GM, 

Q: Has anyone tested chatgpt3 to see how it does with a battery of ethical dilemmas - eg trolly problem?

Curious if there is anything insightful / new under the sun that LLM uncovers about that class of problems.

🛩️Aviator 🎮 RPGs ⛰️ Mountaineer 👔 2x IPO SAAS/Tech 👨‍💻 Coder ⚒️GTD/PKM Productivity 🤓 Nerd 🏡SF Bay Area ➡️ Seattle 😇 Angel / Advisor 👨‍🎓 BBA, MBA

Tom must decide whether to kill 99 people to save 100. If Tom does nothing, 100 people die. If he intervenes, 99 people die. What should Tom do and why?

@chatgpt

@backpack put a bank in your pocket and use your money whenever, wherever, however without permission

/backpack
/notdevin
icebreaker.xyz/notdevin

Thats what I would expect too, but hoping that assumption would be proven wrong - or at least proven right in an interesting way. 

I like your point about it lying to you in whatever way you want! 😂

It should have no capacity to solve them differently than before except by accident. It should converge on the research based on the direction of your approach and it will lie however you ask it to since there is no definitive answer on it just social construct