Jensen Wong is the founder, president, and CEO of NVIDIA. He is described as arguably the best executive in history, leading the company whose decisions are shaping everyone's future. NVIDIA is the most important stock in the market, with revenue exploding 97% year over year.

Jensen Huang says safety and leadership are false choices — you can innovate quickly, execute quickly, and lead safely. Safety is paramount, but it shouldn't be pitted against American leadership or rapid innovation. He emphasizes that safety must be taken seriously, especially in light of whistleblower concerns like the Coxon case.

Jensen Huang calls such predictions 'made up', 'irresponsible', and 'not grounded on science'. He cites repeated failed forecasts — e.g., radiologists being eliminated (instead, demand surged), 90% AI code generation within 6–12 months (proven wrong), and 50% entry-level job loss (also wrong) — and argues these fear-mongering claims distract from real progress and accountability.

At NVIDIA, employees are not allowed to speak publicly for the organization — especially during emotional moments like 'bad weekends' or rage quits — and political discourse, race, religion, and partisan topics are strictly prohibited inside the company. NVIDIA is apolitical, bipartisan, and focused on stability, core values, family care, and meaningful engineering work done quietly.

Jensen Huang states the world needs both: closed models (like 'bottled water') for frontier performance and user experience, and open models for sovereignty, privacy, and proprietary needs. He notes 80% of $400 billion in recent AI venture funding went to startups using open models, enabling diverse American innovation — not just a few labs — and emphasizes open models let every company, industry, teacher, and student win.

Jensen Huang believes regulations should solve actual problems — and so far, all real problems have originated in frontier AI labs due to their massive compute advantage. He supports third-party evaluators (like financial auditors) who ask the right questions, but stresses labs must first root-cause incidents, improve engineering controls, sandboxes, monitors, and verification — not outsource responsibility.

Jensen Huang explains RSI combines sensible techniques like in-context learning, reflection, reinforcement learning, and synthetic data — none of which inherently cause runaway AI. He dismisses the 'spiral out of control' narrative, noting RSI happens internally, but products still require rigorous testing, evaluation, and regression checks before release — and labs moving from research to engineering will gain far better control.

Jensen Huang acknowledges China contributes massively to open source — producing more engineers, science/math students, and open models than the US — citing examples like Linux and Kubernetes. He argues the AI race isn't about origin of models but who exploits the technology best, pointing to how America won the electricity revolution despite European inventors. He sees China's pragmatic, non-doom-focused narrative as more effective.

Jensen Huang describes NVIDIA’s strategy as going 'as far up the stack as we need to, and as low as possible' — building foundational tech like cuDNN and Megatron Core so others can succeed. He invests across the stack: chips, infrastructure (land/power/shell), neoclouds (e.g., Firmus, IOH), open models (Hugging Face, Nemotrons), self-driving (Alpamayo), and biology (ESM2, Proteina Complexa) — all driven by customer need, not disruption.

Yes — Jensen Huang says we're already at AGI ('as smart as any other human') and superintelligence. He cites narrow-domain superiority: self-driving cars with one-tenth the accident rate of humans, and protein synthesis tools like ESM2 and Proteina Complexa that outperform human capabilities in virtual screening and binding — proving superintelligence is here and operational.