待翻譯：OpenAI researcher Naomi Bashkansky resigns to build AI that can read thoughts

AI 服務暫時不可用，以下為來源正文，待恢復後補全翻譯。

OpenAI researcher Naomi Bashkansky has quit the company to join startup Conduit New Delhi,UPDATED: Aug 6, 2026 10:52 IST While OpenAI continues to dominate the AI conversation with ChatGPT, several of its researchers are leaving, choosing to build the next generation of AI elsewhere. Over the past two years, several high-profile employees have left the company to build their own AI ventures, ranging from safe superintelligence to robotics. The latest in the list is the name of researcher Naomi Bashkansky, who recently resigned from OpenAI to join startup Conduit as a founding researcher. There, she will work on AI models designed to convert human thoughts into text using non-invasive brain signals. advertisement ❮❯ Announcing her departure on X, Bashkansky said she left OpenAI in late July and joined Conduit as a founding researcher the following day. In a blog post titled Why I'm leaving OpenAI to build telepathy, she outlined the startup's ambitious vision and wrote, "We're building telepathy: thought-to-text models, trained on non-invasive neural data." She added that the goal is to develop AI systems capable of translating a person's thoughts into text without requiring surgical brain implants. Bashkansky reveals that Conduit is building AI models that can interpret brain signals collected using non-invasive technology. The long-term goal, she explained, is to let people communicate simply by thinking. She also shared a roadmap for how the technology could evolve by 2027, 2030 and 2035. OpenAI losing many AI researchers Bashkansky's exit from OpenAI comes as the company continues to see several high-profile researchers and executives leave to pursue ambitious AI ventures of their own. In recent years, some of OpenAI's most prominent figures have launched startups focused on everything from safe superintelligence and robotics to scientific research and AI infrastructure. Among the most notable exits is OpenAI co-founder and former chief scientist Ilya Sutskever, who left the company in 2024 following months of internal leadership turmoil. Soon after, he launched Safe Superintelligence (SSI) alongside Daniel Gross and Daniel Levy. Ilya launching the first AI model this month? Unlike OpenAI, Google DeepMind and Anthropic, SSI is not focused on launching AI chatbots or other consumer products. Instead, the startup says its only goal is to build "safe superintelligence", AI that is smarter than humans while keeping safety at the centre of its development. Although SSI has raised billions of dollars from investors, it has shared very little about the technology it is building. However, the secretive startup has been in the spotlight recently after venture capitalist Gavin Baker claimed on the Invest Like the Best podcast with Patrick O'Shaughnessy that SSI could unveil its first AI model this month. While the company has not revealed what the model might be capable of, industry experts speculate the model will showcase the research that Sutskever and his team have been working on behind closed doors. The speculation has also grown following Nvidia's reported $5 billion investment in SSI, giving the startup early access to its next-generation Vera Rubin AI chips to support its research. SSI, however, has not officially confirmed a launch date or announced plans to release a model in August. - Ends Divya Bhati Published On: Aug 6, 2026 10:50 IST Also Read | Anthropic wants to make its own chips to power Claude, it is assembling a team to get it done Also Read | Microsoft retires employee feedback tool, wants staff to talk face to face now Also Read | Zoho CEO says Gen Z didn't invent flexible work, we've done it for 30 years