Nvidia's Nader Khalil and Sydney Sykes will lead a session at TechCrunch Disrupt 2026, examining the open vs closed AI debate for startups. The session will explore the trade-offs between open and proprietary AI models and whether either approach can offer a lasting competitive advantage.
Nvidia reported that 145 papers accepted at ICML 2026 cited its Nemotron open models and datasets, showing the rapid advancement of open models.
At the same time, proprietary frontier labs continue pushing model capabilities forward, resulting in a market where the question is increasingly less about whether open models can be useful and more about where each approach makes commercial sense.
The discussion will focus on the practical implications of choosing between open and proprietary models, including cost, control, and the need for infrastructure maintenance. Khalil and Sykes will unpack these questions, offering insights for founders, investors, and line-of-business leads.
"The future is not proprietary versus open, but proprietary and open," said Jensen Huang, Nvidia's CEO, at GTC earlier this year. This statement highlights the hybrid reality of AI development, where companies are combining open and proprietary models rather than treating them as mutually exclusive.
The announcement follows the release of Nvidia's Nemotron 3 Super, an open 120-billion-parameter model designed for agentic workloads. This hybrid approach may ultimately be the most interesting part of the debate, as companies seek to balance flexibility with control.
Nvidia did not say whether one approach is definitively better than the other, and the decision remains complex due to shifting economics and capabilities. The session will provide a platform for understanding the trade-offs and making informed choices.
Source: techcrunch
