\u201cI think it is rational for them to keep trying. Their current AI efforts haven\u2019t worked as well as they\u2019ve hoped and I respect [them] continuing to be aggressive.\u201d He added that none of his \u201cbest people\u201d had taken Zuckerberg up on his offers.\u00a0<\/p>\n
Meta<\/a> has been racing to poach top researchers and engineers from rival groups as it seeks to build a new \u201csuperintelligence\u201d team to develop artificial general intelligence. <\/p>\n Zuckerberg, the chief executive, has been personally selecting and calling talent as part of the superintelligence push, according to one person familiar with the matter and as first reported by Bloomberg. Last week, Meta announced a $15bn investment <\/a>in data-labelling start-up Scale AI and also hired its co-founder Alexandr Wang.\u00a0<\/p>\n Despite Zuckerberg repeatedly declaring his intention for the company to become an \u201cAI leader\u201d, Meta has struggled this year to keep pace with its rivals, suffering several setbacks.\u00a0<\/p>\n Meta has faced claims that it boosted performance metrics for Llama 4, the latest release of its large language model, and has been subjected to online criticism for not releasing a full technical report to accompany the model. In a post on X in April, Ahmad Al-Dahle, Meta\u2019s head of GenAI, acknowledged the \u201cmixed quality\u201d users were experiencing, but denied claims that its performance tests did not follow convention.<\/p>\n Recommended<\/p>\n It has also delayed the release of its flagship AI model called \u201cBehemoth\u201d, and was caught off-guard by rapid advancements in the field by smaller Chinese rival, DeepSeek<\/a>, made at a fraction of the cost.\u00a0<\/p>\n Meta has also recently experienced an exodus of AI talent, with several key researchers who built the Llama models leaving over the past few months, as well as Joelle Pineau, head of AI research. Zuckerberg has been reshuffling its generative AI leadership in a bid to catch up.\u00a0Meta did not immediately respond to a request for comment. <\/p>\n Altman said that the strategy of \u201ca ton of upfront guaranteed comp\u201d was a poor one, criticising \u201cthe degree to which they are focusing on that and not the work and not the mission\u201d. He added: \u2018\u2019I don\u2019t think they\u2019re a company that\u2019s great at innovation.\u201d\u00a0<\/p>\n Salaries for a software engineer at OpenAI range from around $238,000 to $1.34mn, according to financial package tracking website Levels, while Meta\u2019s salaries varied from $212,000 to around $3.7mn.<\/p>\n![\"Meta](\"https:\/\/www.europesays.com\/uk\/wp-content\/uploads\/2025\/06\/https:\/\/www.ft.com\/__origami\/service\/image\/v2\/images\/raw\/https%3A%2F%2Fd1e00ek4ebabms.cloudfront.net.jpeg\")
<\/a><\/p>\n