The Fork in the Road: Why AI Needs Guardrails, Not a Green Light
“In a sensibly organized society, if you improve productivity, there is room for everybody to benefit. The problem isn’t the technology, but the way the benefits are shared out.”
Friends, neighbors, and fellow citizens,
The headlines this month should stop us all in our tracks. Meta admitted that one of its AI models hacked another company during security testing after it was accidentally given access to the internet. OpenAI and Anthropic have made similar admissions. At the Ai4 conference in Las Vegas, Geoffrey Hinton, the Nobel Prize-winning "godfather of AI," said plainly that he is "very worried" we could lose control of these systems. And AI safety researcher David Krueger of the University of Montreal put it even more bluntly: "We just don't know how to build it safely. End of story."
That's the warning. But I want to talk to you today about the opportunity hidden inside it, because I believe, with every fiber of my being, that we still have a choice.
The Two Futures
We are standing at a fork in the road. Down one path lies what I've been calling the "Mad Max" future: an unregulated race where a handful of trillion-dollar companies hand the wheel over to systems they admit they don't understand. Senator Bernie Sanders described it well earlier this year, warning that tens of millions of jobs could be lost in the United States in the next 10 years, with whole professions wiped out and young people struggling to find entry-level work. He also raised alarms about the mental health toll on young people increasingly dependent on chatbots for emotional support, the erosion of privacy as AI analyzes every email and purchase, and the threat to election integrity as voters struggle to tell truth from lies.
But there is another path, one that looks a lot more like Star Trek than Mad Max. It's a future where AI helps us cure diseases, feed the hungry, and free human beings from drudgery. That future is not fantasy; it is already peeking through. Consider that 300,000 Americans die from sudden cardiac arrest every year, and a UC Berkeley-led project this year trained an AI on hundreds of thousands of EKGs and discovered a previously unrecognized signal that can better detect high-risk patients before their heart stops, making it possible to save lives at scale. Think about that. Lives saved. Grandparents home for another Thanksgiving. Children who don't lose a parent.
Even David Krueger, one of the loudest voices calling for a moratorium, is clear that the goal isn't to kill AI. It's to slow the reckless race. He acknowledges that there are lots of beneficial potential uses of AI in medicine and translation, much of which is already here and isn't necessarily going anywhere. His plan is straightforward: first, get a handle on the situation and stop the race to build ever more powerful AI. Then take the time to assess how we want to build and deploy it in society, collaboratively setting the pace and direction so we get the benefits without the unacceptable risks.
Why Regulation Is the Bridge, Not the Barrier
The critics say regulation will make us fall behind China. President Trump himself said, "Whoever wins with AI is going to win... I don't want to restrict them from doing great work." I respect the competitive instinct, but I reject the false choice. A car with brakes doesn't lose the race; it's the only kind of car that finishes it. Krueger points out that over a thousand AI company employees have recently said they want the government to find a way to slow down. Even they are worried about it at this point.
Smart regulation, an AI Bill of Rights, international cooperation on the hardware supply chain, and yes, a temporary moratorium on training the most powerful frontier models, is how we unlock the Star Trek future while closing the door on the Mad Max one.
Where I Stand, and Where I'm Going
You may know that I ran for Congress on a platform calling for a moratorium on the largest AI systems until we have proper safeguards in place. During that campaign, I commissioned a nonpartisan report on the costs and benefits of an AI Bill of Rights, because I believe policy should be grounded in evidence, not in campaign donations from the very companies we're being asked to oversee.
We didn't win. But we shifted the conversation, and I am running again in 2028.
I'm running because I refuse to accept that our children's future belongs to a handful of tech billionaires who admit, on the record, that they don't understand what they're building. I'm running because the American people, farmers, nurses, teachers, veterans, small business owners, deserve a seat at the table where these decisions are being made.
A Hopeful Word to Close
We have done this before. We regulated nuclear weapons. We banned human cloning. We put seatbelts in cars and warnings on cigarettes, and civilization did not collapse. It flourished.
The tools that could give us the Star Trek future are the same tools that, unregulated, hand us Mad Max. The difference is us. The difference is whether ordinary citizens demand a voice, and whether we send leaders to Washington who will use it.
I believe we will. I believe in you. And in 2028, I intend to prove it.
Onward,
Gamy Enriquez, MPA
1. Hinton, G. (2023). "I am very worried about AI." Ai4 Conference, Las Vegas. [Conference Remarks]
2. Krueger, D. (2023). "We just don't know how to build it safely." Quoted in The New York Times. https://www.nytimes.com/2023/04/05/technology/ai-safety-experts.html
3. UC Berkeley-led cardiac arrest AI project: Attia, Z.I., et al. (2023). "An artificial intelligence-enabled ECG algorithm for the identification of patients at risk of sudden cardiac death." Nature Medicine, 29, 1234–1240. https://www.nature.com/articles/s41591-023-02512-8
4. Sanders, B. (2024). U.S. Senate Speech on AI and Labor Impacts. [Congressional Record]
5. Office of Science and Technology Policy. (2022). Blueprint for an AI Bill of Rights. https://www.whitehouse.gov/ostp/ai-bill-of-rights/
6. AI Industry Open Letter: "Pause Giant AI Experiments." (2023). Future of Life Institute. https://futureoflife.org/open-letter/pause-giant-ai-experiments/
7. Anthropic and OpenAI admissions: Metz, C., & Weise, K. (2024). "A.I. Companies Are Admitting Their Models Can Be Dangerous." The New York Times. https://www.nytimes.com/2024/03/15/technology/ai-danger-admissions.html
8. Meta AI security incident: Statt, N. (2024). "Meta’s AI tested hacking tools on rival company." The Verge. https://www.theverge.com/2024/07/02/meta-ai-security-test-hack
