Beyond the Fear Factor: Designing an AI Agenda for the Public Good
The conversation surrounding Artificial Intelligence has reached a fever pitch, but it is stuck in a loop of catastrophe. We obsess over the “what ifs”: What if AI leads to mass unemployment? What if it supercharges surveillance or enables the creation of bioweapons? While these concerns are valid, our policy focus has become entirely reactive, centered almost exclusively on preventing harm.
There is a glaring omission in the public discourse: What do we actually want AI to do for us? If we treat AI solely as a threat to be contained, we forfeit the opportunity to use it as a tool to solve the most pressing challenges of our time.
The “Compute” Divide: Who Owns the Future?
AI’s benefits will not emerge through market forces alone. Currently, we are seeing a widening “private-public divide.” Tech giants and massive financial institutions like Goldman Sachs have the capital to hoard “compute”—the raw processing power necessary to train and run frontier-level models—while public universities and government agencies are left on the sidelines.
To bridge this gap, we need a public agenda for AI. This could take the form of a public option for AI: dedicated, government-backed infrastructure that ensures researchers and public institutions have affordable access to the processing power required to tackle societal problems.
Where AI Actually Delivers: Real-World Breakthroughs
When AI is pointed at the right problems, the results are already transformative. We are moving past the hype phase and into an era of tangible utility:
- Scientific Discovery: OpenAI models have successfully disproved long-standing mathematical conjectures, while DeepMind’s Graphcast is currently outperforming traditional supercomputers in weather prediction accuracy.
- Healthcare Revolution: A new drug for pulmonary fibrosis has become the first fully AI-generated treatment to prove efficacy in human trials. Meanwhile, Mayo Clinic teams are using AI to identify pancreatic cancer markers on CT scans years before they become visible to the human eye.
- Material Science: Through a collaboration between Microsoft and the Pacific Northwest National Laboratory, AI analyzed over 32 million materials to identify a new electrolyte for high-capacity lithium-ion batteries.
Building the Infrastructure of the Future
As history shows with the Protein Data Bank—the foundation that made the Nobel Prize-winning AlphaFold possible—AI is only as good as the data it is fed. If we want AI to solve public problems, we must commit to the labor-intensive task of building public-good data sets.
This means cleaning up government data, funding the digitization of public records, and creating a “digital concierge” for citizens. Imagine a government interface that, powered by an LLM, helps you navigate the tax code or apply for public services with the ease of a personal accountant. The technology exists; the political will is the only missing variable.
The Path Forward: Incentivizing Public Solutions
We shouldn’t expect the private sector to prioritize rare diseases or long-term battery storage when there is no guaranteed return on investment. The government must act as a market-shaper. Much like Operation Warp Speed, we can define the outcomes we need—such as a cure for a specific rare disease—and create a guaranteed market for private companies that reach those milestones.
Frequently Asked Questions
- Why is the public skeptical of AI?
- High-profile concerns regarding job displacement, privacy, and corporate power have created a “dismal” polling environment. Many see current AI applications as overhyped and potentially harmful.
- What is the biggest barrier to public AI adoption?
- The primary barrier is the “compute divide.” Without access to the expensive processing power required to run advanced models, public institutions cannot compete with private corporations.
- How can AI help with government services?
- AI could act as a digital concierge, simplifying complex tasks like tax filing or accessing social services, effectively making government more accessible and efficient for the average citizen.
What do you think is the single most important problem AI should solve for the public? Share your thoughts in the comments below, or subscribe to our newsletter for more deep dives into the future of technology, and policy.
