
Meta CEO Mark Zuckerberg published a 6,500-word essay arguing that superintelligence must be widely available and not centralised by a few companies or governments. He also announced a new open weight AI…
Meta CEO Mark Zuckerberg published a 6,500-word essay arguing that superintelligence must be widely available and not centralised by a few companies or governments. He also announced a new open weight AI model called Glimmer, developed from Muse Spark, which uses less computing power while enabling AI agents. The Hindu reports that Zuckerberg called for the US to compete with China and prevent 'government tyranny' over AI, backing open models as a key advantage.

Zuckerberg rejected fears of mass unemployment, writing that AI will create new jobs like personal biologists and world builders. He envisions everyone having a personal AI agent for daily tasks. He argues distributed AI is safer than concentrated power, adding that human control must be preserved. Times Now quotes him saying the defining question is who will have access to superintelligence.
Two narratives battle: AI will destroy jobs vs it will create new ones, and closed vs open models. Neither side is proven. Zuckerberg's open model push risks misuse by bad actors, while closed systems concentrate power. The real test will be how Meta's Glimmer competes with Chinese open models like DeepSeek. If Glimmer lags in adoption or safety, Zuckerberg's vision weakens. If it thrives, his argument for distributed intelligence gains ground. Watch performance metrics and regulatory response in coming months.
Sources (2): timesnownews.com, thehindu.com
This story was synthesised by AI from the 2 sources linked above.
Updated: this story now draws on 2 sources.