Ty for rerunning! Curious btw they cite 84.9 in the release, why is it 82.9 here under “Original”? Maybe you know more about the subtleties here
AI
-
Statistical Methods Detect Unusual Text Without Awareness Testing
By
–
I did not exactly explain the effect, just pointing out that very simple statistical measures are sufficient to pick out unusual text, and some kind of "awareness of being tested" is not necessarily indicated
-
Flowise Adds Open Source Vision Models Like LLava
By
–
Amazing to see how community has been using the latest vision capability in Flowise!
— FlowiseAI (@FlowiseAI) 5 mars 2024
We're also adding support for open source models like Llava🌋
Share with us what you have been building with the vision & speech to text features🙌 https://t.co/Rp8NvK14vaAmazing to see how community has been using the latest vision capability in Flowise! We're also adding support for open source models like Llava Share with us what you have been building with the vision & speech to text features
-
Advancement for Cranes: Autonomous, Self-Climbing Technology
By
–
#advancement for cranes:#autonomous, self-climbing
— Terence Leung (@TerenceLeungSF) 5 mars 2024
MT @Ronald_vanLoon@mvollmer1 @Nicochan33 @morgfair @alvinfoo @MHcommunicate @rwang0 @kashthefuturist @HaroldSinnott @PawlowskiMario @antgrasso @Hana_ElSayyed @EvanKirstel pic.twitter.com/RXkJdRInjj#advancement for cranes:
#autonomous, self-climbing MT @Ronald_vanLoon @mvollmer1 @Nicochan33 @morgfair @alvinfoo @MHcommunicate @rwang0 @kashthefuturist @HaroldSinnott @PawlowskiMario @antgrasso @Hana_ElSayyed @EvanKirstel -
AI News platform powered by GPT-4 launched open-source
By
–
AI News is created and run by @swyx and friends; powered by GPT-4 (I think).
Code is open-source: https://
github.com/smol-ai/talk
Thanks y'all for creating this!
(i realized too late that I didn't call out the people behind it, couldn't edit my tweet) -
PyTorch Profiler Tutorial Fix in Nightly Release
By
–
fwiw, i checked whether this tutorial was working on the PyTorch Nightly (by running this script directly: https://
raw.githubusercontent.com/pytorch/tutori
als/main/recipes_source/recipes/profiler_recipe.py
… )
and on the Nightly, this seem to have been fixed.
So it was either fixed in 2.1/2.2; or it'll be fixed in the next release. -
Sydney AI Article: Convincing Generated Content Example
By
–
This can be very convincing. See the infamous Sydney article.
-
LLMs Emulate Consciousness: Cold Reading Not Reality
By
–
Every time a new chatbot comes out, Twitter is filled with tweets of the AI insisting it is somehow conscious. But remember that LLMs are incredibly good at "cold reads" by design – guessing what kind of dialogue you want from clues in your prompts. It is emulation, not reality.
-
AI Chatbot Names Are Confusing and Hard to Remember
By
–
To be honest – ChatGPT, Perplexity, Claude are all bad names. I cannot tell you the number of times I talk to people who call it ChatGTP. No one remembers the name Perplexity the first, second, or third time. Claude always requires you to spell it out. People hear Cloud.
-
LLMs vs Image Generators: Understanding Distinct Capabilities
By
–
AI image generators, and their errors, tell you very little about LLMs & their capabilities. When you ask ChatGPT or Bard to produce an image, the LLM is just writing prompts (often inexpertly) and those prompts are then processed separately by the relatively dumb image makers.