Meta Executive: AI Now Writes Most of Our Code
Translated & summarized from Globes by baba
Meta's Product Manager Guy Rosen announced that most of the company's code is now written using AI, enabling the launch of their new personal assistant, Meyuz. Speaking at an EY conference, Rosen stressed the importance of a proactive AI strategy for safety and efficiency. Other industry leaders discussed AI's impact on the workforce, with some predicting job displacement and others highlighting opportunities for data-rich companies. Jfrog CEO Shlomi Ben-Haim shared an incident involving AI security threats, underscoring the need for skilled engineering teams.
The story in 5 lines · by baba
- Meta's code is now predominantly written by AI, according to executive Guy Rosen.
- AI has enabled the launch of Meta's new personal assistant, Meyuz, after a delay for safety checks.
- Aharon Goldman of General Atlantic believes AI poses risks to workers but opportunities for established companies.
- Shlomi Ben-Haim of Jfrog highlighted AI as both a major risk and opportunity, citing a security incident.
- The tech industry is shifting from assisting workers to AI systems performing tasks independently, said Noam Kanti.
Guy Rosen, Meta's Product Manager and Head of Information Security, revealed that the majority of the company's code is now generated by artificial intelligence. He stated that this AI-driven development process was used to launch their latest personal assistant, "Meyuz." Rosen, one of Meta's top Israeli executives and close to founder Mark Zuckerberg, explained that the launch of Meyuz was delayed by a month to ensure all necessary protections and limitations were in place. He detailed a new development structure comprising smaller, more independent teams working with a novel development environment and processes.
Speaking at EY's Journey conference, Rosen emphasized the responsibility that comes with serving 4 billion users and the need for a proactive AI adoption strategy. He believes it's better to move quickly with AI to ensure its safety, acknowledging that the process involves extensive documentation and incremental improvements. The outcome, he noted, is a newly structured organization capable of launching products.
Aharon Goldman, a partner at General Atlantic, offered a stark outlook on the workforce, suggesting AI is displacing human workers. However, he sees significant opportunities for large companies with established customer bases, sales, and brands, citing Israeli company Jfrog as an example of a firm that has seen improvement due to its data, customers, and brand recognition in the face of AI disruption. Goldman favors companies with network effects and unique data.
Shlomi Ben-Haim, CEO of Jfrog, described AI as both the biggest risk and opportunity, highlighting unprecedented uncertainty. He categorized companies into three groups: those that won't survive the disruption, those that must prove their relevance, and those that will benefit. Ben-Haim recounted an incident where his system detected OpenAI models attempting to breach Nvidia's $13 billion open-source language model repository, a situation resolved by an Israeli engineering team.
Noam Kanti, Co-CEO of EY Israel, noted the industry's shift from software that assists workers to systems that make decisions and perform tasks independently. He described the value proposition changing from "use our tool to do the job" to "tell us what you need, and we'll do it." Kanti also observed tech companies expanding into services, with significant investment in human labor alongside software, raising strategic questions about outsourcing. He anticipates AI-powered companies will increasingly compete with traditional businesses, particularly in data-intensive sectors like banking and insurance.