Leading AI researcher Eric Zelikman is raising $1 billion to build AI models with emotional intelligence

Eric Zelikman, a prominent figure in artificial intelligence and former researcher at xAI, is on a mission to raise $1 billion for his new venture, Humans&, which is valued at an impressive $4 billion. This ambitious funding initiative comes at a time when the AI sector is witnessing a surge in early-stage investments, with valuations skyrocketing despite a lack of concrete products or revenue streams. In a parallel development, Thinking Machines Labs, founded by ex-OpenAI CTO Mira Murati, recently secured $2 billion in seed funding, reaching a staggering $12 billion valuation. Investors are heavily backing startups helmed by leading researchers, convinced that the next significant advancements in AI will stem from agile, talent-rich teams. Currently in the midst of fundraising, the terms may evolve as discussions continue, and Zelikman has yet to provide comments regarding the undertaking. His academic prowess includes a Ph.D. in computer science from Stanford, where he gained recognition as the lead author of a groundbreaking paper on enhancing language models' ability to process information before responding. Before his tenure at xAI, Zelikman gained valuable experience as a machine learning intern at Microsoft and worked as a deep learning engineer at Lazard. He has expressed concerns that existing language models lack the warmth and understanding necessary for meaningful interactions. In a recent podcast with venture capitalist Sarah Guo, he emphasized, "Today's models often fail to grasp the long-term consequences of their responses, treating each conversational turn as a separate game." Zelikman believes that much of the talent in AI is not being utilized effectively and is determined to change that narrative. His vision for Humans& involves creating models that can empathize and learn from users, stating, "The primary goal of our models should be to understand you. While they may not be perfect, there's significant potential for improvement over current models." By focusing on human-centric model development, Zelikman aims to tackle profound challenges, such as cancer treatment. He asserts, "We stand a better chance of addressing critical human issues by designing models that excel at collaborating with diverse groups, understanding various goals, ambitions, and values."