By: Stella Rehbein
March 16, 2026
The New York Times Debate at the World Economic Forum 2026, curated by DEBATABLE, explored the AI moment we are in, and the dominant narratives about the pace of AI advancement. This piece covers the New York Times Debate and the key outcomes that followed from it.
The sudden acceleration of AI's capability has destabilised our collective assumptions about the future, forcing leaders to forecast something not just uncertain, but close to unknowable. Entire sectors, national strategies and geopolitical alliances are reorganising around the projection that AI will continue to innovate towards a point of unfathomable excellence. The ambition of individual leaders might be economic advantage, but the collective hope is that AI will clarify pathways beyond human capacity. Where humans have failed, the thinking goes, AI will deliver.
But is this utopian framing simply rhetoric, driven by a mix of human ambition and fear of being left behind? With bottlenecks in energy supply threatening further innovation, and AI's own energy demands working against decarbonisation goals, are we risking too much of our future on AI, and using it as an excuse to avoid our own responsibility? Or is this precisely the moment to treat AI as a genuine tool for advancing climate-focused solutions?
The team arguing for the motion comprised Varun Sivaram, founder and CEO of Emerald AI, Bolor-Erdene Battsengel, founder and CEO of AI Academy Asia, Max Klymenko, content creator, and Adam Grant, the Saul P. Steinberg Professor of Management and Psychology at Wharton School of Management.
Varun argued that AI will succeed by becoming far more capable than humans at solving global challenges such as poverty and climate change, in part by "bringing clean energy onto the grid and power cheaper", creating incentives for countries to adopt it. He cited DeepMind's discovery of 380,000 stable compounds with potential uses in solar storage and carbon capture.
Bolor drew on her experience at the World Bank, where she witnessed emergency relief, medical supplies, food and winter coats delayed by three months due to bureaucracy. "Bureaucracy doesn't just slow us down," she said. "It contributes to killing." She argued that AI can help humans "be efficient, be fast, be inclusive and be accessible."
Max approached the question from his experience as a content creator, identifying humanity's core problem differently from Bolor: not inefficiency, but ego, which he believes AI has the power to transcend.
Adam drew on his background as a psychologist to argue that AI can empathise with and persuade people more effectively than humans do. He noted that after a ten-minute conversation with AI, a quarter of people presenting conspiracy theories abandon those beliefs. He also pointed to AI's role in creativity, as well as concrete examples of AI's impact, such as matching patients with lifesaving drugs.
Each debater on the 'For' team identified a different kind of human failure, from practical problems like poverty and climate change, to emotional shortfalls in empathy, to what Max called the spiritual failure of ego. Despite these differences, all four concluded that AI is well suited to succeed in these areas, whether through empathy, efficiency or by transcending human limitations.
The team arguing against the motion comprised Gerard Reid, co-founder of Alexa Capital, Kay Firth-Butterfield, CEO of Good Tech Advisory, Adele Zeynep Walton, digital activist and independent journalist, and Kirsten Dunlop, CEO of Climate-KIC.
Gerard opened not by addressing AI directly, but by challenging the premise that humans have failed, pointing to declining unemployment and poverty rates, the peaceful reunification of Germany, and the end of the World Wars as evidence of human progress over the past century.
Kay described AI as a tool created by "imperfect people and imperfect data", one that is biased, lacks emotionality and remains dependent on human development to succeed. She called it "a mimic with a good memory", and pointed to its role in generating deepfakes and legal confusion, alongside concerns that reliance on AI in education is eroding critical thinking. Her conclusion was that AI repeats human bias rather than creating something genuinely new.
Adele spoke from personal experience about the emotional risks of treating AI as a substitute for human support and connection.
Kirsten questioned the framing of success and failure altogether, arguing that AI might help a company succeed while contributing to broader societal failure. In her view, the core issues facing society are problems of incentive and legitimacy, not efficiency or optimisation, and these require human solutions.
Where humans fail most consistently, and at the greatest scale, is long-term thinking, climate change being the clearest example. She noted that "we escape accountability, truth and epistemics", and questioned whether AI can help us engage in the value-based judgements that long-term solutions require. In her view, AI's success in creating lasting change is inseparable from humans first succeeding in governing and shaping it responsibly.
The jury comprised Matthew Prince, co-founder and CEO of Cloudflare, Kate Kallot, founder and CEO of Amini AI, Mirjana Spoljaric Egger, president of the International Committee of the Red Cross, and Cully Cavness, co-founder, president and chief strategy officer of Crusoe.
