I know you don't want them to want AI, but (www.anildash.com)

Mozilla’s community debate over adding an “AI pane” to Firefox has ignited fierce backlash—largely from privacy- and IP-conscious contributors—but Rodrigo Ghedrin’s framing that “nobody wants AI in Firefox” misses the broader reality the author observed at MozFest: while many technologists distrust Big Tech’s AI for harms like content appropriation, labor erosion, environmental cost and misinformation, hundreds of millions of mainstream users actively choose AI tools for amusement and utility. That split explains the heat: experts see legitimate risks; ordinary users find value in image and chat generators and will keep using them whether or not experts approve. The practical takeaway for the AI/ML community is that integration, not purity, is the battleground. Browsers can either cede users to opaque “anti-browser” AI apps or offer safer, privacy-preserving alternatives. Key technical and product moves recommended: ship a global “disable AI features” toggle (and be transparent about maintenance costs), market Firefox as an AI-aware privacy browser, and enable local LLMs or extension-based models so users can run models client-side. These choices implicate security, privacy, and UX design trade-offs (e.g., sandboxing, data flows, model telemetry), and they reshape how we mitigate harms—by building and shipping better defaults rather than simply criticizing current platforms.