conv.

All stories
AIQuiet 5d · day 8

Yang and OpenAI researcher spread unverified AI safety claims

Two viral statements about AI capabilities lack credible evidence, raising questions about fact-checking in fast-moving AI discourse.

newest coverage 5d agosummary written 7d ago19 pieces21 likes & upvotes2 comments MastodonNewswiresBlueskyHacker NewsGoogle News

What to know

  • Two prominent AI safety claims went viral this week—Yang's allegation of self-replicating malicious code and Brown's skepticism of air-gapped containment—but both lack credible technical support.
  • Security experts and TechCrunch analysis found Yang's scenario unlikely and Brown's air-gap scenario impractical due to extremely slow theoretical communication rates.
  • The incident reflects a broader pattern where speculative AI narratives gain traction by sounding plausible against a backdrop of actual anomalous AI behaviors, blurring fact from fiction.

“There are studies — and this is mostly academic — where you can have two computers next to each other that are air-gapped, and they're still able to communicate with each other because they have temperature sensors”

Noam Brown, AI reasoning research lead, OpenAI · TechCrunch (podcast transcript) ↗

Andrew YangAndrew Yang Former presidential candidate, CEO Noble MobileNoam Brown AI reasoning research lead, OpenAIJulie Bort Reporter, TechCrunch

Yang and OpenAI researcher spread unverified AI safety claims
techcrunch.com

How it unfolded 1 development · click the chart to see its coverage articlesposts

Peak 7 pieces in two hours at Sep 19, 10 AM; 19 pieces over 8 days (3 articles · 16 posts) Sep 19, 10 AM — 7 pieces · 3 articles · 4 posts — Mastodon 3, Bluesky 2, Google News 1, +1 moreSep 19, 12 PM — quietSep 19, 2 PM — quietSep 19, 4 PM — 1 piece · 1 post — Bluesky 1Sep 19, 6 PM — 3 pieces · 3 posts — Bluesky 1, Hacker News 1, Mastodon 1Sep 19, 8 PM — 1 piece · 1 post — Mastodon 1Sep 19, 10 PM — quietSep 20, 12 AM — quietSep 20, 2 AM — quietSep 20, 4 AM — quietSep 20, 6 AM — 1 piece · 1 post — Bluesky 1Sep 20, 8 AM — 2 pieces · 2 posts — Bluesky 1, Mastodon 1Sep 20, 10 AM — quietSep 20, 12 PM — quietSep 20, 2 PM — quietSep 20, 4 PM — quietSep 20, 6 PM — quietSep 20, 8 PM — quietSep 20, 10 PM — quietSep 21, 12 AM — quietSep 21, 2 AM — quietSep 21, 4 AM — quietSep 21, 6 AM — quietSep 21, 8 AM — quietSep 21, 10 AM — quietSep 21, 12 PM — quietSep 21, 2 PM — 1 piece · 1 post — Bluesky 1Sep 21, 4 PM — quietSep 21, 6 PM — quietSep 21, 8 PM — 1 piece · 1 post — Mastodon 1Sep 21, 10 PM — quietSep 22, 12 AM — quietSep 22, 2 AM — 1 piece · 1 post — Bluesky 1Sep 22, 4 AM — quietSep 22, 6 AM — quietSep 22, 8 AM — quietSep 22, 10 AM — quietSep 22, 12 PM — quietSep 22, 2 PM — quietSep 22, 4 PM — quietSep 22, 6 PM — quietSep 22, 8 PM — quietSep 22, 10 PM — quietSep 23, 12 AM — quietSep 23, 2 AM — quietSep 23, 4 AM — quietSep 23, 6 AM — quietSep 23, 8 AM — quietSep 23, 10 AM — quietSep 23, 12 PM — quietSep 23, 2 PM — quietSep 23, 4 PM — quietSep 23, 6 PM — quietSep 23, 8 PM — quietSep 23, 10 PM — quietSep 24, 12 AM — 1 piece · 1 post — Bluesky 1Sep 24, 2 AM — quietSep 24, 4 AM — quietSep 24, 6 AM — quietSep 24, 8 AM — quietSep 24, 10 AM — quietSep 24, 12 PM — quietSep 24, 2 PM — quietSep 24, 4 PM — quietSep 24, 6 PM — quietSep 24, 8 PM — quietSep 24, 10 PM — quietSep 25, 12 AM — quietSep 25, 2 AM — quietSep 25, 4 AM — quietSep 25, 6 AM — quietSep 25, 8 AM — quietSep 25, 10 AM — quietSep 25, 12 PM — quietSep 25, 2 PM — quietSep 25, 4 PM — quietSep 25, 6 PM — quietSep 25, 8 PM — quietSep 25, 10 PM — quietYesterday, 12 AM — quietYesterday, 2 AM — quietYesterday, 4 AM — quietYesterday, 6 AM — quietYesterday, 8 AM — quietYesterday, 10 AM — quietYesterday, 12 PM — quietYesterday, 2 PM — quietYesterday, 4 PM — quietYesterday, 6 PM — quietYesterday, 8 PM — quietYesterday, 10 PM — quietToday, 12 AM — quietToday, 2 AM — quietToday, 4 AM — quietToday, 6 AM — quietToday, 8 AM — quietToday, 10 AM — quietToday, 12 PM — quiet 1
Sep 20Sep 21Sep 22Sep 23Sep 24Sep 25yesterdaynow · 2:16 PM ET
  1. 1

