NVIDIA’s Khalil and Sykes Speak at Disrupt 2026
▼ Summary
– A session at TechCrunch Disrupt 2026 led by Nvidia executives will explore the strategic trade-offs between using open and proprietary AI models for startups.
– Founders face critical decisions regarding cost, control, and infrastructure that impact their product development and competitive advantage in a rapidly shifting market.
– The gap between open and proprietary models is closing, with Nvidia citing significant research adoption of its Nemotron open models alongside continued advancements by proprietary labs.
– Nvidia CEO Jensen Huang suggests a hybrid future where both open and proprietary approaches coexist rather than framing them as mutually exclusive options.
– Discussion leader Nader Khalil brings expertise from his background in AI infrastructure and his previous co-founding of Brev.dev to address builder-centric challenges.
Choosing an AI foundation is one of the most critical strategic moves a startup can make. Founders must decide whether to rely on proprietary frontier models for speed, leverage open-source architectures for control, fine-tune custom versions, or deploy solutions locally. This choice ripples through every aspect of a business, influencing infrastructure costs, profit margins, operational speed, and long-term differentiation. There is no single correct path, but selecting the wrong model can create lasting liabilities as market dynamics shift.
This complex dilemma takes center stage in the session “The Open vs. Closed AI Debate Is Just Getting Started” at the Builders Stage during TechCrunch Disrupt 2026. Scheduled for October 13-15 in San Francisco, the discussion is led by Nader Khalil, Director of Developer Tech at Nvidia, and Sydney Sykes, Global Head of VC Partnerships. They will explore the tangible trade-offs between open and proprietary ecosystems, examining whether either approach offers a sustainable competitive edge in today’s rapidly evolving landscape.
The Closing Gap Between Open and Proprietary Models
The distinction between open and closed AI is becoming increasingly nuanced. According to Nvidia, the momentum behind open-source technology is accelerating. In July, the company reported that 145 papers accepted at ICML 2026 cited its Nemotron open models and datasets. These research efforts span diverse fields, including robotics, autonomous vehicles, and biomedical science, highlighting the practical utility of open frameworks.
Simultaneously, proprietary labs continue to push the boundaries of model capabilities. Consequently, the industry conversation has shifted from questioning the usefulness of open models to determining where each approach makes commercial sense. Nvidia itself rejects a binary view of this issue. During GTC earlier this year, CEO Jensen Huang argued that the future lies not in choosing between proprietary and open systems, but in utilizing both.
For founders, this hybrid reality presents difficult questions. If two models deliver comparable performance, does lower cost become the deciding factor? Does greater data control outweigh the benefits of a managed service? How should a product roadmap align with a model that may change significantly within six months? These are the practical challenges Khalil and Sykes aim to address, moving beyond abstract philosophy to focus on actionable business decisions.
Dual Perspectives on Infrastructure and Venture Strategy
Khalil brings a deep technical background to the debate. Before joining Nvidia, he co-founded Brev.dev, an AI infrastructure company acquired by Nvidia in July 2024. His work focused on simplifying access to GPU resources across public cloud, private cloud, and on-premises environments. Nvidia’s documentation emphasizes that such tools allow developers to avoid vendor lock-in, providing flexibility in how AI software is deployed.
Sykes complements this view with expertise from the venture capital ecosystem. As Nvidia’s Global Head of VC Partnerships, she understands what makes startups investable and scalable. Her perspective focuses on the business requirements behind technical choices, offering insight into how model selection impacts fundraising narratives and market positioning.
Together, these leaders provide a comprehensive look at the AI stack. They examine what developers need to build efficient systems and what entrepreneurs must do to create defensible businesses. This dual lens helps attendees understand the intersection of technical execution and commercial viability, a crucial consideration for any organization building AI-driven products.
Defining Competitive Advantage Beyond the Model
A common misconception is that the model itself serves as a primary moat. In reality, if competitors have access to the same proprietary APIs or open weights, differentiation must come from elsewhere. True defensibility often stems from proprietary data, specialized workflows, distribution channels, customer relationships, or unique product experiences.
Choosing an open model offers flexibility and potential control over data, but it also introduces responsibilities around deployment, optimization, and infrastructure maintenance. The economics of running local models can vary significantly depending on workload and scale. Nvidia is actively supporting this open ecosystem with initiatives like Nemotron 3 Super, launched in March. This open 120-billion-parameter model is designed for agentic workloads, and companies are already experimenting with combining it with proprietary models rather than treating them as mutually exclusive options.
This hybrid approach may ultimately define the most successful strategies. It allows organizations to balance the innovation of frontier models with the cost-efficiency and control of open-source alternatives. Understanding how to integrate these technologies is key to building resilient AI products.
Strategic Implications for Founders and Investors
The insights shared at TechCrunch Disrupt 2026 extend far beyond engineering teams. For founders, the decision on which AI stack to adopt directly influences fundraising stories, product roadmaps, and overall margins. Investors benefit from understanding where value resides in the stack, helping them distinguish genuine defensibility from superficial product layers built on third-party models.
Line-of-business leaders must consider how model choice affects procurement processes, security protocols, and data governance. Developers and students gain valuable context on how today’s technical decisions shape tomorrow’s business models. The goal is not to declare a philosophical winner between open and closed AI, but to clarify the trade-offs involved.
Attendees will have the opportunity to engage with Nader Khalil and Sydney Sykes on the Builders Stage. Registration is currently available, with savings of up to $200 offered before prices increase on September 25 at 11:59 p.m. PT. This session provides a vital forum for making informed decisions about your AI strategy.
(Source: TechCrunch)
