The Narrow Path - the big picture

While most AI development is beneficial, artificial superintelligence threatens humanity with extinction. We have no method to currently control an entity with greater intelligence than us. We currently have no ability to predict the intelligence of advanced AIs prior to developing them, and we have incredibly limited methods to accurately measure their competence after development.

We now stand at a time of peril. Companies across the globe are investing to create artificial superintelligence – that they believe will surpass the collective capabilities of all humans. They publicly state that it is not a matter of “if” such superintelligence might exist, but “when”.

We do not know how to control AI vastly more powerful than us. Should attempts to build superintelligence succeed, this would risk our extinction as a species. But humanity can choose a different future: there is a narrow path through.

A new and ambitious future lies beyond a narrow path. A future driven by human advancement and technological progress. One where humanity fulfils the dreams and aspirations of our ancestors to end disease and extreme poverty, achieves virtually limitless energy, lives longer and healthier lives, and travels the cosmos. That future requires us to be in control of that which we create, including AI.

======

The above paragraphs appear at the start of the document "A Narrow Path" which was recently published by Control AI. The document sets out a comprehensive policy plan split into three phases: Safety, Stability, and Flourishing. To be clear, this is no anti-AI campaign, but rather an initiative to “build a robust science and metrology of intelligence, safe-by-design AI engineering, and other foundations for transformative AI under human control”.

Shortly after the launch of "A Narrow Path", another document was published: "The Compendium", which covers what can be seen as the larger picture behind the thinking of "A Narrow Path".

The Compendium starts as follows:

Today, ideologically motivated groups are driving an arms race to AGI (Artificial General Intelligence), backed by Big Tech, and are vying for the support of nations. If these actors succeed in their goal of creating AI that is more powerful than humanity, without the necessary solutions to safety, it is game over for all of us. There is currently no solution to keep AI safe.

In order to do something about these risks, we must understand them fully.

The Compendium aims to present a coherent worldview explaining the race to AGI and extinction risks and what to do about them, in a way that is accessible to non-technical readers who have no prior knowledge about AI. You can read this end-to-end or a-la-carte. Each section is standalone.

======

Both documents are full of rich analyses, and are likely to transform public discussion about the future of AI. Both are described as work-in-progress, which means that each of us has the chance to suggest improvements.

======

In this webinar, you will hear from two of the authors of these documents, along with a number of expert commentators who bring their own varied perspectives.

The webinar will include time for audience questions and feedback.

This event will be hosted on Zoom. To register, visit https://us02web.zoom.us/webinar/register/WN_Ve2NriaoTpC4LWXl4Dmf0w.

There will be no charge to attend the webinar.

======

About the panellists:

• Chris Scammell is the COO of Conjecture and one of the principal authors of The Compendium
• Andrea Miotti is the Executive Director of Control AI and the lead author of A Narrow Path
• Robert Whitfield is Chair of Trustees of One World Trust
• Mariana Todorova is a core member of the team in the Millennium Project studying scenarios for the transition between AI and AGI
• Daniel Faggella is the Founder and Head of Research of Emerj

======

The webinar will start broadcasting at 4pm UK time on Sat 9th November. To find this time in other timezones, use https://www.timeanddate.com/worldclock/converter.html?iso=20241109T160000&p1=136

Please log into Zoom up to 10 minutes ahead of the start time of the event, so you won't miss the start of the live broadcast.

As the discussion proceeds, attendees will be welcome to raise questions and vote to prioritise questions raised by others.

======

To register for this event, click here.