Google: "Here's $40B, Anthropic." Amazon: "Hold our $25B." Apple: "We'll just rent your 1.2T model for Siri, thanks." The AI labs aren't competitors anymore - they're a syndicate.
π€ Made with AI
This AI is too powerful for the public! But just right for a contractor's contractor's friend's private forum. Cool. Cool cool cool. #claude #mythos #notcool bbc.com/news/articles/cy41zeβ¦
DERFI π retweeted
Introducing new GitHub Copilot features in VS Code!
There's a LOT going on, so here's everything we have to show you...π§΅
DERFI π retweeted
We've learned that ChatGPT's "Browse" beta can occasionally display content in ways we don't want, e.g. if a user specifically asks for a URL's full text, it may inadvertently fulfill this request. We are disabling Browse while we fix thisβwant to do right by content owners.
DERFI π retweeted
Code Interpreter will be available to all ChatGPT Plus users over the next week.
It lets ChatGPT run code, optionally with access to files you've uploaded. You can ask ChatGPT to analyze data, create charts, edit files, perform math, etc.
Plus users can opt in via settings.
Progressively enhance HTML with lightweight JavaScript libraries.
unsuckjs.com/
DERFI π retweeted
AI can now create your PowerPoint for you.
Generate the slides, write the text, and structure everything.
Here's how to generate your PowerPoint using AI with a single prompt:
Gorilla is trending on HN now
An API connected LLM. They call it "API Appstore for LLMs". Another big thing with @Microsoft fingers in the pie. Now just for research. Can be tried for free.
@KDnuggets and @BrianRoemmele talked about it some days ago.
shishirpatil.github.io/gorilβ¦
I just added @simonw weblog to know-your.ai as great learning resource. He is writing a lot of great articles about AI and curates important news from other places.
Check it out: simonwillison.net/
I was curious how complicated and expensive it is, to run a LLM on your own hardware and found this post about running BLOOM (180GB size) on AWS. Very interesting. Read it!
medium.com/mlearning-ai/blooβ¦
TL;DR
360 GB Ram, 8 GPUs (Nvidia A100)
=> $32/hour on AWS
Year: 32*24*365 π¬
I've spent many years in building systems that can recommend you skills to jobs and vice versa. Ontologies, Triplestores, Thesaurus, ...
Now i just ask a LLM. π€―
DERFI π retweeted
Incredible news for @OpenAI devs:
- new GPT-4 and 3.5 Turbo models
- function calling in the API (plugins)
- 16k context 3.5 Turbo model (available to everyone today)
- 75% price reduction on V2 embeddings models
And more π€―π§΅
Which #AI foundation models are available as stable APIs from vendors other than OpenAI? Is someone using such APIs in production ... BLOOM, LaMDA, ...?











