@aminuyaro_@spectatorindex inflation affects the value of money. It means you can buy less with the same amount of money. So anything money can buy is affected too.
mf trained a simple model to translate ada-002 embeddings back to text and found something interesting:
sentence embeddings have arithmetic properties
the classic "king - man + woman = queen" still works just like w/ token embeddings — but can go a few steps beyond that as well
A small, extremely simple automaton that reproduces a boat in the waves of the sea
[📹 https://t.co/2CTgaNHYNd: https://t.co/P2kFEBURHf]
https://t.co/SYefxnDpYt
@mezaoptimizer I agree.
We should also remember that “important” is a relative concept. When determining whether we are in an important century we should ask:
Important with respect to what exactly?
@mezaoptimizer Why exactly would a digital AI get a vote? on what basis?
You're using the word "mind" loosely here, equating "digital" minds to human minds. I think that's a fundamental flaw in your logic.
Although I do agree AI will be a risk to democracy, it's not due to voting.
Why is @Apple iPhone syncing so dysfunctional?
I just want to copy my own music to the phone.
No cloud services, no backups, no additional features.
Just: rsync -az <from-laptop> <to-iphone>
@GeologicalSETI great idea about the database of experiences but how to filter out hoaxters and false accounts?
Also how to collect the stories from experiencers? Public website/form? Or a more private method?
A survey of LLMs with a practical guide and evolutionary tree.
Number of LLMs from Meta = 7
Number of open source LLMs from Meta = 7
The architecture nomenclature for LLMs is somewhat confusing and unfortunate.
What's called "encoder only" actually has an encoder and a decoder (just not an auto-regressive decoder).
What's called "encoder-decoder" really means "encoder with auto-regressive decoder"
What's called "decoder only" really means "auto-regressive encoder-decoder"
https://t.co/dtEFW3UU6S
@mezaoptimizer This assumes a digital “mind” can be conscious. There is no question of “moral worth” if that’s not possible. Personally I don’t think it is.
@steventey@vercel I’ve also used vercel for my latest AI project and I’m really digging it!
It’s like a generative wikipedia based on #ChatGPT
https://t.co/WmcV5OGrI2
I’ve been developing an idea lately.
I call it gencarta.
It’s like a generative Wikipedia based on #ChatGPT. It generates an article, sub-sections, related topics, images and hyperlinks.
Looking for people to try it, and give me feedback.
Thoughts?
https://t.co/qND1lRaw0d
@BrianRoemmele This is great! Will you publish a “how to guide” or at least a tweet with high level steps needed? Or maybe I missed it if you shared that already
@algekalipso I love hearing your insights but sometimes the concepts and language you use is outside of my understanding. Have you thought of condensing your thoughts in ways that a layman might find it easier to understand?