What researchers are saying

Lambert says Huang’s exit leaves US open models without their patron

Nathan Lambert, AI researcher and author of Interconnects

Lambert argued that Jensen Huang’s resignation is a setback for open-weight models in the US, because his funding and political weight gave a young open-model economy time to develop.

Nathan Lambert argued that Nvidia still has a commercial reason to support open models, since a market of many model builders limits the power of a few large customers that design their own chips. He wrote that the company’s first task now is to stabilise its operations, and that Huang’s founder-style funding of open models will very likely dry up.

In his view the open-model economy is still young. Its strength is enterprise post-training and ownership of models; its weakness is that compute is scarce, so open models are served at high margins and are far from a commodity, which favours closed labs with integrated chips and data centres.

His main near-term concern is hasty regulation at state level and then nationally, driven by political pressure for AI safety. He wrote that Huang had used his influence to protect open models and that no one of similar standing is now in that role.