The Great AI Alignment Love Match

The 💘 Great AI Alignment Love Match 💘 is a five-minute quiz to decide if you are more of a Safetyist, Accelerationist, Skeptic, or Normalist.

The point of this quiz — other than as an homage to the online quizzes of the early 2000s — is to explore Jonathan Zittrain’s “AI Triad” of Safetyists, Accelerationists, and Skeptics.

Those three camps mostly don’t describe actual people — most of us are much more conflicted — but instead describe bodies of research that fail to engage with each other.

The idea is that the research groups answering three questions …

• What to do about the parts of AI that are transformatively good? (Accelerationists)
• What to do about the parts of AI that are transformatively bad? (Safetyists)
• What to do about the parts of AI that are bad in normal economic and social ways? (Skeptics)

… are systematically disincentivized to talk to each other, and end up using incompatible language and missing each other’s key results — a shame when all three camps are trying to figure out how to make a vast technological shift better for humanity.

So, while this quiz cheerfully tries to plop you in one of the camps, I hope it actually offers some windows into different ways to look at the problem.

As a fun bonus, we can also have AI chatbots take the quiz themselves. Here’s where ChatGPT landed the first time I tried that …

and here’s Claude: