Martineski

joined 1 year ago
MODERATOR OF
[–] Martineski@lemmy.fmhy.ml 7 points 1 year ago* (last edited 1 year ago) (3 children)

I'm not going to fact check this number and I will just edit the title with it.

[–] Martineski@lemmy.fmhy.ml 2 points 1 year ago

It looks so pretty.

[–] Martineski@lemmy.fmhy.ml 5 points 1 year ago

There should be a button for it as you can see in the screesnhot that the other person in the comments sent.

[–] Martineski@lemmy.fmhy.ml 3 points 1 year ago* (last edited 1 year ago)

I mean it's not that much for people interested in the topic. And we will learn a lot about capabilities of it.

[–] Martineski@lemmy.fmhy.ml 9 points 1 year ago

Lmfao, what about the other creativity levels?

[–] Martineski@lemmy.fmhy.ml 8 points 1 year ago* (last edited 1 year ago) (6 children)

Does anyone hare have access to image sending? I would like you to test those images out:

1.) First try out the one sent by OP

Try prompting each image 3 times to check the consistency (each prompt at different creativity?) and each time you send an image make a new conversation to avoid it getting info about the images in the context window. Thank you to whover does that.

[–] Martineski@lemmy.fmhy.ml 5 points 1 year ago

From what I see the image is from 2018 so probably.

[–] Martineski@lemmy.fmhy.ml 1 points 1 year ago (1 children)

Oooh, ty for the context.

 
 
 

Superintelligence when? This question is more urgent than ever as we hear competing timelines from Inflection AI, OpenAI and the leaders at the Center for AI Safety. This video not only covers the what was said, it offers some data on what superintelligence is projected to be capable of.

I discuss things that can hasten those timelines (see the new Netflix doc, Top 1% for Creativity and the AI Natural Selection paper) or slow it down (ft. Yuval Noah Hariri, the new Jailbroken paper, and more). And I end on some reflections of what it might mean to interact with a superintelligence (ft Douglas Hofstadter).

 
 
 

I explained the situation to the poster where I told him that antiwork posts take too much space on the front page of our instance. Posts from antiwork shouldn't now take too much space unless more people start posting at which point it will be impossible to slow down sublemmy and it will settle down on the front page permamently lol.

I just wanted to clarify that before I get any complaints. Have a great day guys.

 
 
 

The ⁠alignment-minetest project just put out a new blog post detailing what we've been working on for the past several months.

The post is titled "Minetester: A fully open RL environment built on Minetest" and covers:

  • The minetester framework and how it relates to existing efforts in ai for Minecraft

  • A PPO baseline and environment customization using the framework.

  • Basic interpretability work we did on the learned PPO policy.

  • General and specific takeaways from our work so far.

  • Next steps for the project.

Relevant Links:

Blog Post: https://blog.eleuther.ai/minetester-intro/

Minetest Channel: https://discord.com/channels/729741769192767510/1014999314835181650

Minetester Repo: https://github.com/EleutherAI/minetest/

Minetester Baselines: https://github.com/EleutherAI/minetest-baselines/

Interpretabilty Notebook: https://github.com/EleutherAI/minetest-interpretabilty-notebook

Special thanks to @rkla @nev and @Eduuu for their major contributions to the project. Additional thanks to @josiah_wi and @Delta for their contributions.


Martineski: I got the announcement on this dc server: https://discord.gg/AY2Hg2qj

Direct link to the announcement: https://discord.com/channels/729741769192767510/794042109048651818/1127288927888343084

 
 
view more: ‹ prev next ›