LAISH Discussion "The Intelligence Curse"
Join us for an open discussion of The Intelligence Curse by Luke Drago and Rudolf Laine, a six-chapter essay series about what happens to ordinary people if AI automates most work.
The authors argue that today’s states and companies have an incentive to invest in regular people because they depend on human labour and taxes. If AGI removes that dependence, the incentive weakens. They compare it to the “resource curse,” where governments rich in natural resources neglect their citizens. Powerful AI would hollow out organizations from the bottom up, close off social mobility and make capital and control over AI far more important than human talent.
The authors don’t treat this as inevitable. They propose three kinds of technology and institutions to break the curse:
Avert AI catastrophes by hardening the world, without relying on centralized control
Diffuse AI, building tools that augment people and put control in regular people’s hands
Democratize institutions, keeping them tied to human needs as AI grows more powerful
We’ll begin with a short recap of the argument, then open the floor. Questions we might explore:
Is the resource-curse analogy convincing? Where does it hold, and where does it break down?
Will AI really hollow out organizations from the bottom up, and are we already seeing it in entry-level hiring?
Would states and companies really stop caring about people, or do democratic and market pressures push the other way?
Are “avert, diffuse, democratize” realistic, and do they conflict with each other (for example, diffusing powerful AI versus preventing catastrophe)?
What would a good social contract look like if human labour matters less?
What can individuals, communities and Portugal actually do about it?
No expertise needed, and reading the essay beforehand is encouraged, though a summary will be given for anyone who hasn’t. All perspectives are welcome, including sceptical ones.
Read the essay: https://intelligence-curse.ai/
Entrance is free of charge and light refreshments will be provided.
At LAISH, we are dedicated to giving Portugal a voice in the development of safe AI that remains aligned and under human control. We have fellowships, courses, events and publications to help make that a reality. Visit our website for all upcoming opportunities and events: www.lisbonaisafetyhub.org
We look forward to seeing you.