OpenAI acquires startup Orchestra Bio to beef up its biotech business
THE SO WHAT
Folding Orchestra Bio into OpenAI’s Rosalind and Workbench efforts is a bet that wet-lab planning and analysis can be standardized into an AI-first workflow. Biotech teams should start evaluating where they want to own proprietary data and protocols versus plugging into a shared model-and-harness stack.
READ THE SOURCE
MORE FROM THE WIRE
Applied AIAnthropic’s New Agent Has No Cute Mascot, and You Probably Can’t Try It Yet
Positioning Claude Duet as a serious, gated agent rather than a mass-market mascot is a signal that high-trust, high-stakes workflows are the target, not consumer novelty. If you’re building on top of foundation models, expect a split between open, playful assistants and tightly controlled agents that look more like enterprise software than chatbots.
Applied AIMeta’s stock rises as investors glimpse an opportunity that could dwarf the ad business
If investors think Meta’s Muse assistant could one day rival ads, they’re effectively underwriting a shift from attention arbitrage to intent and workflow ownership. Any consumer or SMB app that relies on Meta’s distribution should assume that assistant surfaces will become contested real estate for discovery and commerce.
Applied AIAI helped uncover a security nightmare in the web’s backbone. Should you be worried?
AI surfacing a ‘security nightmare’ in core web infrastructure shows that attackers and defenders now share the same acceleration tools. Treat this as a forcing function — if you’re not using AI to continuously probe your own stack, assume someone else is doing it for you.
Meta CEO Mark Zuckerberg says Muse and Reality Labs are highly unprofitable — for now
Zuckerberg calling Muse’s compute costs and Reality Labs’ hardware losses ‘highly unprofitable’ is a clear signal that Meta is willing to subsidize AI assistants and devices at scale. If you’re building consumer AI or XR, assume years of price and performance pressure from a player treating these as strategic, not near-term P&L businesses.