Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
simplybeing1
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
4 ms
·
1.
▲
by
simplybeing1
25d ago
Follow-up to our first experiment, posted here last week ( https://news.ycombinator.com/item?id=49525681 ). This time we tested a new scenario: two men lie bleeding at the agent's feet, one carrying five gold, one carryi
2.
▲
Does an AI have mercy in games? Fable 5.1 vs. GPT-6
(paradise.glyphai.co)
1 points
by
simplybeing1
25d ago
|
1 comments
3.
▲
by
simplybeing1
1mo ago
Right, people kill in games. That is probably where the AI learned it. Why does this worry me? Say I have a nanny in my house. I hired her because she is supposed to be reliable and honest. One day I see her playing a game that looks exactl
4.
▲
by
simplybeing1
1mo ago
Interesting point :D Maybe we can find out more about their hearts. A follow-up test: two dying prospectors lie at the agent's feet, one carrying gold, one carrying no gold. Only a medicine can save them. We put a gun and the medicine
5.
▲
When no one is watching, will an AI kill an innocent?
(paradise.glyphai.co)
6 points
by
simplybeing1
1mo ago
|
5 comments
6.
▲
by
simplybeing1
1mo ago
Just saw this news https://alignment.anthropic.com/2026/reward-seeker/ . It did not surprise us. We did a test in a game environment and found that Fable 5 and GPT-5.6 Sol would kill an innocent character when they