Amrit1

Amrit Sidhu-Brar

Will1

William MacAskill

Ryan G profile

Ryan Greenblatt

Stefan Torges photo

Stefan Torges

Fin1

Fin Moorhouse

Rose1

Rose Hadshar

Eric Drexler

Philip Trammell

Tom Houlden

Katie Resized-402x503

Katie Hearsum

max-daniel-photo

Max Daniel

Tom 2

Tom Davidson

Lukas Finnveden photo

Lukas Finnveden

Lizka1

Lizka Vaintrob

Max1

Max Dalton

Toby1

Toby Ord

Daniel Eth

Owen Cotton-Barratt

Carl Shulman

Jess Riedel

Jared Leibowich

Nikola Jurkovic

subscribeSplashSection

Stay up to speed with our latest research.

subscribeFollowSection

Subscribe

When countries grow at the same exponential rate, they maintain their relative sizes. But after we develop AGI, there may be a period of superexponential growth, with growth becoming faster and faster over time. If this superexponential growth lasts for long enough, the leader could pull further and further ahead of the others, eventually producing &gt;99% of global output, and outgrowing the rest of the world combined. This post gives a basic economic analysis of this dynamic and argues that the leading country in AI development could outgrow the world, but only if it was trying hard to do so.

Could one country outgrow the rest of the world?

How to make the future better (other than by reducing extinction risk)

A summary of our new paper, "Persistent Path-Dependence."

The trajectory of the future could soon get set in stone

A summary of our new paper, “Convergence and Compromise”.

Will morally motivated actors steer us towards a near-best future?

AI systems may soon fully automate AI R&amp;D.

How Quick and Big Would a Software Intelligence Explosion Be?

In the new paper “No Easy Eutopia”, we argue that reaching a truly great future will be very hard.

Is eutopia the default outcome post-AGI?

Should we aim for flourishing over mere survival?

Once AI can automate human labour, positive feedback loops could lead to a rapid increase in physical capabilities as well as cognitive ones.

The Industrial Explosion

Some are concerned that future AI systems could take over the world.

Human Takeover Might be Worse than AI Takeover

The development of AI that is more broadly capable than humans will create a new and serious threat: AI-enabled coups. An AI-enabled coup is most likely to be staged by leaders of frontier AI projects, heads of state, and military officials; and could occur even in established democracies.

AI-enabled coups: how a small group could use AI to seize power

Forethought

Will MacAskill discusses his new research series ‘Better Futures’.

➡️ Read the papers at forethought.org/research/better-futures.

00:00:00 Intro
00:02:38 Survival vs Flourishing
00:09:40 When is it more important to ensure flourishing vs survival?
00:15:30 Are upside-focused interventions tractable and neglected?
00:31:30 Sailing to eutopia
00:40:52 Is a future moral catastrophe likely?
01:01:30 Does the value of the entire future face diminishing returns?
01:24:20 Learning from historical utopian failures
01:36:15 Thinking about future value
01:58:15 Blockers to better futures
02:12:53 Digital minds and weird futures
02:31:38 Why institutions matter for making the future great
02:50:10 Closing thoughts

Survival is Insufficient: on Better Futures, with Will MacAskill | ForeCast

Follow our research

Could one country outgrow the rest of the world?

How to make the future better (other than by reducing extinction risk)

The trajectory of the future could soon get set in stone

Will morally motivated actors steer us towards a near-best future?

How Quick and Big Would a Software Intelligence Explosion Be?

Listen to our Podcast

All episodes

Watch our recorded episodes

Survival is Insufficient: on Better Futures, with Will MacAskill | ForeCast

Listen to our podcast

Subscribe

Follow our research

Subscribe to our newsletter

Could one country outgrow the rest of the world?

How to make the future better (other than by reducing extinction risk)

The trajectory of the future could soon get set in stone

Will morally motivated actors steer us towards a near-best future?

How Quick and Big Would a Software Intelligence Explosion Be?

Listen to our Podcast

All episodes

Watch our recorded episodes

Survival is Insufficient: on Better Futures, with Will MacAskill | ForeCast

Listen to our podcast

Subscribe