Ben's Bites
Posts
Build an AI App, without code, in 5 minutes

Build an AI App, without code, in 5 minutes

Ben Tossell
November 30, 2022

I'm back in the tutorial saddle! After so many people tweeted me about tutorials yesterday I thought I'd give it a go. (I built Makerpad, a no-code tutorial site).

Let me know what you think and any requests to do more, on this tweet.

Let's get into it.

🤌 Ben's Picks

I made a tutorial! How to build an AI app, without writing code, in 5 mins. (link) I’d love it if you wanted to share the tweet + reply there with any requests for more tutorials.

Stability AI has released its official API. Why? Well, you can access the outputs by using a tool with a UI or knowing how to code to set it up yourself. So an API gives another way to access the outputs, you can plug them into spreadsheets with Zapier and plug them into an app much easier. (link)
Introducing GPT-JT, a 6B parameter open source model that can outperform many 100B+ parameter models and was trained over slow (1Gbps) internet links. (link)

🛠️ Cool Tools

Postwise - Write, schedule and grow with this AI Twitter tool. (link)
Elephas - Your personal AI writing assistant, is now available across all your Apple devices and iOS apps. (link)
Predis.ai - Create social media posts with AI. (link)
Snipd - AI podcast player to discover & save the best highlights in podcasts, has just released a new feature: Key takeaways - auto-generated takeaways from episodes. (link)
Snipets - Save highlights from youtube, newsletters and Twitter in seconds. (link)

Finta - AI-powered fundraising tool for founders. (link)
OpenAI Cookbook - example code and example prompts for accomplishing common tasks with the OpenAI API. (link)

👋 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.

🔬 Research

SGAM: Building a virtual 3D world through simultaneous generation and mapping. (link)
Exploring two complementary methods for improving existing language models by a large margin without using massive computational resources, from Google AI. (link)
NeuralLift-360: Lifting an in-the-wild 2D photo to a 3D object with 360° views. (link)
Research suggests that data centre hardware used for AI training is experiencing strong cost declines, “Applying Wright’s Law to AI accelerators”. (link)
Decision Diffuser - a conditional diffusion model that outperforms offline RL across standard benchmarks – using only generative modelling training. (link)
A framework for comparing different network architectures using counterfactual simulation testing. The Naturalistic Variation Object Dataset (NVD) for studying the generalization and robustness of modern neural networks. (link)
Using Vision Transformers (ViT) to accelerate data loading and training of neural networks for computer vision. (link)
Improving common sense in vision-language models via knowledge graph riddles. (link)
Graph Neural Networks: a powerful and versatile tool for advancing design, reliability, and security of ics. (link)
Coder reviewer reranking for code generation. (link)
One is All: bridging the gap between neural radiance field architectures with progressive volume distillation. (link)

🤓 Everything else

New documentation on language models used in OpenAI's research is up. (link)
A free, comprehensive course to turn you into an NLP expert. (link)
AssemblyAI’s Winter Hackathon starts December 9th and there’s $50k in prizes up for grabs. (link)
OpenAI’s Whisper ASR is transcribing Apollo 13’s 7,200 hours of mission control chatter. (link)
Woolitize fine-tuned Stable Diffusion model. (link)

Hugging Face is putting on a live event: Diffusion models, today (Wednesday 18:00 CET). (link)
A riddle from GPT-3: “What is quiet enough to hear yet not powerful enough to be heard? (link)
Homebrew AI club are hosting a Holiday Social for AI Builders in San Francisco on Thursday 8th December. (link)
Some examples of OpenAI davinci-002 versus davinci-003 using Todoist's Suggest Tasks from AI. (link)
A low-poly-world fine-tuned Stable Diffusion model. (link)
Bleeding Edge (an AI news aggregator) is launching AMA’s. This Thursday 10:30am PT they’re hosting an AMA with the team behind Meta AI’s Universal Speech Translator project. (link)
Alberto back with another great piece, this time “Stable Diffusion 2 Is the First Artist-Friendly AI Art Model”. (link)
Tutorial: How to set up your own AI chatbot for your journal entries. (link)
NVIDIA wins NeurIPS Awards for research on generative AI, generalist AI agents. (link)