Matthew Prince picked up on the debaters' discussion of incentive failure, asking, "What are AI's incentives?" and pressing further: could AI help illustrate the very incentive failures that prevent us from solving these problems, acting as a neutral arbiter? This pushed the 'For' team to clarify AI's intentions and the 'Against' team to consider whether AI could help address human incentive failures directly.
Kate Kallot asked debaters to consider whose success and failure AI is actually promoting, citing the 2.6 billion people without internet access. If AI is to succeed for everyone, she argued, rather than "just serve one side of the population", it needs to be genuinely inclusive. She returned to Matthew's point, asking what incentive AI itself will have to keep "driving the world in the right direction."
Mirjana Spoljaric Egger asked how debaters would address the global trend of growing inequality, one of the defining challenges of our time, particularly since the pandemic. She questioned whether AI can meaningfully solve this, and how humans will retain agency over the technology and accountability for its consequences.
Cully Cavness argued that both teams had drifted from the actual motion, debating whether AI is good or bad rather than whether it can succeed where humans have failed. He challenged the 'For' team to focus on AI's concrete potential to address the climate crisis, through fusion, flexible power consumption, and carbon capture and storage, and pushed the 'Against' team to acknowledge the role technology and innovation have already played in shaping the lives we have today.
By the audience 'clap-o-meter', the 'Against' team won. As with the Women's Forum debate earlier in the 2025 to 2026 season, the more significant outcome was the willingness of speakers and audience to engage seriously with a question this consequential.
1. Uplifting diverse perspectives and spreading truth is vital to rebuilding collective trust
A consistent thread across the debate was the link between truth and trust. In a related conversation at the same WEF 2026 gathering, André Hoffman, co-chair of the World Economic Forum, and Meredith Kopit Levien, president and CEO of the New York Times, discussed how continuously seeking truth and remaining open to difficult conversations rebuilds the connections that trust depends on. Truth here is not one person's correct perspective, but an attitude of staying open to a range of backgrounds and views, which together build a fuller picture of the world.
The debate format itself plays a part in this. It creates a structured space to disagree openly, surfacing tensions that might otherwise stay hidden, and in doing so builds the kind of mutual understanding that trust requires.
AI's relationship to truth and trust cuts both ways. Adam Grant argued that AI can promote truth by helping disprove misinformation, including conspiracy theories. Kay Firth-Butterfield argued the opposite, pointing to AI's role in generating deepfakes and amplifying bias. Both are true: AI has the potential to either rebuild trust through truth, or deepen division through misinformation, depending on how it is used.
2. Humans must take collective action to ensure AI is grounded in human values
This leads to a second outcome: people need to actively define their values in relation to AI, rather than letting its use develop by default. Jury member Mirjana Spoljaric Egger asked plainly: who is going to be accountable? Responsibility for AI's impact sits with all of us. But as Kirsten Dunlop pointed out, we are currently navigating real "value conflicts", with competing agendas shaping how the technology develops. Some applications build trust and solve problems; others create confusion and division.
Setting clearer collective intentions for AI requires genuine reflection on what we value. If, as Kirsten suggested, we choose to define success as including the wellbeing of all people and the planet, then priorities follow naturally: as Kate Kallot argued, ensuring wider access to the internet and inclusion in AI's benefits becomes essential, rather than optional. Equally, if we choose to value the planet, AI can be directed towards solving the climate crisis through renewable energy, as Varun Sivaram discussed. AI can serve collective wellbeing, but only if the human intention behind its use is clear and value led.
3. AI is a mirror, and an opportunity for reflective learning
The third outcome is that AI can act as a mirror, reflecting back the values embedded in its creation. Kay Firth-Butterfield described it as "a machine that is built on human frailties, designed to repeat those human frailties... a mimic with a good memory." Meredith Kopit Levien made a related point in the conversation referenced above, describing algorithms as "adjusting to things they already know and already believe in."
This reflective quality cuts both ways. AI repeats existing human bias, but as Kate Kallot suggested, it might also help us see our own biases more clearly than ever before. There is an opportunity here that hasn't existed in quite this form: by treating AI as a tool for studying the values that shape it, we gain a clearer view of where we currently stand, and where we might choose to shift.
This report draws on the World Economic Forum 2026 Cocktails and Conversations series, hosted by the New York Times and Intent, and the New York Times Debate curated by DEBATABLE.