AI Pendant, Another AI wearable, now from Rewind

In today’s email:

🧞‍♀️ AI Pendant- Another AI wearable, now from Rewind
💆‍♂️ Automata: Building an AI That Builds Itself
🎶 Spotify in the Works with Prompt-Driven AI Playlist
🛠 Various AI-related tools and platforms. Easiest way to run LLM locally, animate anything, generate your body transformation using AI, a private, custom-tailored & secure Legal AI copilot, the fastest and smartest domain search in the galaxy, and more.

AI Pendant- Another AI wearable, now from Rewind [Link]

AI wearables are the theme of this week. Besides Rewind, we had 4 other announcements in the past few days:

OpenAI, in collaboration with iPhone designer Jony Ive, is in the works to create an AI-integrated device with a potential investment of up to $1B from SoftBank.
Humane, having secured over $230M in funding, introduces a wearable accessory that understands you and functions as your personal aide, set to launch on 11/9.
Meta recently unveiled AI-infused Ray-Ban smart glasses capable of visual identification and answering queries.
Recently, Avi showcased Tab, a pendant accessory that serves as a personal creative aide, designed to understand you intimately. The first 100 unit was sold out at $600.

Introducing Rewind Pendant – a wearable that captures what you say and hear in the real world!

⏪ Rewind powered by truly everything you’ve seen, said, or heard
🤖 Summarize and ask any question using AI
🔒 Private by design

Learn more & preorder: rewind.ai/pendanttwitter.com/i/web/status/1…

— Dan Siroker (@dsiroker)
Oct 2, 2023

Turn almost perfect pictures into invaluable memories [Link]

RealFill is an innovative inpainting model that uses reference images to accurately fill in missing parts of a target image. It outperforms other methods in tests on a new benchmark.

Using up to five reference images, RealFill fine-tunes a pre-trained inpainting model to grasp the scene’s nuances. The adapted model then completes the target image’s gaps.

RealFill effectively in-paints target images, matching reference images even with varied lighting, viewpoints, and motion.

Comparison with Baselines

Paint-by-Example lacks detailed scene fidelity.
Stable Diffusion In-painting doesn’t always align with references.
RealFill consistently matches the reference images.

RealFill is time-intensive due to fine-tuning.
It struggles with major viewpoint changes and challenges faced by the base model, like text

Automata: Building an AI That Builds Itself [Link]

Podcast: Why Your AI Startup is Probably* Already Dead – Special Guest Scott Belsky [Link]

Adobe Chief Strategy Officer Scott Belsky joins the Morins and the Lessins as the Fifth Wheel for an episode about all things AI. Plus, inside the FTC’s case for breaking up Amazon.

Other stuff

Promptbreeder employs large language models like GPT-3 to iteratively improve text prompts. But here’s the magic – it doesn’t just evolve the prompts themselves. It also evolves how the prompts are generated in the first place. [Link] RECOMMENDED
Tutorial: Getting the most out of Midjourney with your own private Discord Server [Link]
A cool demo of Runway ML [Link] RECOMMENDED
A Hustler’s Haven: AGI House – Breeding Ground for AI Maestros [Link] RECOMMENDED
Google Accused of Stifling Bing’s Growth [Link]
Pixel 8 Leaks Reveal Enhanced AI & New Cases [Link]
Microsoft’s Yusuf Mehdi on the company’s AI transformation [Link]
A 166-page report from Microsoft qualitatively exploring GPT-4V capabilities and usage [Link]
Amazon Anthropic: Poison Pill or Empire Strikes Back [Link]
Spotify in the Works with Prompt-Driven AI Playlist [Link] RECOMMENDED
Tom Hanks Warns Fans to ‘Beware’ of a Dental Plan Ad Featuring an AI Version of Him [Link]