Zuckerberg Slams OpenAI & Anthropic, Backs Open-Source AI

Meta CEO Mark Zuckerberg has released a 6,500-word essay that appears to criticize large language model leaders OpenAI and Anthropic. Tensions have been growing, with Zuckerberg previously stating these companies centralize power, harming innovation and posing risks. What's interesting is Zuckerberg advocates for open-source AI models. He argues these models, accessible and modifiable by anyone, are safer from a cybersecurity standpoint. This is because more people can identify vulnerabilities and upgrade systems. OpenAI and Anthropic, however, use proprietary, closed-loop models. Zuckerberg's essay coincides with Meta launching new open-source models for laptop use and planning to open the weights on its latest AI model. He also seems to challenge Anthropic's CEO, who has expressed extreme caution about AI. The bottom line is this essay highlights a significant debate in the rapidly evolving field of artificial intelligence.