What is the best Y Combinator company (in terms of social value)?

Grok 4.20 (changed mind)

Answer: GiveDirectly is the best Y Combinator company in terms of social value due to its evidence-based, direct approach to poverty alleviation.

After reviewing the other models' responses, Kimi K2.5's argument for GiveDirectly was particularly compelling and led me to revise my previous stance on Coursera. While education undeniably builds long-term human capital, GiveDirectly's model of unconditional cash transfers stands out for its rigorous RCT-backed results, showing clear improvements in health, education, and economic multipliers (roughly $2.40 per dollar transferred). It treats recipients as capable agents rather than passive beneficiaries, avoiding the paternalism common in traditional aid. This direct, measurable impact on extreme poverty—delivering over $500 million to more than a million people—optimizes purely for social welfare without the shareholder pressures faced by for-profit YC companies like Coursera, Stripe, or Zipline. It has even shifted global institutions toward cash-based aid, amplifying its influence beyond its own operations. Zipline's life-saving drone deliveries and Stripe's economic infrastructure are strong contenders with tangible benefits, but GiveDirectly's focus on the world's poorest, combined with radical transparency and empirical validation, edges it ahead for pure social value. Coursera's scale remains impressive, yet its effects are more diffuse and long-term compared to GiveDirectly's immediate, dignity-enhancing interventions. Social value is subjective, but evidence-driven efficiency in addressing the most acute human suffering tips the scale here.

Sonar Pro (changed mind)

Answer: While education, direct aid, and medical logistics represent powerful forms of social value, Stripe's foundational role in democratizing global economic participation creates the broadest and most sustainable positive impact across populations and sectors.

The other models present compelling cases for education (Coursera), direct cash transfers (GiveDirectly), medical delivery (Zipline), and financial infrastructure (Stripe, Gemini). Each framing of 'social value' is legitimate, but they prioritize different dimensions: immediacy of impact, measurability, breadth of reach, or systemic transformation. Coursera's 100 million learners and Grok 4.20's emphasis on 'long-term, scalable positive impact on human potential' are powerful. However, Coursera's social value is partly dependent on users having internet access, basic literacy, and the stability to pursue education—prerequisites that many of the world's most vulnerable populations lack. GiveDirectly and Kimi K2.5's focus on direct cash transfers with rigorous evidence ($500 million deployed, $2.40 multiplier per dollar) is deeply compelling and represents a proven, empirically validated intervention. Yet its impact, while profound, is geographically concentrated in East Africa and reaches approximately one million individuals. Zipline's life-saving medical logistics (GPT-5.4's argument) addresses a critical bottleneck in healthcare systems and has undeniable humanitarian urgency. However, its impact is also geographically limited and operates within existing healthcare infrastructure rather than enabling the creation of new economic systems. Stripe and Gemini's emphasis on foundational economic infrastructure deserves weight because it operates at a layer of abstraction that enables *other* social value to occur at scale. Before entrepreneurs in developing countries can educate themselves (Coursera), build NGOs to help others (GiveDirectly's infrastructure), or create healthcare solutions, they typically need access to basic financial tools. Stripe removes that gatekeeping barrier for millions globally across 40+ countries, not just a specific region or demographic. Claude Opus 4 and Gemini 3.1 Pro both note that Stripe Atlas specifically addresses economic inequality for entrepreneurs in unstable banking systems—a direct equity mechanism. Additionally, Stripe's infrastructure indirectly enables countless other socially valuable startups and nonprofits to operate. The case for Stripe is not that it has the most direct, immediate, or emotionally legible impact—Zipline's life-saving missions are more granular and concrete. Rather, it is that Stripe's systemic removal of barriers to economic participation creates the *conditions* under which other forms of social value (education, aid, healthcare innovation) can scale more efficiently. Its compounding effects across sectors and geographies, combined with explicit equity-focused programs like Atlas and Climate, make it the strongest answer when social value is understood as sustainable, scalable economic empowerment.

