1/ 🧵Have you guys heard of "StyleDrop" yet?

It lets you input ONE image as a style reference, and then prompt NEW images in that style.

(It's a Google Research paper, so you know they will wait until the LAST moment to give us access).

-thread- 2/ So here's a flower that is melting gold. Using this image as the style reference, you can start prompting LETTERS! Adobe Firefly, Google is looking at you 👀

1/9 📜On Friday, @HBCoop_ hosted a 3 Hour Twitter Space about latest developments in Generative AI and VR/AR (So far, 2,000 people tuned in!)

Speakers included:
@Scobleizer, @bilawalsidhu, @AnnGreenberg, @angrypenguinPNG

I used Claude-Instant-100K (not ChatGPT) to take notes. twitter.com/i/web/status/1… 2/🍏 The Apple AR/VR headset was a primary focus. Talk of eye sensors tracking your gaze, even predicting where you'll look next. Imagine typing just by looking at a keyboard? is the market ready for mass adoption? The next 6-9 months should tell. #AppleVR

1/9✍️ Just finished listening to a 45 minute podcast episode from @AiBreakfast with @SumanPersonalAI from @Personalai.

Took some notes for you guys (so you don't have to).

Here's the first key idea: "Don't solve the human, solve the problem." 2/9✍️ I think that's the mantra of Personal AI?

Apparently it started with filling in the gaps in blind people's vision. Then they transitioned into filling the gaps in our memory and cognition.

Pretty freaking cool.