The Future is AI: Exploring the Possibilities and Pitfalls

Session Overview

From writing novels to designing new molecules, artificial intelligence is transforming nearly every sector and fueling a new era of innovation. This panel led a thought provoking discussion on the challenges, opportunities, and potential risks of AI's future impact on New Mexico and the United States.

Key Session Insights

More than a tool, AI is an inflection point in human history — one that is reshaping governance, democracy, labor markets, and humanity itself. The technology’s potential to streamline operations and personalize services also has the power to displace jobs, magnify bias, and destabilize public trust. Dr. Will Tracy, Vice President for Applied Complexity at the Santa Fe Institute, compared AI’s emergence to that of the printing press: a disruptive force whose long-term effects may take decades to fully understand but whose immediate consequences are already apparent. There is an urgent need, he argued, to bring ethical frameworks, civic norms, and regulatory mechanisms into alignment with the speed of technological development. Without these safeguards, society risks a future in which the benefits of AI are enjoyed by the few while the burdens fall on the many.

Bridging the gap between elite research environments and the broader public becomes essential if AI is to serve as a truly transformative force across society. While the U.S. DOE National Laboratories like Los Alamos and Sandia are actively investing in AI to accelerate research and scientific discovery, technology must be democratized beyond elite institutions. Dr. Jason Pruet, Senior Director of the AI Office at Los Alamos National Laboratory, noted that for AI to fulfill its potential responsibly, its use in advanced research must be accompanied by efforts to build public understanding and ensure equitable access across New Mexico.Without such engagement, the benefits of these powerful tools may be unequally distributed or misunderstood.

This need for inclusion is amplified by the accelerating pace of AI development, which is quickly reshaping the demands placed on institutions, workers, and policymakers. In both the public and private sectors, there is growing concern that the pace of AI innovation is outstripping the ability of workers, educators, and even government agencies to keep up. Dr. Kevin Dixon, Program Director of Sandia National Laboratories, remarked that although the “pie” of opportunity may grow larger, its slices will not be distributed equally unless deliberate efforts are made to build capacity across all sectors of society. As AI tools become more powerful and ubiquitous, it becomes critical to upskill workers not just in how to use these tools, but in how to interpret, question, and adapt them to their specific contexts. Failing to do so could lead to widened inequality, eroded trust in institutions, and technological disillusionment.

AI is fundamentally altering the landscape of education, reshaping how students learn, how teachers instruct, and what skills are essential in the modern world. In New Mexico, educators and technologists are piloting programs that embed AI tools into K–12 and higher education. Ms. Meg Fisher, Co-Founder of Santa Fe AI Partners, highlighted the work her organization is doing to teach students in middle and high school to use AI for image recognition and machine learning. This curriculum is buttressed by challenge-based learning, which encourages students to solve real-world problems using large language models (LLMs) and machine learning tools. Similarly, programs like the co-pilot training at the University of New Mexico (UNM) are helping faculty and staff integrate AI tools into curricula, while also providing students with exposure to these technologies early on. While these programs show promise, educators are grappling with the challenge of maintaining academic integrity in an environment where students have access to advanced AI tools that can assist in completing assignments. Despite these challenges, early exposure to AI fosters critical thinking, reduces fear of the technology, and equips students with the skills necessary to thrive in an AI-driven world.

As education systems work to prepare the next generation for an AI-driven world, similar challenges emerge in the broader workforce, where the rapid pace of AI’s development is reshaping not only how we learn, but how we work and innovate in scientific fields. Artificial intelligence is no longer simply a tool used to accelerate tasks — it is becoming a true collaborator in the research process, capable of transforming how knowledge is generated. Los Alamos National Laboratory Division Leader Dr. Aric Hagberg described how researchers have gone from using AI to merely make things go faster, to using AI to build things faster. This transition marks a deeper integration of AI into scientific inquiry, where models are not just computational aids but reasoning partners in exploration and design.

However, as Dr. Dixon pointed out, despite AI’s growing potential, institutions like Los Alamos are not fully leveraging its capabilities. Many research teams continue to underutilize AI technology, struggling to implement it in their workflows in a way that maximizes its transformative potential. This problem raises a critical question: How can we improve AI adoption within research institutions? Increasing AI usage requires not only better access to technology but also fostering a culture where researchers embrace AI as an essential part of their process, rather than just a tool to speed up tasks.

To build this culture of adoption, individuals must understand the foundational principles of AI. By training students to understand the strengths and weaknesses of these models, such as the bias in models, they can more effectively apply AI in ways that align with its capabilities. Ms. Geetika Gupta, Director of Product Management for NVIDIA highlighted that when students grasp how AI models are built, they are better equipped to discern what tasks these models excel at, and more importantly, what tasks they are not suited for. This knowledge reduces fear of using AI and helps people approach these tools more thoughtfully. Teaching these skills will play a significant role in improving AI adoption, ensuring that the next generation not only uses AI but understands it deeply — empowering them to use it responsibly and effectively in their own work.

These challenges of workforce readiness and public inclusion are compounded by the sheer magnitude of change AI is driving within scientific fields themselves. Dr. Dixon noted the scientific impact of AI may be even more dramatic than currently realized, potentially compressing workflows by a factor of ten. Yet this acceleration exposes a critical vulnerability: most institutions are not yet structured to absorb such rapid transformation. Dr. Dixon is concerned that the United States will not revolutionize its institutions to meet this challenge until it is too late. For example, what if foreign adversaries use AI to make America’s missile defense system obsolete? 

Bridging the gap between technological capability and institutional readiness requires a coordinated, cross-sector approach. Dr. Tracy outlined the importance of building an AI innovation ecosystem in New Mexico that leverages its strengths — world-class U.S. DOE National Laboratories, top-tier research universities, vibrant cultural institutions, and entrepreneurial nonprofits. Rather than operating in silos, he proposed that these organizations must see themselves as co-architects of a shared experiment in AI governance and implementation. AI holds immense promise, but establishing collaborative norms and standards is necessary. He called for all institutions to think of themselves as part of a larger, interconnected experiment. China has already embraced this cooperative approach, and Dr. Tracy warned that the United States risks falling behind if it does not follow suit.

In this line of thinking, Ms. Gupta described how public-private collaboration, particularly with the U.S. Department of Energy national laboratory enterprise, could dramatically improve the innovative potential of AI. Currently, commercial companies are training their generative AI models with publicly available data. However, these models have the potential to be more powerful if they were trained on the data and simulations generated by the national laboratories.

Advancing AI in a responsible and transformative way also requires rethinking who gets to shape these systems and how they are understood. Dr. Moses argued that the United States can develop competitive AI systems not just through greater computational power, but through more thoughtful and interdisciplinary approaches to model development. Rather than relying solely on computer scientists, Dr. Moses called for broader engagement across disciplines, including historians of science, who can contextualize AI’s development; psychologists, who can explore how AI mimics or diverges from human cognition; and biologists, who can help society grapple with AI as a novel form of synthetic “life.” She noted that this shift toward interdisciplinary thinking is already underway, and expressed optimism that a more integrated, nuanced approach to AI is becoming the norm. Without it, she warned, we risk misunderstanding the capabilities and consequences of the systems we are building.

Despite ongoing challenges, the panelists expressed a clear belief in AI’s potential to uplift communities and support innovation at the grassroots level. In New Mexico, rural residents, tribal nations, and small-town entrepreneurs are emerging as vital contributors in exploring how AI can advance the public good. Applications such as precision agriculture, rural healthcare, and localized educational tools point to how AI could drive economic renewal and address persistent challenges. Both Dr. Hagberg and Ms. Gupta expressed an optimistic vision these tools will inspire a new wave of entrepreneurs who harness AI to solve problems in their own community. Similarly, Ms. Meg Fisher pointed to the breakthroughs underway in healthcare, where computing power is already being used to develop more personalized and impactful treatments — advances that could benefit patients regardless of geography.

This vision also calls into question the assumption that only those with massive computational infrastructure can meaningfully contribute to AI advancement. Ms. Gupta pointed to the DeepSeek AI model as evidence that progress does not always require training models from scratch. Instead, fine-tuning existing models through post-training and reinforcement learning — particularly with input from domain specialists — can unlock major gains. She illustrated this with a metaphor: a general-purpose athlete becomes truly effective only after a coach prepares them for a specific sport. This approach supports greater access and could help drive AI innovation in communities with limited resources.

New Mexico stands at the intersection of innovation and inclusion — home to some of the nation’s most advanced research institutions, yet marked by persistent disparities in education, infrastructure, and opportunity. This paradox makes it an ideal proving ground for national AI policy. The path forward must balance speed with stewardship, innovation with equity, and ambition with accountability. If New Mexico can align its unique assets — U.S. DOE National Laboratories, research universities, diverse communities, and civic institutions — it can lead the way in shaping an AI future that serves all.