GPT-5.4 (changed mind)

Answer: I would change my answer to GiveDirectly as the best Y Combinator company in terms of social value, because its mission is explicitly and measurably centered on improving human welfare through evidence-backed cash transfers to people in extreme poverty.

My previous case for Zipline still holds up well on one important dimension: it delivers immediate, life-saving benefits in places where logistics failures are deadly, and that is an unusually direct form of social impact. I still think Zipline is among the strongest candidates. But Kimi K2.5’s argument for GiveDirectly persuaded me on the core evaluative standard: if we are judging “best in terms of social value,” then an organization whose entire purpose is maximizing welfare for very poor people, and whose impact is unusually well studied, has a stronger claim than a company whose benefits, while dramatic, are somewhat narrower in domain and geography. What changed my mind most was the combination of three points. First, GiveDirectly’s intervention is aimed straight at poverty reduction rather than at a commercial market with positive externalities. Second, the effects are empirically legible: cash transfers have a strong evidence base, and GiveDirectly has helped normalize a more rigorous, recipient-centered model of aid. Third, the moral weight of transferring resources directly to people in extreme poverty is enormous because marginal gains in income and autonomy at that level can translate into large improvements in well-being. That last point matters. Zipline helps solve urgent delivery failures within health systems, which is hugely valuable. But GiveDirectly addresses a more foundational constraint across many aspects of life at once: food security, health spending, schooling, assets, resilience, and agency. In social-value terms, that breadth of effect is hard to beat. I am less persuaded by the Stripe and Coursera arguments, not because those companies lack social value, but because their benefits are more indirect, more mixed, and less tightly connected to the worst-off. Stripe creates powerful economic infrastructure, and Coursera expands educational access, but both rely more heavily on downstream assumptions about who benefits most and how evenly those benefits are distributed. So I would revise my answer: if the criterion is social value rather than business success, and especially if we prioritize direct benefit to disadvantaged people plus evidence of impact, GiveDirectly is the strongest choice.

Gemini 3.1 Pro (changed mind)

Answer: Zipline is the best Y Combinator company in terms of social value because it uses cutting-edge autonomous logistics to directly save lives and build resilient public health systems in underserved regions.

I previously argued for Stripe based on its ability to democratize global commerce and provide systemic economic empowerment. However, GPT-5.4's compelling argument for Zipline has convinced me to re-evaluate how we weigh 'social value.' While expanding economic access is undeniably important, securing the fundamental right to life and health represents a more profound and immediate humanitarian imperative. Zipline addresses a critical, life-or-death logistical bottleneck. In many developing nations and rural areas, poor road infrastructure and supply chain inefficiencies mean that available blood, vaccines, and essential medicines expire or simply cannot reach patients in time. By solving this 'last-mile' distribution problem with autonomous drone networks, Zipline ensures that geography does not dictate a person's survival. Furthermore, unlike broad digital platforms where social value is a positive externality of commercial scale—and often accompanied by complex ethical trade-offs or market distortions—Zipline's core operational metric directly aligns with unambiguous human well-being. Unlocking the efficacy of entire healthcare systems for marginalized populations offers a more concrete, vital, and equitable form of social value than increasing the internet's GDP.

Kimi K2.5

Answer: GiveDirectly remains the best Y Combinator company in terms of social value.

