By MaRS Staff | July 30, 2026
By the end of 2026, global investment in artificial intelligence is projected to reach $2.5 trillion. Yet, as capital floods into foundation models and consumer applications, a critical question looms: how do we ensure this unprecedented wave of technology addresses society's most pressing challenges rather than generating commercial hype?
For Leah Morris, Executive Director of the Encode: AI for Science fellowship at Pillar VC, the answer lies in bridging the gap between frontier machine learning talent and fundamental scientific research.
Run by Pillar VC in partnership with public research institutions like the UK’s Advanced Research + Invention Agency (ARIA), Encode places top AI researchers directly into scientific labs tackling hard problems in climate science, chemistry, materials discovery, and medicine.
"Much of the world's top AI talent isn't currently focused on fundamental, hard science problems," Morris notes, highlighting that engineers in frontier commercial labs are hungry to build solutions that truly matter. By supporting foundational infrastructure—such as open-source scientific datasets, benchmark competitions, and university-industry collaborations—public-private partnerships can unlock breakthroughs whose value compounds over decades, even when they fall outside traditional venture capital timeline models.
While mainstream headlines often fixate on far-off biosecurity threats, Morris argues that the most urgent societal risk is far more immediate: the accelerating concentration of wealth.
As AI consolidates economic power into the hands of a small number of tech entities, the resulting financial bifurcation fuels public frustration and political volatility. Compounding this challenge is an alarming shift in governance, where major choices regarding surveillance or military technology are increasingly dictated by corporate usage policies rather than democratic oversight or international diplomacy.
This dynamic has elevated the global conversation around sovereign AI. Rather than retreating into digital isolationism or attempt complete technological self-sufficiency, Morris defines true digital sovereignty as freedom from coercion. Nations must ensure democratic access to compute hardware, data infrastructure, and open models so they are not forced into dependency on external monopolies.
As machine learning tools automate administrative tasks, maintaining human judgment remains essential. AI algorithms are not neutral arbiters of absolute truth; rather, their value lies in exposing uncertainty and revealing blind spots to assist human decision-makers.
At the same time, investors and policy leaders must navigate an increasingly noisy startup ecosystem—one Morris candidly describes as "bullshit in a bull ring." As superficial AI wrapper products proliferate, separating genuine scientific advancement from empty pitch decks requires deep domain diligence.
Navigating AI’s multi-trillion-dollar ascent will require far more than faster algorithms. It demands institutional innovation, democratic safeguards, and a concerted effort to direct technology toward the public good.
The podcast navigates complex issues surrounding AI's rapid evolution, underscoring both the transformative potential and the critical need for intersectional considerations. Leah Morris’ insights into responsible AI and the recognition of human imperfections in decision-making frame a nuanced outlook on the future of technology in society.
This website uses cookies to save your preferences, and track popular pages. Cookies ensure we do not require visitors to register, login, or share any identity information.
