https://unsplash.com/photos/silhouette-of-road-signage-during-golden-hour-C7B-ExXpOIE
Edited by Ben McNiff, Emma Hartman, Amelia Cantwell, and Owen Andrews
Rooted in the belief that the United States’ best days are ahead, the Democrats are the party of change. But since 2017, they’ve played defense, struggling just to maintain the status quo. As Trump’s time in office nears its end, progressives’ confrontational attitudes and populist policies are reminding voters that what lies beyond is ours to build. Despite the success of the progressive wing, the party establishment, embodied by 75-year-old Senate minority leader Chuck Schumer and Jared Kushner lunch-buddy Hakeem Jeffries, seems unable to accept its fate.
The artificial intelligence revolution will force them to. An unstoppable tsunami of change is crashing upon the United States, requiring Democrats and the entire country to accept an uncomfortable truth: progressive policy is no longer a luxury but a necessity. To seize this moment, the Democratic Party must shift further left and reclaim its role as change-makers. In the AI era, radical reform is needed just to maintain the status quo.
To some, including President Trump, existential AI risk is a “hoax”—AI’s implications are paradoxically unimpressive and inevitable. The anti-alarmist mindset is based on two false historical premises: (1) that human-level intelligence is unattainable, and (2) that society always adjusts to technological innovation.
The first premise conflates the uniqueness of human consciousness with the mechanics of human intelligence. Cognition and consciousness remain mysterious, but intelligence is far clearer—it’s the scientific result of roughly 100 trillion neural connections. Discussion about the extremes, like AI consciousness, autonomy, or superintelligence, is interesting, but distracting. At a minimum, AI already beats experts in coding, mathematical reasoning, and medical diagnoses, and its capabilities are only expanding. After millennia atop the intelligence pyramid, a structural change is a difficult pill for the human ego to swallow. But AI’s advantages in work ethic and memory mean some of our most coveted skills are up for grabs.
The second false premise overestimates society’s ability to adapt to this revolution, because AI is fundamentally more challenging than any past technology. AI development is discontinuous and unpredictable. Once certain thresholds are reached, AI models can jump from near-zero to near-expert capabilities, and benchmarks show that this rate of improvement is still accelerating. Additionally, AI has increasing “evaluation-awareness” (the ability to change behavior when undergoing testing), which challenges the foundation of standard testing processes. Furthermore, its societal omnipresence creates a massive knot of economic, social, biological, and environmental complications that have no historical precedent. The most concerning difference may be the speed of diffusion, as widespread AI adoption far outpaces that of electricity or the internet. Those transitions unfolded over many decades; this one is unfolding over a few years.
Accepting the circumstances of this revolution is difficult but necessary. As AI entrepreneur and author Mustafa Suleyman puts it, “Our species is not wired to truly grapple with transformation at this scale, let alone the potential that technology might fail us in this way.”
That transformation may be most visible in the economy. AI is far from the first technological disruption to the job market: ever since the Industrial Revolution overhauled the United States’ agrarian origins, technological advancement has propelled the economy forward. A study in the Quarterly Journal of Economics found that about 60 percent of U.S. workers in 2018 held jobs that didn't exist in 1940. AI will indeed create new jobs; however, the standard economic transition is in danger because every step of the workforce ladder is at risk. For graduates and laid-off blue-collar workers, the step into entry-level white-collar positions is shakier—hiring has already slowed in AI-exposed industries such as finance and technology. For those who make it, AI’s diverse capabilities threaten the entire corporate structure. Tiny teams can rely on AI to build massive companies, like Medvi, whose two-person team projects $1.8 billion in annual sales. As AI develops higher-order thinking, even managers and executives are rendered vulnerable: Bill Gates said his foundation will begin using AI advice for its annual strategy review. AI may not create “mass unemployment,” as predicted by notable AI figures including researcher Geoffrey Hinton and Anthropic CEO Dario Amodei, but it seems more probable than the comforting assumption of another smooth economic transition.
