Think About What AI Future You Want 🤔

Effort Level: ●○○○○ (Less than 1 hour)
Areas: Self-Education

Summary

In order to make sure that the AI future goes the way you want it to, you first have to understand what you want! It’s worth spending 30-60 min writing down a rough draft of your thoughts on this. We’ve found that often simply the act of putting words to paper is enough to clarify a lot of your thoughts.

The exact format of what to do here can vary from person to person. We would recommend at least thinking through the following 5 questions:

  1. What excites you most personally about AI’s potential?
  2. What worries you most personally about AI’s risks?
  3. What role should humans play in an AI-enabled world?
  4. What is the worst/best case scenario for an AI future? Is scarcity solved? Does everyone get wiped out by AI?
  5. What are warning signs that AI is on the path to becoming superhuman (but is not there yet?)? What are warning signs that AI is on the path to fizzling out?

If you want more questions to think about see the following:

Personal Impact Questions

  • How would you want AI to change your daily routine in 10 years? What would stay the same?
  • What’s something you currently struggle with that AI could help solve, and how would that feel?
  • What human activities or experiences do you hope will never be automated or replaced?

Values and Relationships

  • How should AI systems treat human dignity and autonomy? What does that look like in practice?
  • What role should AI play in how we connect with other people? Should it bring us together or stay out of the way?
  • How do we ensure AI amplifies the best of human nature rather than our worst impulses?

Work and Purpose

  • If AI could do many jobs, what would give your life meaning and purpose?
  • What kinds of work should remain fundamentally human, and why?
  • How might we restructure society if human labor becomes less economically necessary?

Decision-Making and Control

  • When should humans be in the loop for important decisions, and when is it okay for AI to decide?
  • How much should you understand about how an AI system works before trusting it with something important?
  • What would ideal human-AI collaboration look like in areas you care about?

Societal Structure

  • How should the benefits of powerful AI be distributed across society?
  • What institutions or safeguards would you want in place as AI becomes more capable?
  • How do we preserve human agency and choice in a world with very smart AI systems?

Failure and Risks

  • What’s your biggest worry about AI development, and how would your ideal future address it?
  • How should society handle AI systems that make mistakes or cause harm?
  • What would responsible AI development look like to you?

Cognitive Strangeness

  • What if an AI could think a million thoughts in the time it takes you to blink? How would you even communicate with something like that?
  • Imagine an AI that understands your personality better than you do and can predict your choices before you make them. What happens to your sense of free will?

AI Systems Themselves

  • If an AI claimed to be suffering or to love someone, how would you know if it was telling the truth? Does it matter?
  • Should an AI system have rights itself?

Control and Agency:

  • What if an AI systems could manipulate human behavior but chose not to? How would you know the difference?
  • Imagine an AI system that was as persuasive as the best PR organizations today. What happens to truth and democracy?

Next Steps

After writing this, feel free to check out other steps that you can do, including: