Aran Komatsuzaki
RT @kalpeshk2011: To detect text written by LMs like #ChatGPT, many methods have recently emerged: DetectGPT, watermarks, GPTZero.

We present a paraphrasing attack that can drop their detection rates to <10%.

To defend against it, we propose detection with retrieval.
https://t.co/OrVzmVIKjI
🧵👇 https://t.co/HgwBxPToka
tweet

Stable Diffusion 🎨 AI Art
RT @DiffusionPics: Reversion : Textual Embeddings for Relations Between Objects #AIArt

#StableDiffusion2 / #StableDiffusion https://t.co/RjTHZvE8nv
tweet

Yann LeCun
Debate at 5:30 EST today:
“Do large language models need sensory grounding for meaning and understanding?”

Debaters:
Jacob Browning, David Chalmers, Brenden Lake, Yann LeCun, Gary Lupyan, Ellie Pavlick.

Livestream: https://t.co/90ZCqIEEYk
Event: https://t.co/zxDVr3m1Bc
tweet

Dave Craige
RT @BusinessMaven1: #HustleGPT Day 3 🤨 Website building mainly… after yesterdays ‘breakup’ BUT onwards and upwards! I’ll be updating YouTube with videos of my progress 😅 Stay tuned to see is A.I will make me rich or bankrupt me 💰🤖
tweet

Carlos E. Perez
AI is the medium, and the medium is the message. If you don't grok this, then you won't be groking anything moving forward.
tweet

Earth Is A Sales Funnel For SATAN
another normal day online https://t.co/WFiSv1Pq1H
tweet

KaliYuga
I know the description is garbled but https://t.co/SrDt4ybPez
tweet

Robert Scoble
RT @d_feldman: Theory of mind? https://t.co/boe9fc2kjY
tweet

Robert Scoble
RT @ivanrouzanov: Sad day: Gordon Moore, Intel Co-Founder, Dies at 94.

https://t.co/Mb8xFmDhzf
tweet

Emad
I don't know why there hasn't been more mainstream coverage of @midjourney v5.

It's like crazy amazing and groundbreaking.

🤖🎨🖌️ https://t.co/WLXDGJZB2q
tweet

Project Veritas
RT @BrazilianMario: Woah 👀

@NPR https://t.co/YAr4mgXIiI
tweet