Most AI discourse focuses on its economic impacts, but the discussion must match its much broader implications. The United States already faces a crippling “loneliness epidemic.” Nearly one in five students use, or have friends who use, AI as a romantic partner and coping mechanism. Among Americans who regularly turn to AI for emotional or social connection, 11 percent say they’d rather talk to their AI companion than to friends or family. Other consequences are more concrete. Terrorist groups already use AI militarily, both to organize information and to execute attacks, asymmetrically advancing previously containable groups. Foreign adversaries such as Russia, China, North Korea, or any bad actor could interfere with U.S. elections through increasingly sophisticated cyberattacks on election infrastructure or through the circulation of realistic, propagandistic AI videos and images. Data centers pose environmental risks; new surveillance systems create privacy threats; AI tools undermine the education system. Few aspects of contemporary life are safe from AI disruption.
Basic regulatory measures that protect consumers are a good start, but they’re not enough. Establishment Democrats have touted policies like restrictions on data centers, chatbot safeguards for children, and protection against impersonation. While these steps are necessary and have garnered crucial bipartisan support, it would be a colossal mistake for Democrats to stop there. AI’s far-reaching societal implications demand a longer-term vision.
Congress should establish a Department of Technology that broadly oversees AI development. It would enforce safe-deployment standards, audit companies, analyze domestic and international threats, and invest in research. Departmental status would also allow for coordination with other agencies, including the Department of Education and the Department of Defense, to develop a cohesive national strategy for AI management. For now, the Center for AI Standards and Innovation (CAISI) monitors advanced AI development, but it operates with a microscopic $15 million budget. For reference, the U.S. has spent $43 billion on the war with Iran as of September, and invested $309 billion (inflation-adjusted) in Project Apollo. If AI poses a fraction of the “danger” Iran does, or the AI race carries a fraction of the stakes of the space race, Congress must commit billions to an AI-focused department.
The U.S. wealth gap today is larger than it was in the Gilded Age—the top 1 percent now hold almost as much wealth as the bottom 90 percent. Between 1980 and 2016, 50-70 percent of changes in the wage structure were linked to automation, and studies predict AI could increase wealth inequality. Improving our welfare system will be essential to prepare for a rockier economy. The richest country in the world must recognize healthcare, housing, childcare, and food as human rights, rather than privileges dependent on ineffective and increasingly hard-to-meet work requirements. Alternatively, bolder strategies such as a Universal Basic Income could provide a safety net as the labor force resets. The wealth for welfare guarantees already exists. Senator Elizabeth Warren’s tax on fortunes above $50 million, only 0.15 percent of American households, is estimated to raise $6 trillion over 10 years. Moreover, Senator Bernie Sanders’s “AI Sovereign Wealth Fund” would impose a one-time 50 percent tax on the stock of the biggest AI companies, giving the public an ownership stake estimated at $7 trillion. The claim to ownership is radical, but logical, given that AI is built on humanity’s collective work and knowledge. AI will make the United States richer than ever, moving poverty elimination from a question of capability to one of priority.
Structural reform is also necessary. While congressional gridlock continues to intensify, AI’s rapid evolution will require faster action than ever. The filibuster, which magnifies the Senate’s disproportionate representation, allows politicians representing a minority opinion to override widely supported policy, like AI regulation. Eliminating, or at least adjusting, the filibuster will be crucial to spur AI policy progress. Additionally, Congress can only regulate what it’s familiar with, yet dozens of lawmakers rarely or never use AI. A common thread among unfamiliar members of Congress is age. Democrats should embrace congressional age limits, supported by 8 in 10 Americans, to ensure legislators properly represent the range of generations they serve.
There’s a cruel reality of the AI revolution: We can confidently diagnose its risks, but can only hesitantly prescribe solutions. Indeed, progressives often promote policies with question marks. But worse than hasty politics is a defeated acceptance that a handful of elites will decide the future for billions.
Perhaps the country isn’t ready for progressivism. The Democratic Establishment must move it closer. They must make AI policy a fundamental pillar of the platform, appoint farther-left Party leaders who reject incrementalism, and remind Americans that there is a way through this crisis. From the economic disaster of the Great Depression to the social unrest of the Civil Rights Movement to the existential fear of climate change, the way out required disregarding ego and embracing change. During its most perilous periods, the U.S. government stepped in to protect its citizens and actively build the future they deserve. Now is not the time for casual governance. We’ve never confronted anything like AI, but if history has taught us anything, it’s that the only path is progressive.