Former Intel CEO Pat Gelsinger Launches AI Alignment Benchmark: Flourishing AI

Former Intel CEO launches a benchmark to measure AI alignment | TechCrunch

After former Intel CEO Pat Gelsinger capped off a 40+ year career at the semiconductor giant in December, many wondered where would Gelsinger go next. On Thursday, the former Intel CEO revealed one piece of his next chapter: trying to ensure AI models support a flourishing humanity. In partnership with a "faith tech" company he first invested in roughly 10 years ago called Gloo, Gelsinger launched a new benchmark -- Flourishing AI, or FAI -- to test how well AI models align with certain human values. The FAI benchmark is based on The Global Flourshing Study, a survey directed by Harvard and Baylor University, to measure human well-being around the world. Gloo took six core categories from the study -- Character and Virtue; Close Social Relationships, Happiness and Life Satisfaction, Meaning and Purpose, Mental and Physical Health, Financial and Material Stability -- and added one more, Faith and Spirituality, to test LLMs.

Ex-Intel CEO Pat Gelsinger has a new mission for AI

Former Intel CEO Pat Gelsinger, in partnership with the "faith tech" company Gloo, launched the Flourishing AI (FAI) benchmark on Thursday, to assess artificial intelligence model alignment with human values. Gelsinger's involvement follows his departure from Intel in December, concluding a career spanning over four decades at the semiconductor company. He had initially invested in Gloo approximately ten years prior to this announcement. The FAI benchmark is derived from The Global Flourishing Study, a survey collaboratively directed by Harvard University and Baylor University, which focuses on global human well-being. AWS is getting an AI agent marketplace Gloo adapted six core categories from The Global Flourishing Study for the FAI benchmark: Character and Virtue, Close Social Relationships, Happiness and Life Satisfaction, Meaning and Purpose, Mental and Physical Health, and Financial and Material Stability. The company additionally integrated a seventh category, Faith and Spirituality, for evaluating large language models (LLMs). Gelsinger stated in an interview with The New Stack that he has "lived at the intersection of faith tech my entire life," indicating a long-standing personal interest in the domain. The benchmark aims to provide a standardized method for evaluating whether AI systems contribute positively to human flourishing.