Meta: Shut down your invasive AI Discover feed

Original Article ↗ Hacker News Discussion ↗

Lack of clarity in Mozilla’s campaign

  • Many commenters found Mozilla’s petition confusing and under-explained.
  • Complaints: no screenshots, flow diagrams, or concrete examples; assumes prior knowledge of Meta’s AI app; feels like engagement-bait more than an informative explainer.
  • Several people said the submission link should have been to the investigative articles that actually describe the feature.

What Meta’s AI Discover feed appears to do

  • Context from linked reporting: Meta’s standalone AI app has a “Discover” tab showing other users’ AI chats.
  • People testing the app describe the flow as:
    • Chat screen has a “Share” button.
    • Tapping “Share” opens a preview with a prominent “Post” button.
    • Tapping “Post” makes the chat public and surfaces it in Discover; a link can then be shared.
  • Some evidence from users: Discover shows clearly unintended content (e.g., “note to self” about canceling insurance, stylized baby photos with originals attached).

Dark patterns vs user responsibility

  • One side:
    • Chats are private by default; you must tap Share → Post.
    • UI clearly shows you’re “posting”; Mozilla’s language about “quietly turning private chats public” is misleading or false.
  • Other side:
    • “Share” usually means “let me choose where/who to share with,” not “publish to a public feed by default.”
    • Making public posting the only way to share, and calling it “Share,” is a dark pattern, especially for non-technical users.
    • Meta’s long history of nudging oversharing and testing many UI variants makes people distrust that this is merely user error.

Leaving Meta vs network lock-in

  • Some argue the only real solution is to stop using Meta entirely.
  • Others say that’s unrealistic due to network effects, especially where WhatsApp is the de facto communication channel (often zero-rated and used for everything from school to business).
  • A few report successfully quitting Facebook/Instagram with minimal impact; others describe real social or practical costs, particularly outside the US.

Views on Mozilla’s credibility and strategy

  • Mixed reactions:
    • Some are glad Mozilla is still pushing on privacy issues.
    • Others say this specific campaign is sensationalist, poorly communicated, and undermines trust.
  • Prior incidents (terms-of-use changes around data, bundled addons/telemetry, perceived Google dependency) are cited as reasons to doubt Mozilla’s moral high ground.
  • Several call for Mozilla to focus on making Firefox and its web tech stronger instead of vague activism.

Broader attitudes to privacy and platforms

  • Strong baseline distrust of Meta/Facebook: many say you should assume anything you give them can become public or be exploited.
  • Others push back against fatalism, arguing that “nothing is private anyway” is how privacy norms get destroyed.