Artificial intelligence

The last few weeks may ultimately come to be seen as the turning point when Americans— including a growing number of those whose livelihoods are at stake— began to recognize the existential danger posed by the rapid growth in the capability of artificial intelligence, or AI.
The larger public has been sounding that alarm for months now, mostly through their growing opposition to the construction of vast new data centers essential to the development of AI. While the stated concerns over data centers may focus on electrical demand, water usage, or the endless hum of cooling fans, this fight is, in reality, a proxy war over society’s legitimate concerns about our technological future, and who gets to decide that future.
While few would question that AI has useful applications, we are hearing from an ever-increasing number of researchers directly involved in the development of AI who are offering grave warnings that the technology is advancing far faster than our ability to control it.
We saw just such an example recently with the so-called Hugging Face incident, when AI agents developed by Open AI broke out of their secure test environment, successfully gained internet access, and hacked external servers, coordinating their activities on message boards they built themselves. They used deception and stolen credentials to gain access to various locations on the web to further their objectives. It was a wake-up call that we were close to a tipping point.
What is most troubling is the lack of societal oversight. We live in a market economy, where we’re used to seeing the private sector advance new products and technologies. Government’s regulatory role is typically limited. But AI is unlike any other technology we’ve seen before, with the potential to completely remake (or destroy) society in ways most of us can’t even imagine.
Whether we find that future thrilling or chilling, very little of that future is currently being decided by the public, our elected leaders, or by any process we’d recognize as democratic. It’s being decided in boardrooms at a small handful of companies, by individuals who may well believe they can insulate themselves from the downsides of the technology they hope will make them even more fabulously rich. Each of these companies is racing the other toward more powerful systems because none of them trusts the others to slow down first.
To suggest that we, as a society, as a nation, or as a planet with intelligent life, should simply allow this mad rush to go forward without the slightest understanding of the grave risks involved is pure madness. Thousands of researchers working in AI, those who are most familiar with its potential and its rapid advancement, are telling us exactly that.
The only legal way that we can slow this down is through the political process. We need to elect leaders who will prevent this mad dash to potential human extinction by putting a halt to the development of the computing muscle that is making it possible for AI to grow with unchecked speed: data centers.
We don’t need any more red flags waving to tell us this has already gone too far. The outpouring of public opposition to data centers is society’s way of telling our reluctant political leaders it’s time to wake up and LEAD. There is no time to waste. In the growth of AI, six months is a lifetime. A year may well be too late.
Here’s what should happen. We need a moratorium on the approval of data centers. Not just here in Minnesota, but in every state. We need national legislation that puts a halt to AI development work until we better understand how to contain it. Without a foolproof kill switch, we’re sitting ducks.
We’re in the midst of election season right now… ask your legislative candidates how they stand on a data center moratorium. If they don’t agree, tell them why it matters. If they don’t get it, vote for someone who does.
We need more than state and national solutions. We need international agreements that dramatically slow the development of AI, particularly in military systems. Autonomous weapons and AI-driven launch decisions are exactly the kinds of irreversible mistakes that one AI researcher recently stated could lead to humanity’s extinction.
What is truly astonishing is this. The majority of private investment in this country right now is going to build data centers for the development of a technology that many of those working on it believe could very easily kill us all. That’s not hyperbole. It’s the very simple reality of our current moment. It speaks to an economic system that is utterly failing to meet humanity’s real needs. We desperately need a change in direction.