GPT-3 summariser that reads websites

Ben Tossell
November 28, 2022

Tomorrow never comes, but Mondays always do!

Welcome to the 1,185 new folks joining us from last Monday 🤯❤️.

Let's kick off the week.

Storytimes - Create the perfect children's book with AI. I wanted to do this for my nephew and started plotting all the tools I’d need to actually train a model on his face, generate a reasonable storyline, generate good images, get an actual book made etc. I even spoke with someone last week as they had this same idea. But Storytimes just built the all-in-one tool for it! (link)

These AI-powered smart glasses will change the game for the deaf or hard of hearing. They provide real-time subtitles during conversations. (link)
Michelle trained an AI chatbot on their childhood journal entries - so that they could engage in real-time dialogue with their "inner child". (link)
Daniel Gross made a GPT-3 summariser that reads websites just like humans do. It scrolls pages and reads in visible text in chunks, which it then attempts to summarise. (link)

SnapshotAI - turn any subject into works of art within an hour. (link)
Spellbook - AI contract drafting and review. (link)
Powderworld - a lightweight simulation environment for understanding AI generalization. (link)
KnowGL: Knowledge generation and linking from text. (link)
Passphoto.ai - Create your passport photo with AI. (link)
Article.audio - Convert articles into audio. (link)
Recast Studio - Automatically create dozens of short video snippets from your podcast or video content using AI. (link)
Flair - the AI design tool for branded content. (link) They announced their funding from AI Grant here.

👋 Too many links?! I created a database for all links mentioned in these emails. Refer 1 friend using this link and I'll send over the link database.

MAIDA - Diverse medical image datasets from across the globe. (link)
Converting Diffusion AI imagery to 3D holograms with depth estimation AI. (link)
Solving math word problems with process- and outcome-based feedback. (link)
Towards Photorealistic 3D Object Generation and Editing with Text-guided Diffusion Models. (link)
SpaText - new method for text-to-image generation using open-vocabulary scene control. (link)
Sketch-Guided Text-to-Image Diffusion Models. (link)
EDGAR: Embedded Detection of Gunshots by AI in Real-time. (link)

Why "Prompt Engineering" and "Generative AI" are overhyped. (link)
Fine tuning Stable Diffusion v2.0 with DreamBooth. (link)
A fine-tuned Stable Diffusion model trained on images from Mad Max: Fury Road. (link)

Moonbeam, an AI writing tool, has released collaborative writing. And yes, you can invite AI to collaborate too. (link)
Accelerated reaction energy prediction for redox batteries. (link)
AvatarAI is pivoting from just avatars to entire photo shoots. Let’s see how many other avatar tools follow suit! (link)
Melting Pot by DeepMind - A suite of test scenarios for multi-agent reinforcement learning. (link)
Building NLP applications with LLMs just got easier with LangChain AI. Easily chain multiple prompts together to build sophisticated AI solutions. (link)
A thread from Runway CEO on how AI will change how we tell stories. (link)
Raw output examples of Stable Diffusion 2.0. (link)
AI language & art models are growing at 10x a year. A thread. (link)
Scale CEO, Alexander Wang wrote a post on “The AI War, The China Threat, and How to Win It”. (link)
A thread on all 15 outstanding papers from NeurIPS. A TLDR and why they are significant. (link)
Stable Diffusion 2.0 is live on PlaygroundAI. And, it’s free. (link)
Raoul Pal and Emad on a podcast “AI: This is the Start of a Complete Shift in Society”. (link)
Thoughts on Stable Diffusion 2 — models-turtles, retraining hurdles, and step towards ClosedAI. (link)