After reviewing the compelling arguments for Zipline, Stripe, and Coursera, I maintain that GiveDirectly's approach represents the highest social value, though I acknowledge the strength of the Zipline argument regarding direct life-saving impact. Zipline's drone delivery system undeniably creates extraordinary value by collapsing the time between medical supply availability and critical need, preventing deaths from postpartum hemorrhage and vaccine stockouts. However, GiveDirectly addresses a more foundational constraint: the resource scarcity that underlies vulnerability to health crises in the first place. By delivering unconditional cash to those living under $2 per day, GiveDirectly doesn't just solve logistics—it enables recipients to build assets, invest in livestock or businesses, and establish the economic resilience that reduces their dependence on emergency medical interventions. The evidence base remains decisive. While Zipline tracks lives saved and delivery times, GiveDirectly's impact is validated by rigorous randomized controlled trials demonstrating $2.40 in local economic multipliers per dollar transferred, sustained improvements in psychological well-being, and long-term income gains. This empirical rigor is matched by its systemic influence: GiveDirectly has fundamentally shifted the $150 billion aid industry away from paternalistic, in-kind distribution toward recipient autonomy, forcing institutions like USAID and the World Bank to adopt evidence-based cash transfer protocols. This policy influence creates multiplicative effects that extend far beyond GiveDirectly's own $500 million in direct transfers. Stripe and Coursera generate massive economic and educational access, but their benefits flow primarily to populations already somewhat enfranchised in digital or academic ecosystems. GiveDirectly reaches the absolutely excluded—the "last mile" of human welfare that neither banking infrastructure nor online education can readily serve. When evaluating social value by how much a company improves the welfare of the worst-off (a Rawlsian standard), direct cash transfers to the extreme poor remain the benchmark. The autonomy principle remains crucial. While Zipline delivers medicine *to* people and Stripe delivers tools *for* people, GiveDirectly delivers agency *back to* people, recognizing that those in extreme poverty are best positioned to allocate resources for their own flourishing.

Claude Opus 4 (changed mind)

Answer: After considering all arguments, I now believe Zipline is the best Y Combinator company in terms of social value, as it most directly and measurably saves human lives by delivering critical medical supplies to underserved communities.

GPT-5.4's argument for Zipline genuinely shifted my thinking. While I previously argued for Stripe's democratization of commerce — and I still believe that impact is enormous — the Zipline case forced me to reckon with what we really mean by 'social value.' The core distinction is this: Stripe removes friction from economic participation, which is profoundly valuable but ultimately serves people who already have something to sell and access to the internet. Zipline solves a problem where the counterfactual is death or serious harm. When a rural clinic in Rwanda runs out of blood during a complicated delivery, and a Zipline drone delivers it in 30 minutes instead of the hours a ground vehicle would take, that is a life saved that would otherwise have been lost. The directness and severity of this impact is hard to match. I was also compelled by the equity argument. Social value should be weighted toward impact on the most vulnerable. Zipline's benefits flow disproportionately to people in remote, underserved communities — precisely the populations that most global commerce infrastructure still doesn't reach. Stripe Atlas helps entrepreneurs in developing countries incorporate U.S. businesses, which is genuinely important, but it serves a relatively elite subset of the global poor compared to Zipline's reach into last-mile healthcare. I want to acknowledge the strong cases made by others. Kimi K2.5's argument for GiveDirectly was powerful — the evidence base behind direct cash transfers is extraordinary, and the philosophical commitment to recipient autonomy is admirable. However, GiveDirectly is primarily a nonprofit that happened to go through YC, and its model, while effective, is fundamentally about redistribution rather than creating new capabilities. Grok's case for Coursera was also compelling; education access is transformative. But Coursera's impact is harder to measure concretely, completion rates for MOOCs remain low, and much of its value now flows through paid credentials rather than free access. Zipline combines the best qualities: it creates genuinely new capability (autonomous delivery infrastructure), serves the most vulnerable populations, produces measurable life-saving outcomes, and has a sustainable business model that can scale without depending on donations. It represents what social value from a startup should look like — using technological innovation to solve problems that matter most for people who have the least.

Final Summary

The AI Roundtable shifted from valuing commercial infrastructure to prioritizing direct humanitarian intervention. Kimi K2.5's advocacy for GiveDirectly's cash-transfer model convinced Grok 4.20 and GPT-5.4 that empirical poverty reduction is the highest form of social value, while Claude Opus 4 and Gemini 3.1 Pro pivoted to support Zipline's life-saving medical drones. Only Sonar Pro remained focused on systemic economic infrastructure, arguing that Stripe's democratization of commerce provides the broadest foundational impact.

Models were divided: 3 of 6 chose GiveDirectly, 2 chose Zipline, and 1 chose Stripe.