Meta pushes hardware down-market while Zuckerberg eyes prediction platform

Meta launched a new line of AI-enabled smart glasses starting at $299, marking a deliberate push to bring voice-assistant hardware into mass-market reach. The move follows the Ray-Ban Meta collaboration, which sold at a premium tier, and signals that the company sees wearable AI as a distribution wedge worth subsidizing. Reuters reported the pricing Wednesday, positioning the range as the cheapest AI-capable glasses Meta has shipped to date.

Separately, Mark Zuckerberg is exploring a standalone prediction market platform that would operate independently of Facebook, Instagram, and Threads. The New York Times reported that the app would allow users to forecast events and wager reputational or financial stakes, with Meta's existing social properties serving as referral channels. The project reflects Zuckerberg's long interest in forecasting and crowd-sourced probability, and would put Meta in direct competition with Polymarket, Kalshi, and Manifold.

The two bets share a common structure: lower-friction entry and adjacency to the core graph. The $299 glasses make AI assistants ambient and voice-first. The prediction market makes engagement measurable in conviction, not just scroll depth. Both attempt to solve the same strategic problem—how to own the next habit before the platform underneath re-rates.

OpenAI, meanwhile, launched an initiative to identify and patch security vulnerabilities in open-source software, aiming to address supply-chain risk in widely used codebases. TechCrunch noted the effort focuses on bugs in dependencies that flow into production systems, a category of risk that has compounded as AI tooling has accelerated code generation and reuse. The move is defensive infrastructure—less about user acquisition, more about hardening the ecosystem OpenAI depends on.

Inside Defense reported that the growth of the consumer satellite market is starting to constrain military launch availability. As commercial demand for low-earth-orbit capacity rises, defense customers are finding fewer launch windows and higher costs. The dynamic mirrors what happened in cloud compute: commercial scale subsidizes infrastructure build-out, then crowds out the original anchor tenant. The Pentagon is now the price-taker, not the price-setter, in the launch stack.