Stop Waiting for Tomorrow's AI: The Untapped Potential of Current Technology
While policymakers and researchers fixate on future AI developments, we're missing unprecedented opportunities to solve today's problems with existing technology.
Yesterday, UC Berkeley announced an incredible breakthrough. In conjunction with researchers from UCSF, Stanford, and NVIDIA, they developed an AI model capable of identifying patterns in gene sequences in different organisms. Humans alone would have needed years to accomplish that feat. What's more, the model can spot disease-causing mutations in human genes.
Admittedly, I cannot speculate about the domain-specific importance of this instance of AI progress. However, I do know two things. First, the researchers themselves think it's a big deal and that’s not nothing given their prestigious institutional ties. Second, this development will likely go uncovered by major news outlets and even escape the attention of those who pay close attention to AI. The reason is an excessive focus on how AI may develop rather than how it may be used today to solve important problems.
If AI development ground to a halt today it would still take society years, if not decades, to fully realize the benefits of the technical progress that’s been made since the introduction of ChatGPT-3.5 in 2022. The importance of this work—making the most of the AI that’s already available here and now—has been largely ignored by policymakers and civil society.
Legislation at the state and federal level has focused on the frontier of AI development—attempting to reduce the odds of the most sophisticated models causing harm or ending up in the hands of our adversaries. New York state legislators, for example, are debating legislation that would impose numerous procedural checks on AI labs racing to increase AI capabilities. On the Hill, bills such as the Decoupling America’s Artificial Intelligence Capabilities from China Act of 2025 have garnered attention from politicians and the press.
Academics, nonprofit leaders, and researchers likewise seem distracted by what AI may look like years from now rather than what societal advances are possible with today’s systems. Conference agendas are full of panels debating the AI race with China and the harms that future models may pose.
While some legislative and scholarly attention to the next wave of AI innovation and its geopolitical ramifications is surely warranted, the current ratio between attention to future AI and present AI is off. It’s time to allocate more resources and expertise to existing “everyday” AI—systems that teachers, doctors, lawyers, and the rest of us could use right now to better serve our communities.
Excessive focus on future AI systems (and their risks) has driven many Americans to assume the worst when it comes to AI—rendering them skeptical of even straightforward applications of AI. Polls have long suggested that most Americans have a negative view of this incredible technological advance. As of 2023, the majority of Americans expressed more concern than excitement over the use of AI in daily life. By 2024, the public slightly adjusted its views—a majority regarded AI as producing as much good as bad. A 2025 survey, however, may suggest that public views have since soured. Respondents expressed particularly grave concerns that AI would exacerbate the spread of misinformation.
Exposure to more information on how current AI systems are already improving daily life would presumably lead to a different perspective. Broader dissemination of stories like the Berkeley breakthrough described above could help the public see that we’re currently underutilizing AI systems. In turn, the public would likely have a different perspective on when and how we should lean into AI. Imagine, for example, if more Americans knew the following:
Researchers at Harvard Medical School develop an AI model that accurately detects certain cancers 94 percent of the time. They cite this as evidence that AI-based tools markedly improve the ability of clinicians to spot and treat cancer.
The vast majority of small business owners regard AI as a cost-effective tool that increases their profits as well as their ability to compete with larger firms.
Students equipped with the AI educational tools developed by Quill.org have shown significant improvement in their writing skills even over the course of a single semester. That progress would not have occurred absent AI assistance. As Quill boasts, “2,203 students [in a school district in Kentucky] completed over 120,000 sentences on Quill in the first month alone. That’s equivalent to receiving teacher-written personalized feedback on almost 5,000 essays!”
Greater investment in these efforts would drastically improve the quality of life for millions of Americans. It’s these use cases that merit far more popular attention and political support.
Delayed expansion of these use cases is unacceptable, even when factoring in the known limitations of many AI systems such as persistent problems with hallucinations. There are too many opportunities for AI to aid Americans to tolerate relatively low rates of adoption by the general public as well as by professions tasked with performing tasks essential to the broader community. Future generations will not look kindly upon us if we fail to lean into AI. For sake of comparison, imagine if we learned that scientists sat for years on a new cancer treatment that was many times more effective than existing protocols.
My own take is that we’re already imperiling our future reputation by failing to take advantage of existing AI systems. Our grandchildren will struggle to comprehend why it took so long for us to invest in AI literacy so that everyone knows its risks and harms in addition to how they can use it in their daily lives. They will likely also question why the American Bar Association and related groups did not rally behind AI and undergo substantial institutional reforms to put the technology to its best uses.
As a lawyer and legal scholar, the failure to embrace AI as a means of improving our legal systems and laws has gone on for too long. Yet, the moment of reckoning has not arrived. By way of example, just 11 percent of lawyers in Florida regularly use tools like ChatGPT. AI is something happening to lawyers rather than something lawyers are harnessing in fulfillment of their pivotal societal role. I can’t speak for other professions but my hunch is that their members similarly feel a lack of agency when it comes to learning about and employing AI.
This is not to say that current AI systems are not without risks, biases, and other shortcomings that justify some caution. Bad actors have abused such systems to accelerate their fraudulent activities, for instance. But a defensive and overly pessimistic view of AI will only result in those bad actors having a greater edge over the innocent public. Likewise, our allies and adversaries will not wait for us to catch up when it comes to AI adoption. Americans need to act now before kids around the world assume the mantle of the next AI innovators. We also don’t want to find our workforce operating in the modern equivalent of the Stone Age while international peers receive the training necessary to thrive in the Age of AI.
Debating when and if we will reach artificial general intelligence (AGI)—systems that perform most tasks as well as or better than humans—is a fun and important thought exercise. Speculating about China’s AI capacity also makes for good conference fodder, press releases, and congressional inquiry. The comparatively boring work of integrating AI into our daily tasks, however, is arguably the most important thing to do if America is going to achieve AI dominance.
* * *
The gap between AI's current capabilities and its everyday implementation represents more than a missed opportunity—it's a failure to fulfill our societal obligations. While discussions about AGI and international AI competition capture headlines, the true measure of America's AI success will be determined by how effectively we deploy today's technology to solve real-world problems. Teachers struggling with overcrowded classrooms, doctors facing mounting caseloads, and legal aid attorneys turning away clients due to limited resources already have powerful AI tools at their disposal.
What's missing is the institutional support, training, and public understanding needed to implement these solutions at scale. Rather than allowing fears about future AI to paralyze progress, we must focus on responsible adoption of existing systems. The path to AI leadership doesn't solely run through speculation about tomorrow's breakthroughs—it runs through the classrooms, clinics, and courtrooms where AI can make a difference today.
Thanks for the article, Kevin. Yes, I think many of us - most notably Boomers - have a jaded eye and apprehension of future uses of AI. Your examples help allay much of that concern.