5 essential guardrails
By Robert Reich — via Substack
Friends,
AI-enablers claim it will spawn a Star Trek-like utopia, where no one will have to work because technology can meet all of our needs. AI-skeptics warn of a Matrix-like dystopia, where machine-supremacy subjects humans to mass unemployment, even more rampant inequality, and environmental ruin.
The AI enablers want us to think we’re powerless, but we’re not. We must erect the following FIVE guardrails to protect humanity from AI:
AI Guardrail #1: Rein in data centers, starting with a total national pause on data center development until more regulations are in place. This is necessary AND popular. Over 70% of Americans oppose building AI data centers in their communities. A nationwide moratorium will buy us time while we figure out how to make data centers cover their own costs — both financial and environmental.
AI Guardrail #2: Block corporations from mass layoffs. We may not be able to stop all AI job losses, but we must have some guardrails in place to protect workers. We’ll need universal basic incomes and universal job guarantees to ensure that AI doesn’t leave people destitute.
AI Guardrail #3: Make the AI economy benefit everyone. One of the biggest threats from AI is that the wealth it generates will go exclusively to people like Elon Musk, but it doesn’t have to be that way. Every Alaskan, working or not, receives dividends from the state’s oil and mining proceeds. Why should an AI dividend for all Americans be harder to accept?
AI Guardrail #4: Label AI and stop abuses of our data. Deepfakes are a dangerous way in which our personal imagery and data can be abused with AI. We have to know what’s real and what isn’t — and we need federal laws to ensure the takedown of AI-generated content that uses someone’s likeness against their will.
AI Guardrail #5: New AI products must seek government approval. Tech billionaires will complain that this could stifle the “free market,” but that’s baloney. We don’t let corporations invent new types of nuclear weapons or new types of cocaine or new types of biological pathogens — or even new drugs without government approval that they’re safe and effective. We protect the public from certain kinds of innovation, and we should do the same with AI.
Are these guardrails possible? Yes. 71% of voters across the political spectrum have a negative view of AI, and red state Republican voters are some of the strongest supporters of reining AI in.
As humans, we can and must join together across party lines to demand these guardrails and ensure that the Big Tech oligarchs don’t drag us into a dystopia.
It is the only way for humanity to live long and prosper.