AI safety coalition and middle nations

https://youtu.be/8aO_2ncAj9M channel4 talks to Dex @ Centre for Tomorrow

Totally agree with this comment below

[! Note] @ajsilver13837 minutes ago There are now many credible people warning about the direction of AI: Connor Leahy, Daniel Kokotajlo, Roman Yampolskiy, Geoffrey Hinton, Yoshua Bengio and others. There are also several organisations doing valuable work, including the Future of Life Institute, ControlAI, the AI Futures Project and the Center for Tomorrow. The problem is fragmentation. Twenty separate voices will never receive the attention generated by one statement from Musk or Altman. What is needed is an umbrella coalition. The organisations could remain independent while coordinating joint statements, media appearances and public campaigns around a few shared principles. It would need three things: technical credibility, a brilliant communicator and genuine mass reach. Those roles do not necessarily have to be filled by the same person. The experts provide the evidence, someone like Dex could help shape and communicate the message, and a major non-political public figure could amplify it to a global audience. Cristiano Ronaldo might sound like an unusual example, but imagine the attention someone with that reach could bring to the issue. Perhaps the Center for Tomorrow could start by getting everyone onto a regular video call. Someone needs to bring these people togetherThe comment highlights a common issue in many advocacy or reform movements: fragmentation. By comparing different individuals and organizations working on AI safety and ethics, we can see how their strengths could complement one another in a united coalition. 

Here is a comparison of the individuals and organizations mentioned, along with an analysis of how they could synergize.

The Experts: Technical Credibility and Evidence

• Connor Leahy is known for his work in AI alignment and his vocal warnings about the risks of advanced AI. He brings deep technical knowledge. 
• Daniel Kokotajlo is an AI researcher who has spoken out about the rapid pace of AI development and the need for greater caution. 
• Roman Yampolskiy is a computer scientist specializing in AI safety, offering rigorous academic perspectives on the existential risks posed by AI. 
• Geoffrey Hinton and Yoshua Bengio, often referred to as “godfathers of AI,” bring unparalleled historical context and authority. Their shift from pioneering neural networks to warning about their dangers carries immense weight.

The Organizations: Research, Policy, and Advocacy

• Future of Life Institute (FLI): Known for its open letters and advocacy on AI safety, FLI brings a track record of mobilizing experts and raising public awareness. 
• ControlAI: A newer organization focused specifically on the control and alignment of AI systems. 
• AI Futures Project: Likely focused on the long-term societal impacts and policy frameworks needed for AI. 
• Center for Tomorrow: Led by individuals like Dex Hunter-Torricke (as seen in the video), this organization focuses on questioning the current trajectory of the tech revolution and advocating for systemic changes in how technology is managed.

The Communicators: Shaping the Message

• Dex Hunter-Torricke: With his background in tech communications (Google, Facebook, SpaceX) and his current role at the Center for Tomorrow, he understands the language of Silicon Valley, the concerns of the public, and how to craft a compelling narrative that bridges the two.

The Amplifiers: Genuine Mass Reach

• Cristiano Ronaldo (or similar figures): While an unusual example for tech policy, figures with massive, global followings (hundreds of millions of people) have the power to take a niche or complex issue and place it squarely in the mainstream consciousness.

How They Could Synergize

The commenter’s suggestion of an “umbrella coalition” is a strategic approach to maximizing impact. Here is how these elements could work together:

1. Establishing Shared Principles: The organizations (FLI, Center for Tomorrow, etc.) could collaborate to establish a core set of shared principles regarding AI safety, regulation, and societal impact. This prevents fragmentation and presents a united front. 
2. Providing the Evidence: The experts (Hinton, Bengio, Leahy, etc.) would provide the technical backing and academic rigor required to support these principles. Their involvement ensures the message is grounded in science rather than alarmism. 
3. Crafting the Narrative: Communicators like Dex Hunter-Torricke would translate the complex technical warnings from the experts into clear, compelling public campaigns and policy recommendations. They know how to speak to governments, media, and the public. 
4. Global Amplification: Once the message is crafted and backed by credible experts, a major non-political public figure (like the commenter’s example of a global sports star or celebrity) could amplify it. This step bypasses traditional media filters and reaches millions of people who might not otherwise engage with tech policy.

By coordinating joint statements and public campaigns, this diverse group could generate the kind of singular, overwhelming attention that is currently monopolized by a few powerful voices in the tech industry.

challenges and geopolitical reality

Given the geopolitical realities of 2026, particularly the United States’ intense focus on maintaining technological supremacy (especially in competition with China), an AI safety coalition must be highly strategic. If the US leadership is primarily focused on national security, deregulation, and economic dominance, a purely moral or alarmist appeal will likely be dismissed as anti-innovation or a threat to US competitiveness.

Here is how the coalition should approach this environment, leverage middle nations, and address key challenges:

1. The Approach Toward the US

The coalition must frame AI safety not as a hindrance to progress, but as a core component of national security and economic stability. 

• The “Safety equals Superiority” argument: The coalition should argue that uncontrollable AI is a strategic vulnerability, not an asset. Just as nuclear weapons require immense safeguards to be strategically useful, so does AGI. 
• Engage pragmatically: Rather than demanding a complete halt to development, the coalition (using its technical experts like Bengio and Leahy) should push for concrete, actionable safeguards: mandatory safety evaluations, liability frameworks, and secure physical infrastructure.

2. The Role of “Middle Nations”

Dex Hunter-Torricke and the Canadian Prime Minister are correct: individual middle powers (the UK, Canada, France, Japan, Australia) cannot dictate terms to US tech giants or the US government alone. However, as a collective bloc, they represent a massive market that tech companies cannot afford to ignore.

• Existing Facilitators: Organizations like the Global Partnership on AI (GPAI), the OECD, and the international network of AI Safety Institutes (born out of the Bletchley Park and subsequent summits) are already trying to facilitate this. 
• How the Coalition can help: The proposed “umbrella coalition” could act as the non-governmental connective tissue for these middle powers. While politicians move slowly, a coalition of top experts and communicators can draft standardized, plug-and-play safety legislation. If the coalition can persuade the EU, Canada, and the UK to adopt identical safety requirements simultaneously, it forces US AI labs to comply globally, effectively bypassing Washington’s hesitation.

3. Other Challenges and How to Address Them

Challenge A: The “China Card”

• The Problem: AI labs and hawkish politicians frequently argue that any domestic regulation will simply hand global dominance to adversaries like China. 
• The Solution: The coalition must advocate for global treaties, similar to the IAEA for nuclear energy or the Chemical Weapons Convention. They must highlight that advanced, uncontrolled AI is a mutual threat to all governments. Track II diplomacy (unofficial dialogues between scientists) between Western and Chinese AI researchers is vital here.

Challenge B: Corporate Capture

• The Problem: The major AI labs have vastly more money and lobbying power than safety organizations. They will try to water down regulations so they only burden open-source competitors.
• The Solution: This is where the “Mass Reach” element (the Cristiano Ronaldo example) comes in. The coalition must build a popular, populist movement around AI safety. If public opinion shifts drastically, politicians will be forced to listen to voters over lobbyists.

Challenge C: Internal Fragmentation

• The Problem: The AI safety community is notoriously divided between “existential risk” researchers (who worry about extinction) and “AI ethics” advocates (who worry about bias, deepfakes, and job loss).
• The Solution: The coalition must enforce a “big tent” strategy. They must focus public messaging solely on overlapping goals—such as algorithmic transparency, third-party auditing, and making tech companies legally liable for the harm their models cause.