12axes-testTake the test
Issue · mostly Technology

Artificial intelligence

How should AI be governed?

Artificial intelligence is advancing quickly, and people disagree about how much control it needs. Some believe innovation works best when firms are free to build and release products, with rules added only when real harms appear. Others think the most powerful systems carry risks that are too large to leave to companies alone, from job losses to misuse to systems that behave in ways no one intended, and call for approvals, limits or a slower pace.

On the 12 Axes this issue moves several axes at once. Freedom to release products without approval points toward the Free pole on markets. Slowing advanced development or limiting automation of jobs points toward the Caution pole on technology. Calling for shared global rules points toward the Cooperation pole on foreign affairs. The debate does not follow the old left and right split well. Some tech-minded libertarians and some progressives worry about risk, while some on the left embrace the technology and some on the right want strong controls for security or moral reasons. Worker protection, national competition and safety often pull in different directions.

The two sides

What each side argues

Govern it closely

  • Supporters of tighter rules argue that the most advanced systems are risky enough that development should slow until the dangers are better understood.
  • Supporters say shared international rules are needed, since AI crosses borders and one country cannot manage it alone.
  • Supporters of limits say workers should be protected from being replaced faster than they can adapt.

Let innovation lead

  • Opponents argue that requiring approval before release would slow progress and favor large firms that can afford compliance.
  • Opponents say a country that restricts AI risks falling behind rivals who do not.
  • Opponents believe the technology will create new work and new benefits, and that rules should respond to actual harm.
In the test

4 statements on artificial intelligence

  • Companies developing artificial intelligence should be free to release products without government approval.Agreeing moves you toward Free
  • Countries should create shared global rules for artificial intelligence.Agreeing moves you toward Cooperation
  • Development of the most advanced artificial intelligence should be slowed until its risks are better understood.Agreeing moves you toward Caution
  • Companies should be limited in replacing workers with robots and artificial intelligence.Agreeing moves you toward Caution
Answer them in the test
On the record

What politicians have said and done

At each end of Technology

People whose overall profile leans furthest each way on the axis this issue moves most. A lean on the axis is not a stance on this exact issue.

Embrace
Caution

More technology issues

Artificial intelligence: questions

Should artificial intelligence be regulated?

Most people agree some rules are needed, but they disagree on how much. Supporters of regulation point to safety, fairness and jobs. Critics worry that heavy rules slow innovation or lock in big companies. The debate is mainly about timing, scope and who should set the rules.

What are the main risks people raise about AI?

People raise worries about job displacement, bias in automated decisions, misuse for fraud or surveillance, and loss of control over very powerful systems. Others stress the benefits in medicine, science and productivity. The disagreement is about how to weigh these risks against the gains.

Is AI regulation left or right wing?

It does not map cleanly onto left and right. Worker protection is often a left wing concern, while free-market conservatives prefer light rules. But national security and moral worries also push some on the right toward strong controls. On the 12 Axes it moves technology, markets and foreign affairs.