An Opportune Moment: the Global AI Safety Summit

"Responsible Scaling" means AI companies continue to make dangerous AI with no oversight. If companies capture the AI Safety Summit: humanity is at risk.

We are at a pivotal moment in human history. AI represents the largest threat to humanity this century.

For the first time, the Global AI Safety Summit has been convened, inviting leading politicians and the CEOs of large AI Companies.

But there is a problem.

AI companies are trying to rig the summit, pushing their own agendas and policies in an attempt to continue their race to godlike AI without interference.

This is despite their own acknowledgement that godlike AI could now be the greatest threat to humanity.

If the AGI companies get their proposals endorsed by governments, they will continue on the path to our possible extinction, gambling with all our lives.

Where? From 1-2 November, the UK is hosting the first ever Global AI Safety Summit. Tech CEOs, politicians, and academics are set to meet at Bletchley park to discuss the regulation of AI before it becomes beyond human control.

Who? It was announced at the beginning of October that there would be a guest list of 100 attendees. Until this point (and maybe for the foreseeable future) the full guest list remains a tightly guarded secret.

We know that the list includes high-ranking representatives from multiple countries (think Kamala Harris, Rishi Sunak, Justin Trudeau, Emmanuel Macron and Ursula von der Leyen) and CEOs and representatives from the largest, most powerful AI companies (think the likes of Sam Altman, Dario Amodei, and Demis Hassabis), alongside policymakers and members of civil society.

What? The UK Government has given the summit five objectives. AI companies aim to manipulate these aims to rig the summit towards their own agendas to remain as unregulated as possible.

The biggest AI companies (Anthropic, OpenAI and DeepMind) aim to create smarter-than-human AI. There are currently no methods by which they could ensure that this is safe. AI CEOs have even admitted that if the technology goes wrong it could be “lights out for all of us”.

Even in the face of such serious and existential risks to humanity, AI Companies want the Summit's outcomes to be underwhelming. They want to ensure there are no enforceable policy proposals, and all agreements are voluntary and non-committal.

This is a terrible, insufficient, and unacceptable response to the severity of the danger we face. AI companies cannot be granted unrestricted power over all our futures.

We are at a precipice. The summit has the potential to lead to a safer future for us all: where AI is regulated by concrete and effective policy. Or, it might be rigged by AI companies to entrench apathy and give them unrestricted power over the fate of humanity.We deserve better.