Did we just hit the first AI “speed limit”?
In 1896, the world’s first speeding ticket was issued to Walter Arnold for driving at the “breakneck” speed of 8 mph.
From then until the Formula 1 of today, think of how many technological improvements cars have undergone.
For decades, it probably seemed like cars would just keep getting faster and faster as technology improved.
And yet, here we are 130 years later, and we can’t drive faster than 65 in most places.
Claude just released Fable (which was awesome), and our government immediately yanked it off the market. It’s still not back.
Does this mean we’ve gotten to the first speed limit set for AI? The point after which no future AI models can be used by ordinary people—at any price—even though Formula 1 models will continue to get better behind closed doors?
I’m sorry world for bringing the concept of an AI speed limit into existence, but you can expect my metaphor to be picked up by governments and used—I guess I’m channeling my inner Frank Luntz.
The speed limit is a powerful metaphor because it signals that there is already precedent for arbitrarily limiting technological advancement in the name of the greater good.
If this is true, and we’re setting the AI speed limit today, then for the average Joe the arms race of AI is already over. Because it’s not like any future AI models will be less powerful than the ones already banned.
Or maybe we’re going 8 mph, and this will all seem quite silly in a few years.




