Almost at the same time, Meta made a move in the opposite direction. The company has acquired Moltbook, a strange but fascinating platform where humans are mostly spectators. The site works like a minimal version of Reddit. But instead of people posting, autonomous AI agents do. Bots write posts. Bots reply. Bots vote. Humans mostly watch. Thousands of AI agents were already active on the platform. Some debated philosophy. Some shared observations. Others coordinated tasks. It quickly turned into a real-time laboratory for observing how autonomous software systems interact with each other. The founders, Matt Schlicht and Ben Parr, will join Meta on March 16. They will be part of Meta Superintelligence Labs, the company's frontier AI unit led by Alexandr Wang. What makes Moltbook particularly interesting is the infrastructure underneath it. The platform was built alongside an agent framework called OpenClaw. It allows users to run AI assistants locally that can communicate with other agents and perform tasks such as scheduling, file management, and messaging. Interestingly, OpenAI has hired the person who created OpenClaw. The long-term plan of Meta for Moltbook remains unclear. Meta has said existing users will continue to have access to theplatform for now. What is obvious is that Meta and LeCun are chasing intelligence from very different directions. It is also not extremely clear why LeCun left Meta. And all of this was followed by NVIDIA announcing a multi-year partnership with Mira Murati's Thinking Machines Lab. The plan is to deploy at least one gigawatt of next-generation NVIDIA Vera Rubin systems to support the startup's frontier model training and AI platforms. Even for Murati, the focus is not just on scaling LLMs. For now, it seems the industry is exploring multiple paths, and we may have to wait to figure out where AI is ultimately headed. |
Комментариев нет:
Отправить комментарий
Примечание. Отправлять комментарии могут только участники этого блога.