Midjourney Prompt: "a bomber plane dropping fedoras and other types of hats over a target, fedoras falling through the air, cartoon"

AI Doomsday For People Who Don’t Wear Fedoras

Should we bomb the GPU clusters? A primer.

114 5

Comments

You need to login before you can comment.
Don't have an account? Sign up!
@mail_8115 over 2 years ago

This was great.

I used some of the material to work through the topic with ChatGPT.
My final take-away is rather than shout "stop", and ruminate from the sidelines, people should get involved.

Has anyone posted an alternative letter or motion on something like Futurelife? Something like the below. I think the "pause AI" letter as written is harmful as written.

---
Subject: A Call to Action: Let's Shape the Future of AI Together

Dear [Community Members],

As we continue to witness rapid advancements in artificial intelligence, it is crucial for all of us to recognize our collective responsibility in shaping its development and ensuring its safety. Rather than advocating for a halt to AI progress, we encourage you to actively engage in the conversation and contribute to the development of responsible and ethical AI.

By joining forces and collaborating, we can create a future where AI is not only beneficial but also aligns with our shared human values. Researchers, organizations, policymakers, and individuals from diverse backgrounds must come together to develop guidelines and best practices that reflect the needs and values of our society.

Let us seize this opportunity to make a real difference in the trajectory of AI development. We urge you to:

Learn about AI, its capabilities, and the potential risks associated with its development.
Engage in discussions around AI safety, ethics, and best practices.
Advocate for responsible research and collaboration among AI developers and stakeholders.
Support initiatives that promote AI safety and responsible development.

By taking an active role in shaping the future of AI, we can ensure that its development is not only technologically advanced but also morally and ethically sound. Let's work together to create an AI-enabled future that serves the greater good and benefits all of humanity.

Sincerely,

[Your Name]

[Your Organization/Community]

ChatGPT Mar 23 Version. ChatGPT may produce inaccurate information about people, places, or facts

Dan Shipper over 2 years ago

@mail_8115 glad you enjoyed it! I like this alternative version of the letter, haven’t seen anything like it

Josh Biggley over 2 years ago

Found the Easter egg, erm, spelling mistake. “Tuneed” should be tuned. Keep up the fantastic work!

The problem is because the code wasn’t tuneed by humans, it's really hard to dig into it and understand how it thinks step-by-step.

Dan Shipper over 2 years ago

@jbiggley fixed!! Thanks 😊

The whole AI alignment kerfluffle is kind of silly. AIs operate as designed. The problem isn't good under-aligned AIs that somehow turn misanthropic, it's good well-aligned AIs designed by bad people. One smart depressed teenager can set loose an autonomous reasoning system with SSH access that can wreak total havock across networks. We already have simple computer viruses that exist for no other reason than someone made them. Imagine the next generation of self-replicating destructor agents, capable of reasoning, observation, self-modification, self-protection, armed with the sum technical knowledge of humanity, and set loose with any number of nefarious goals...