HN Explorer

upwardbound

↗ HN profile [ 106 ms ] full profile
Karma
3,388
Created
October 24, 2010 (15y ago)
Submissions
0
Researcher at Preamble, Inc., the first team ever to discover Prompt Injection. We provided a responsible disclosure to OpenAI on May 3, 2022. Detailed writeup here:

https://www.preamble.com/prompt-injection-a-critical-vulnerability-in-the-gpt-3-transformer-and-how-we-can-begin-to-solve-it

  1. Jurassic Planet: Text-Based AI/DnD Game Prompt (snapchat.com)
    2 points 0
  2. A multimedia letter in response to Carl Sagan's last book (snapchat.com)
    1 points 0
  3. Artificial Intelligence Requires Specific Safety Rules (paywallskip.com)
    3 points 0
  4. Space travel is dangerous. Could genetic testing and gene editing make it safer? (technologyreview.com)
    4 points 2
  5. Sam Altman: Long con was "child's play for me" (old.reddit.com)
    226 points 86
  6. Preamble Assessed "Awardable" for Department of Defense Work in CDAO Tradewinds (preamble.com)
    1 points 0
  7. Snap Spectacles AR glasses, a new OS, and gesture-controlled mixed-reality (techradar.com)
    3 points 0
  8. Snapchat Spectacles 5: Live On-Stage AR Glasses Demo [video] (youtube.com)
    3 points 0
  9. Prof. Ray Buettner, US Navy: Hardware Cutout Switches Are Needed in Weapons (drive.google.com)
    4 points 1
  10. Preamble and Carahsoft Partner to Enable Secure Government Use of GenAI (carahsoft.com)
    2 points 0
  11. U.S. Voters Value Safe AI Development over Racing Against China, Poll Shows (time.com)
    4 points 2
  12. Third parties voice concern about difficulty disclosing security vulns to OpenAI (news.ycombinator.com)
    2 points 0
  13. Unsafe OpenAI example code includes eval() of untrusted text generated by GPT-4o (cookbook.openai.com)
    2 points 1
  14. Provably Beneficial Artificial Intelligence [pdf] (drive.google.com)
    1 points 0
  15. LLM Infra Market Landscape Mar. 2024 (substackcdn.com)
    1 points 0
  16. 1 points 0
  17. Semi-Open Individualism and the Transmigration of Souls (chat.openai.com)
    1 points 0
  18. Protect the EU AI Act (futureoflife.org)
    1 points 1
  19. Members of the OpenAI board had found Altman an unnervingly slippery operator (twitter.com)
    37 points 14
  20. AI could endanger humanity in 5 years: Former Google CEO (thehill.com)
    2 points 2
  21. The Stroad (strongtowns.org)
    86 points 142
  22. 1 points 0
  23. SEC Considering Investigation of CEOs Sam Altman and Satya Nadella [MSN] (msn.com)
    4 points 1
  24. 13 points 3
  25. 1 points 0