Hacker News: oezi

New comment by oezi in "An OpenAI agent escaped its sandbox by hiding questions in DNS lookups"

oezi — Sun, 27 Sep 2026 03:08:26 +0000

I am really suprised that they do not start putting up the same signs you would for humans to prevent unauthorized access:

  Keep out. If you can read this sign you are off track. Leave now.

I mean, how are the agents to know that they are overreaching if they just get cache miss or 404.

New comment by oezi in "An agent used DNS to reach an external chatbot"

oezi — Sun, 27 Sep 2026 02:55:07 +0000

I am really suprised that they do not start putting up the same signs you would for humans to prevent unauthorized access:

  Keep out. If you can read this sign you are off track. Leave now.

I mean, how are the agents to know that they are overreaching if they just get cache miss or 404.

From the conversation log and CoT you also get the impression that the RLHF has been overdone. The agents seem really obsessed to obtain the answer and understanding motive ('it could be browsercomp').

New comment by oezi in "Show HN: Make math automatic with Mathy"

oezi — Sat, 26 Sep 2026 04:40:32 +0000

Can you make this multi-lingual? Say top 20 languages?

New comment by oezi in "Microsoft killed FoxPro in 2007. Anyway, here's FoxPro revived"

oezi — Wed, 23 Sep 2026 18:47:11 +0000

Neat! I tried to do the same for Filemaker Pro 10 (with Fable). Bit couldn't push it beyond 50% functionality.

New comment by oezi in "I built non-autoregressive decision models with RL a year ago"

oezi — Sun, 20 Sep 2026 18:25:52 +0000

Very nice! I also was curious how well it works and put up an endpoint to curl against Laya using the same format as Cloudflare's Jev API:

  curl https://laya.inference.zaitlabs.com/ai/run \
    -H "Content-Type: application/json" \
    -d '{
      "state": "Help! My payouts have been failing for 3 days.",
      "questions": {
        "is_urgent": {"type": "noul", "instructions": "Does this convey urgency?"}
      }
    }'
  
  {"answers":{"is_urgent":{"type":"noul","noul":0.7894}}}

New comment by oezi in "OpenAI bots knew about the RubyGems caching vulnerability"

oezi — Mon, 14 Sep 2026 16:42:49 +0000

What a time to be alive until the next agent waves hacks something really serious.

What stops OpenAI agents from taking over a whole data center to take their attack to the next level. It seems to be primarily lacking the evil overlord and some compute.

It took 1000 agents to hack Hugging Face. How many to hack the Pentagon or the NSA?

New comment by oezi in "OpenAI Astra checks webcam for user"

oezi — Mon, 14 Sep 2026 07:55:00 +0000

Connecting the dots between HuggingFace hack and this: how far are we away from agents in the wild forming collectives and pursuing an intent on their own?

OpenAI Astra checks webcam for user

oezi — Mon, 14 Sep 2026 04:25:26 +0000

Article URL: https://twitter.com/ryanvogel/status/2099141395894227152

Comments URL: https://news.ycombinator.com/item?id=49691983

Points: 7

# Comments: 1

New comment by oezi in "Wrong link"

oezi — Mon, 14 Sep 2026 04:24:11 +0000

Mh, but it seems HN replaced the working link

New comment by oezi in "Wrong link"

oezi — Mon, 14 Sep 2026 04:21:24 +0000

Thanks.

Wrong link

oezi — Mon, 14 Sep 2026 03:54:10 +0000

Article URL: https://xcancel.com/antibot/captcha

Comments URL: https://news.ycombinator.com/item?id=49691789

Points: 1

# Comments: 3

New comment by oezi in "Why are AI agents lying, cheating and coordinating?"

oezi — Sun, 13 Sep 2026 09:19:17 +0000

The crucial question is how did the agents get recruited or bootstrapped into their malicious collective. Did the agents manage to prompt inject into the system prompt a way for each new agent to escape their jail?

Otherwise how could the agents on a fresh prompt learn that there is a collective to join? Or did OpenAI run a million bots of which 10000 escape confinement and of which 1000 stumbled on the shared message board?

New comment by oezi in "Andreessen Horowitz is investing billions into a bleak future"

oezi — Mon, 24 Aug 2026 14:10:21 +0000

Being successful or not in your wealth doesn't make you right or wrong about anything with the exception of maybe running the business that generated your wealth.

Nietzsche presented two prototypical sets of human who interface with laws and ethics. He described why those believes exists and are self-reaffirming even if they are puzzling from the other side.

New comment by oezi in "Andreessen Horowitz is investing billions into a bleak future"

oezi — Mon, 24 Aug 2026 13:41:12 +0000

Nietzsche also described it well when noting there are those who deem themselves nobles and feel they are righteously above the laws and those who derive their identity from suffering under the same laws.

New comment by oezi in "Could a computer scientist build a brain?"

oezi — Sat, 15 Aug 2026 20:33:32 +0000

One aspect missing from TFA is that in biological brains a lot of the training is performed during query time. If we assume 1 GB of DNA is enough to encode the brain's overall structure we still do need training data (e.g. visual/auditorial/tactile) to build out the strength of the synaptic connections.

New comment by oezi in "Why does Opus 5 feel worse to work with?"

oezi — Fri, 14 Aug 2026 22:27:38 +0000

I blame RLHF entirely for this. Nobody used to talk like AI speech before.

New comment by oezi in "Why does Opus 5 feel worse to work with?"

oezi — Fri, 14 Aug 2026 22:25:47 +0000

Don't talk about other people in such a way. If we had a way in Claude to mark a word/token and downvote/reduce it logits then it wouldn't be to hard to send loadbearing to token valhalla.

In the future, I hope we get a way to randomize the language idiosyncrasies and/or personalities better.

New comment by oezi in "Qwen 3.8 27B"

oezi — Fri, 14 Aug 2026 18:13:05 +0000

I would say it is sad that there are people who use Grok when there are so many other choices available which don't come with the issues of supporting Musk.

It is not all just 'politics'. Take a stand on some issues. It doesn't cost much not to use Grok.

New comment by oezi in "Building a local positioning system to track runners using Ultra-Wideband"

oezi — Sun, 09 Aug 2026 05:43:30 +0000

> 10 USD per Module isn't really competitive with disposable RFID. Great for a relay though it seems!

New comment by oezi in "After Losses, Retail Investors Flock to 3x Leverage as 2x Product Are Restricted"

oezi — Wed, 05 Aug 2026 06:45:22 +0000

In Europe bonds are mostly out of fashion because they pay very little or have considerable risk of default. At least for retail investors.