The California Legislature is right in that the civilizational implications posed by AI technology necessitate a public policy response. But the response has been a cacophony of onerous regulations, wokeism, and a fundamental misunderstanding of emerging technologies.
Take California’s Senate Bill 1047, which recently passed out of the Senate. It is rife with the kind of suffocating regulations you would expect out of California. For example, the bill requires developers of AI models to demonstrate that there is no possibility that their covered model has a “hazardous capability” before they can initiate training. This is an impossibly high standard for major incumbents and a death knell for smaller competitors.
What’s more, the bill codifies the state’s vision of a cloud platform owned and hosted by the state, CalCompute, which would aid in “fostering equitable innovation” and assist in operationalizing the state’s $67 billion effort to embed “DEIA [diversity, equity, inclusion, and accessibility] in state operations.” This effectively enshrines the Google Gemini approach of only producing AI-generated images of black and Native American Founding Fathers.
That California is spending billions on a woke operating system might be comical if the race for AI legislation in the U.S. weren’t being driven by the states.
Considering the most recent substantial national tech policy laws are from the 1990s (e.g., the Communications Decency Act of 1996 and the Children’s Online Privacy Protection Act of 1998), it is no exaggeration to say the battle for tech policy is coming down to the states. While President
