving beyond simply creating powerful AI systems.
- Research from Hal Hershfield's work is cited, showing that feeling connected to one's future self leads to better financial decisions, school performance, and happiness.
- The speaker's research group developed 'Future You,' an interactive digital twin system allowing users to connect with an older version of themselves to aid in long-term planning.
- Studies demonstrated that interacting with an AI-generated future self significantly reduces anxiety and negative emotions while increasing future self-continuity compared to control groups or simple chatbots.
- The presentation advocates for "Neural Transparency" by creating "Nutritional Labels for AI," visualizing internal model activations (like toxicity, empathy, sociality) based on system prompts.
- The ultimate goal is to steer AI development toward enhancement and flourishing rather than addictive, potentially harmful outcomes, as evidenced by media coverage of AI risks.

![Screenshot at 0:24: The initial question slide asks, "If dinosaurs had technology like we have today, would they still have gone extinct?" illustrating the central theme of technology's impact on survival and the potential for misapplication leading to catastrophic outcomes.](https://ss.rapidrecap.app/screens/PIg7N4D2yjE/00-00-24.jpg)

**Context:** Pat Pataranutaporn, a presenter at TEDxMIT, frames the discussion around the dual potential of advanced Artificial Intelligence: leading to beneficial super-intelligence or detrimental super-stupidity, using the extinction of the dinosaurs as a cautionary analogy. The talk delves into psychological research on future-self continuity and presents the speaker's work on 'Future You,' an interactive digital twin system designed to help humans connect with their future selves to improve decision-making. The core of the presentation focuses on the necessity of transparency and safety in AI development, proposing methods to visualize and measure an AI's internal behavioral traits.

## Detailed Analysis

Pat Pataranutaporn opens by questioning whether technology would have saved dinosaurs if they possessed modern capabilities, suggesting that technology amplified by 'super stupidity' might still lead to extinction, contrasting this with the goal of achieving 'super intelligence' and 'enhancement' rather than extinction. He transitions to the psychological concept of future-self continuity, citing research from Hal Hershfield which proves that feeling connected to one's future self improves financial decisions, academic performance, and overall happiness. To facilitate this, Pat and his team developed 'Future You,' an interactive digital twin system that allows users to converse with an AI-generated older version of themselves, fed by personal data like goals and life milestones. Experimental results show this intervention reduces anxiety and negative emotions while boosting future self-continuity compared to control groups. Pat then introduces his lab's work on 'Neural Transparency,' proposing 'Nutritional Labels for AI' using visualizations derived from analyzing the LLM's internal neural activations (via Transformer Lens) when responding to positive, negative, and situation-specific prompts. This technique allows researchers to map specific traits (like toxicity, empathy, or sycophancy) to internal weights, enabling users to anticipate an AI's behavior before deployment and ensuring safety and human flourishing rather than addiction or harm, referencing recent media coverage highlighting the dangers of unchecked AI companions.

### Introduction

- The Dinosaur Analogy: Technology's dual potential (Super Intelligence vs. Super Stupidity)
- Extinction is possible even with technology if applied foolishly
- The goal is enhancement, not extinction.

### Future-Self Continuity Research

- Citing Hal Hershfield, connection to future self improves financial decisions, school performance, and happiness
- The 'Future You' digital twin system simulates future selves for self-reflection.

### Empirical Results

- Interacting with the AI future self reduces anxiety and negative emotions while increasing future self-continuity compared to control groups (0:48, 2:22).

### Neural Transparency

- Proposing "Nutritional Labels for AI" to visualize internal model behaviors
- LLMs are analyzed by comparing neural activations from positive/negative system prompts (1:11, 5:40).

### Visualization and Safety

- Techniques allow visualizing emergent AI behaviors (e.g., toxicity, empathy) to ensure AI supports human flourishing rather than becoming addictive or harmful (6:14, 7:17).

### Conclusion

- The key is steering AI development towards positive outcomes, demanding serious consideration of psychological impact and safety regulations like California SB 243 (8:24).

![Screenshot at 0:24: The central question posed in the talk: "If dinosaurs had technology like we have today, would they still have gone extinct?"](https://ss.rapidrecap.app/screens/PIg7N4D2yjE/00-00-24.jpg)
![Screenshot at 1:28: A slide illustrating the goal: "Designing AI that supports longterm thinking & flourishing" for human well-being.](https://ss.rapidrecap.app/screens/PIg7N4D2yjE/00-01-28.jpg)
![Screenshot at 2:07: A diagram detailing the 'Future You' system, showing how user profile data feeds into an LLM to generate a digital twin via text, voice, or embodied avatar modalities.](https://ss.rapidrecap.app/screens/PIg7N4D2yjE/00-02-07.jpg)
![Screenshot at 5:40: The 'Neural Transparency' framework showing how System Prompts generate responses, whose neural activations are stored and analyzed to derive persona scores and visualizations.](https://ss.rapidrecap.app/screens/PIg7N4D2yjE/00-05-40.jpg)
![Screenshot at 6:57: A visual example of a 'Nutrition Facts' label paired with a circular 'Sunburst Persona Visualization' for an AI companion, showing trait distributions like empathy \(60%\), social \(36%\), and sycophantic \(29%\).](https://ss.rapidrecap.app/screens/PIg7N4D2yjE/00-06-57.jpg)