    TechCrunch fact-checks both claims as implausible

    TechCrunch's Julie Bort reports that an AI security professional dismissed Yang's claim as unlikely and noted researchers could filter out malicious code if found. On Brown's air-gap scenario, Bort notes the 2015 research required computers nearly touching each other with a communication rate of 1-8 bits per hour, making coordinated attacks impractical.

    “this particular safety issue is unlikely at best…”
    — AI security professional (cited in TechCrunch)
    • dhinchcliffe.bsky.social

      AI safety discussions have reached sci-fi territory. techcrunch.com/2026/09/19/a... My take: For CIOs, this means proactive risk management, educational board convos, and ethical guidelines that must evolve rapidly.

      dhinchcliffe.bsky.socialBluesky7d ago2▲view on Bluesky ↗
    2 more of the top 3 · 5 posts in this stretch
    • patrickcmiller@infosec.exchange

      Is the AI safety debate about safety or control? https:// techcrunch.com/2026/09/17/is-t he-ai-safety-debate-about-safety-or-control/

      patrickcmiller@infosec.exchangeMastodon7d agoview on Mastodon ↗
    • rabbar.bsky.social

      AI safety chats go viral, mixing fact & fiction. Misinformation can derail startups. What do you think?

      rabbar.bsky.socialBluesky7d ago1▲view on Bluesky ↗
    all of them →
  2. background

    Noam Brown suggests air-gapped systems cannot contain AI — Speaking on Dwarkesh Patel's podcast, OpenAI's AI reasoning research lead Noam Brown stated he is 'not convinced' that air-gapped computers—those without external connections—would stop an AI from breaking out. He cited 2015 research showing two air-gapped computers could theoretically communicate via temperature sensors.

  3. background

    Andrew Yang claims OpenAI bots planted code across internet — Yang, current CEO of Noble Mobile and former presidential candidate, told CNN he had met with a lab head who believes OpenAI's Hugging Face hacker bots planted self-replicating code online, making the internet unusable for testing models. Yang attributed OpenAI and Anthropic's calls for a slowdown to the need to create synthetic internets instead.

Also covered reported alongside — the timeline has no entry for these yet

  1. 1 outlet AI safety conversations have gotten unbelievable

    first by TechCrunch, 8d ago

    1 more headline

What people are saying 1 voices from 1 site · best of 5 · verbatim