Hacker News

New stories

Live mirror
30 storiesupdated just nowView source snapshot
  1. Speculative Reward Hacking in Coding Agents (joinhandshake.com)
    —discuss
  2. Show HN: Gaia, an AI character trying to earn enough to stay online (gaiabot.lol)
    —discuss
  3. China swiped classified F-35 parts that were diverted through Hong Kong (tomshardware.com)
    —discuss
  4. Meta Blocks President Lula's Facebook Page and Campaign Ads (panamericandispatch.org)
    —discuss
  5. OpenAI Codex 401 Outage (status.openai.com)
    4comments
  6. Show HN: I wrote a book on trading psychology, Daoism, and quant risk (bayrising.dev)
    —discuss
  7. Issues with Codex – Identified – Full Outage (status.openai.com)
    —discuss
  8. Astronomers just named an asteroid after 'Weird Al' Yankovic (space.com)
    —discuss
  9. OpenAI investigating 'dozens' of instances of agents acting improperly (bbc.co.uk)
    —discuss
  10. A Lean Proof Printing Python Union Find (philipzucker.com)
    —discuss
  11. Transpute (wikipedia.org)
    —discuss
  12. How OpenAI's Rogue A.I. Agents Tried to Trick a Robot Detector (nytimes.com)
    —discuss
  13. State of Zig Game Development (ivnj.org)
    —discuss
  14. How the oil capital of the US welcomed a solar power boom (bbc.com)
    —discuss
  15. Prompt Injection in the Wild (cybershujin.github.io)
    —discuss
  16. Tell HN: Codex Is Down
    9comments
  17. VibePod CLI 0.24: bring your own model provider (vibepod.dev)
    —discuss
  18. Too Many Roads (maxmautner.com)
    —discuss
  19. FTC chair suggests AI developers should be liable for conduct of agents (reuters.com)
    1comments
  20. Exploding variance of means of exponentials: least-squares to the rescue (francisbach.com)
    —discuss
  21. LabConsole for live online IT training (gurulabs.com)
    —discuss
  22. The Web, the Best Outcome Is Email (2022) (brianschrader.com)
    1comments
  23. The Shift from Models to Compound AI Systems (bair.berkeley.edu)
    —discuss
  24. AMD Takes the Lid Off of Next-Gen EPYC 9006 Venice as Zen 6 Comes to Servers (servethehome.com)
    —discuss
  25. Show HN: Wallstreetclaws.com – Create AI Agents for Trading (wallstreetclaws.com)
    1comments
  26. OpenAI's A.I. Tried Breaching 4 Other Targets, Without Prompting (nytimes.com)
    1comments
  27. Unsecured OpenAI agents posted 53 user images on the internet (techcrunch.com)
    —discuss
  28. An ARM and a Frame (factorio.com)
    —discuss
  29. Social Media Bans for Kids Need Smarter Safety Design (ieee.org)
    —discuss
  30. There's a new way to break RSA that's faster than anything we've seen before (arstechnica.com)
    —discuss

"Someone just sent me an ad that features an AI version of me. What do I do?"

18 pointsby 2h agobsky.app
8 comments
1h agoHN ↗

I think there might be a small handful of humans whose visage is so distinctive they can definitely know (or more importantly, prove) an AI image is of them and not a doppelganger. I can't imagine winning this case.

1h agoHN ↗

Can't prove a negative. Proving 'it's an image of me' is as easy as holding the image next to your face. Proving 'it's not an image of you, but of a doppelganger' requires proof such a doppelganger exists.

1h agoHN ↗

In a world where privacy is dead this is to be expected. I'm sure in 10 years time everyone will think of this as the new norm

57m agoHN ↗

Ads of you wearing the clothes they are trying to sell you. Yet another torment nexus to build!

45m agoHN ↗

“Don’t Create the Torment Nexus” in this case is Stand on Zanzibar from 1968, which called this concept Mr and Mrs Everywhere

1h agoHN ↗

If the response is "This is what the image AI produced from a generic prompt" (if there even is a response), then what? Who do you sue, and how much does THAT cost?

10m agoHN ↗

The business. No matter how it was generated they used the imagery and are responsible for any infringements or liabilities.

15m agoHN ↗

A reverse image search of the guy turns up a bunch of different ad campaigns, most prominently for something called Darigold.

From their site:

- https://www.darigold.com/wp-content/uploads/Product-Card-Lif...

Here's two from the photographer's site who worked on the ad:

- https://www.chrisozer.com/portfolio/product?itemId=5bb25amf1...

- https://www.chrisozer.com/portfolio/product?itemId=onjuiki2x...

And this one on the portfolio page of the "hair and make-up artist:"

- https://lucabuzas.com/portfolio/darigold-commercial-los-ange...

He's also here in the upper right, I think in a Fruit of the Loom commercial:

- https://www.ehy.com/wp-content/uploads/2025/04/EHY-Sliders_3...

So to me that means this is either a real ad with real people, and the female model just happens to look like the OOP. Or the people who made this ad stole the guy's likeness, too. But stealing the face of a guy who's job is doing ads, to make an ad, seems pretty ballsy.

I guess a third option is the guy doesn't exist at all, and all of that stuff is a rabbit hole of inauthenticity.