---
title: "The emergence of superintelligent AI - The Intercom Blog"
description: "In their podcast, Intercom’s Fergal Reid and Emmet Connolly examine superintelligent AI, where systems recursively improve beyond human intelligence, now considered a realistic possibility after evidence like Othello-GPT demonstrating reasoning. Researchers from the Max Planck Institute and the Machine Intelligence Research Institute warn that containing such systems may be extremely challenging or impossible, posing an existential risk to humanity. The discussion weighs the need to regulate frontier labs closely while avoiding overregulation that curtails benefits, and notes that computation’s ubiquity makes bans difficult. They address the alignment problem and the debate between doomers and skeptics, arguing that AI is net positive up to human-level intelligence. However, they caution that the threshold beyond human-level requires serious societal engagement, and that the timeline remains uncertain, ranging from years to decades."
---

# The emergence of superintelligent AI - The Intercom Blog

[Intercom](https://yomu.fyi/company/intercom) · Oct 5, 2026

**Type:** Explainer

## Summary

In their podcast, Intercom’s Fergal Reid and Emmet Connolly examine superintelligent AI, where systems recursively improve beyond human intelligence, now considered a realistic possibility after evidence like Othello-GPT demonstrating reasoning. Researchers from the Max Planck Institute and the Machine Intelligence Research Institute warn that containing such systems may be extremely challenging or impossible, posing an existential risk to humanity. The discussion weighs the need to regulate frontier labs closely while avoiding overregulation that curtails benefits, and notes that computation’s ubiquity makes bans difficult. They address the alignment problem and the debate between doomers and skeptics, arguing that AI is net positive up to human-level intelligence. However, they caution that the threshold beyond human-level requires serious societal engagement, and that the timeline remains uncertain, ranging from years to decades.

## Context

Podcast discussion between Intercom’s Director of Machine Learning and VP of Product Design on the emergence, risks, and regulation of superintelligent AI.

## Takeaways

- Recent evidence from systems like Othello-GPT suggests AI is starting to understand and reason, shifting perspectives on the plausibility of superintelligence.
- Researchers warn that containing superintelligent AI may be extremely challenging or impossible, posing an existential risk.
- Regulating AI is difficult due to the ubiquity of computation, requiring a balance between oversight of frontier labs and allowing non-dangerous AI to progress.

## FAQ

### What evidence supports the possibility of superintelligence?

Systems like Othello-GPT, trained on game moves, demonstrate reasoning about board state, showing AI understanding. Researchers previously skeptical now acknowledge this shift, increasing confidence in the plausibility of superintelligence.

### What are the main regulatory challenges for AI?

Computation is everywhere, making bans difficult. Overregulation can curtail benefits, so frontier labs need close oversight while non-dangerous AI should maintain a 'move fast' approach to allow progress.

### Is superintelligence likely to happen soon?

The timeline is uncertain, ranging from years to decades. Recent evidence increases confidence, but the exact threshold is unknown, and the discussion emphasizes that society should engage with these issues soon.

**Tags:** [AI](https://yomu.fyi/topic/ai), [AI Governance](https://yomu.fyi/topic/ai-governance), [Superintelligence](https://yomu.fyi/topic/superintelligence)

- Source: [Intercom](https://www.intercom.com/blog/videos/the-emergence-of-superintelligent-ai)
- Source URL: https://www.intercom.com/blog/videos/the-emergence-of-superintelligent-ai
- Ingested by Yomu: 2026-10-05T18:00:46.474Z

[Read original post](https://www.intercom.com/blog/videos/the-emergence-of-superintelligent-ai)