Welcome to another week in the world of AI, with Ben's Bites. A big hello to the 1,709 people who joined us since last week. It's a snowy one here in the UK and I attempted to play golf on a very frozen course yesterday. (No, it didn't go well)!

Ben's Bites Hackathon begins today! We have $27k in cash prizes to give away. It is your last day to register so please do so here³⁶⁸. And join the Discord here³⁸⁴. I'll be sending out today/tomorrow any details that you need; guidelines, rules, voting system, prize criteria etc.

There's a community meetup in San Francisco on Wednesday - 14th Dec²⁹⁵, and Shre who helps me with BB will be there too.

🤌 Ben's Picks

• Marques Brownlee posted a video: The Truth About AI Getting "Creative". (link⁵³¹)

• Dreamlike Diffusion 1.0 - Stable Diffusion trained on high quality art. (link⁷³⁸)

🛠️ Cool Tools

• Dramatron - a new tool for writers to co-write theatre and film scripts with a language model, by DeepMind. (link⁴⁹⁰)

• GPT Visual Studio Code Extension - Use GPT-3 to generate documentation and get help debugging your code. (link⁵⁰⁰)

• Get an AI generated recipe based off ingredients you have at home. (link^708†)

• Vidyo - Make short videos from long ones instantly. (link⁷²⁷)

• Polycam AI texture generator. Type anything and watch it come to life in 3D. (link⁶²²)

• Ready-made forms for instructing ChatGPT to do things without repeating yourself, you can create your own and share them with others. (link^763†)

• Fuzzy Compiler - make up your own code, no worrying about syntax. Translates to working code w comments. (link⁴⁵⁷)

• AI content detector - enter the text and it’ll tell you if it’s AI-generated or real. (link⁵³⁷)

• A Discord bot for ChatGPT. (link³²³)

• Felvin - Wix for generative AI applications. (link⁵⁹⁸)

• GPT-3.5 powered terminal commands. (link⁴¹⁴)

• chatGP-T1000 - An AI shape-shifter bot, that assumes identities understands your natural language, and replies with chatGPT responses, in character with deepfaked audio and lipSync. (link⁴⁷⁶)

• TaleTime - A.I. children's storybook generation. (link^646†)

• ChatGPT prompt curation to use ChatGPT better. (link⁶⁷⁵)

• Chat with GPT with your voice in your native language. (link³⁶³)

• Make Logo AI - Unique HD logos, generated by AI. (link⁷⁸⁸)

• HowToReplyTo.com - AI-powered replies for every message. (link⁵⁹¹)

• YeahWrite - a really nice writing tool using AI. (link⁶²¹)

• HoDoI - Ask AI questions and get the real answers. (link⁵⁵⁶)

Join Discord³⁸⁴ | Hackathon³⁶⁸ | Partner with us¹⁶⁹  | Host or join a meetup¹⁶⁰

🎓 Learn

• Almost automatically generate blog posts on Webflow with this simple Zapier integration with OpenAI. (link³⁷³)

• Data cleaning and preprocessing for beginners. (link²⁴⁶)

🔬 Research

• Video-text modelling with zero-shot transfer from contrastive captioners. (link¹⁵²)

• LoopDraw: a loop-based autoregressive model for shape synthesis and editing. (link¹⁶⁸)

• Hyperbolic audio source separation. (link¹⁶¹)

• Phone2Proc: Bringing robust robots into our chaotic world. (link¹³⁶)

• ORCa: Glossy objects as radiance-field cameras. (link¹⁵⁰)

• PromptonomyViT: Multi-task prompt learning improves video transformers using synthetic scene data. (link¹⁴⁹)

• BigScience: A case study in the social construction of a multilingual large language model. (link¹⁴²)

• Seeing a rose in five thousand ways. (link²⁵⁵)

• SmartBrush: Text and shape guided object inpainting with diffusion model. (link¹⁵⁹)

• Sparse Upcycling: Training mixture-of-experts from dense checkpoints. (link¹²⁸)

• Ego-body pose estimation via ego-head pose estimation. (link¹²²)

• LORA - Low-rank adaptation for fast text-to-image diffusion fine-tuning. (link¹⁴⁴)

• DIP: Differentiable inter-reflection-aware physics-based inverse rendering. (link¹³⁸)

• Memories are one-to-many mapping alleviators in talking face generation. (link²¹⁹)

• PALMER: Perception-action loop with memory for long-horizon planning. (link¹⁶⁶)

• Masked lip-sync prediction by audio-visual contextual exploitation in transformers. (link¹⁵⁸)

• Training-free structured diffusion guidance for compositional text-to-image synthesis. (link¹³⁶)

👋 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.

🤓 Everything else

• An author showing how ChatGPT could completely disrupt their industry. (link⁵⁵⁴)

• How Amazon Robotics is working on new ways to eliminate the need for barcodes. (link²⁵⁷)

• Why Google missed ChatGPT. (link⁴²⁵)

• ChatGPT's math-nerd alter ego. (link²⁴⁴)

• Using ChatGPT to write and illustrate a children's story about "a robot that wanted to be a human." (link²²⁹)

• You can now do video classification in just 3 lines of code thanks to the new video-classification pipeline in Hugging Face transformers. (link¹⁸³)

• Sarah Guo wrote about her EOY prediction for AI. (link³⁵¹)

• Surveillance robot dog unveiled in Ferndale. (link²⁸⁵)

• AI’s Jurassic Park moment. (link³¹⁶)

• Stumbling with their words, some people let AI do the talking. (link^222†)

• David Friedberg’s prediction is that “it seems more likely that AI won't replace the creator in the traditional model but rather will become the personalized content creator for each individual…” (link²³⁴)

• A thread of all the use cases where ChatGPT lacks and its output is not correct. (link²²⁵)

• A diffusion model trained on couches - Couch Diffusion. (link¹⁹⁰)

• A dreambooth model trained on a diverse set of analog photographs - Analog Diffusion. (link⁵⁴⁷)

• A collection of 50 curated AI designs. (link³³⁷)

• Disputing a parking fine with ChatGPT. (link³⁷⁷)

🧑‍💻 Who's hiring in AI

VEED.IO - Simple Online Video Editing. VEED is hiring AI / ML engineers⁹⁴ to level up its creative toolkit and make it more magical.

Buildspace - where builders, build! They're looking for an ML/AI instructor to build their new course¹²².

Add your company here¹²³ | Join the Talent pool¹¹⁹

🖼 AI IMAGES OF THE DAY

Someone on Reddit is creating fake selfies going through history. They're pretty spectacular!

👋 SEE YA