Daily issuepublished

Proof that using GPT-4 improves your performance

Using ChatGPT custom instructions for fun and profit

By 7 min read
Proof that using GPT-4 improves your performance

In today’s email:

  • 🧠 Consultants using ChatGPT-4 outperformed those who did not, by a lot

  • 🧩 Generative AI is just a phase. What’s next is interactive AI.

  • 👩‍💻 I Saved 60 Hours of Coding With The "RJP" Prompting Technique

  • 🧰 Various AI-related tools and platforms. Open source AI agent community, AI-generated games, AI Jeopardy, AI survey maker, one-click competitor research, instant 3D face recognition, free, AI-powered research tool for scientific literature, and more.

Sign Up | Advertise | ChatGPT

Consultants using ChatGPT-4 outperformed those who did not, by a lot [Link]

Consultants using GPT-4 completed 12% more tasks, executed tasks 25% faster, and achieved 40% higher quality results. Story illustration

For the last several months, I been part of a team of social scientists working with Boston Consulting Group, turning their offices into the largest pre-registered experiment on the future of professional work in our AI-haunted age. Our first working paper is out today. There is a ton of important and useful nuance in the paper but let me tell you the headline first: for 18 different tasks selected to be realistic samples of the kinds of work done at an elite consulting company, consultants using ChatGPT-4 outperformed those who did not, by a lot. On every dimension. Every way we measured performance.

DeepMind’s cofounder: Generative AI is just a phase. What’s next is interactive AI. [Link]

Story illustration

DeepMind cofounder Mustafa Suleyman wants to build a chatbot that does a whole lot more than chat. In a recent conversation I had with him, he told me that generative AI is just a phase. What’s next is interactive AI: bots that can carry out tasks you set for them by calling on other software and other people to get stuff done. He also calls for robust regulation—and doesn’t think that’ll be hard to achieve.

Suleyman left DeepMind and moved to Google to lead a team working on AI policy. In 2022 he founded Inflection, one of the hottest new AI firms around, backed by $1.5 billion of investment from Microsoft, Nvidia, Bill Gates, and LinkedIn founder Reid Hoffman. Earlier this year he released a ChatGPT rival called Pi, whose unique selling point (according to Suleyman) is that it is pleasant and polite. And he just coauthored a book about the future of AI with writer and researcher Michael Bhaskar, called The Coming Wave: Technology, Power, and the 21st Century's Greatest Dilemma.

DeepMind discovers that AI large language models can optimize their own prompts [Link]

Story illustration

In a new paper, researchers from DeepMind propose a new way: Optimization by PROmpting (OPRO), a method that uses AI large language models (LLM) as optimizers. The unique aspect of this approach is that the optimization task is defined in natural language rather than through formal mathematical definitions.

The researchers found that, on small-scale optimization problems, LLMs can generate effective solutions through prompting alone, sometimes matching or even surpassing the performance of expert-designed heuristic algorithms. However, the true potential of OPRO lies in its ability to optimize LLM prompts to get maximum accuracy from the models.

Generative Image Dynamics [Try it yourself]

Story illustration

We present an approach to modeling an image-space prior on scene dynamics. Our prior is learned from a collection of motion trajectories extracted from real video sequences containing natural, oscillating motion such as trees, flowers, candles, and clothes blowing in the wind. Given a single image, our trained model uses a frequency-coordinated diffusion sampling process to predict a per-pixel long-term motion representation in the Fourier domain, which we call a neural stochastic motion texture. This representation can be converted into dense motion trajectories that span an entire video. Along with an image-based rendering module, these trajectories can be used for a number of downstream applications, such as turning still images into seamlessly looping dynamic videos or allowing users to realistically interact with objects in real pictures.

Other stuff

  • How Silicon Valley doomers are shaping Rishi Sunak’s AI plans [Link]

  • Sam Altman’s Net Worth: How wealthy is the CEO of ChatGPT-maker OpenAI [Link]

  • The Executive Trying to Make Microsoft Smarter About AI [Link]

  • GPT-4 is not getting worse! [Link] RECOMMENDED

  • AI may outperform most humans at creative thinking tasks [Link]

  • China’s AI boom depends on an army of exploited student interns [Link] RECOMMENDED

  • I Saved 60 Hours of Coding With The "RJP" Prompting Technique [Link] RECOMMENDED

  • Using ChatGPT Custom Instructions for Fun and Profit [Link] RECOMMENDED

Superpower ChatGPT 5.0.0 has been released. 🎉

The most powerful release yet with Prompt Chains, AutoComplete Menu, Quick Sync, Custom Instruction Profiles, and many more features.

Superpower ChatGPT Extension on Chrome

 

Superpower ChatGPT Extension on Firefox

 

Editor's Pick

WebTranspose - Supercharge your LLMs with Website Data [Link]

Editor's Pick ✨

Agents - An Open-source Framework for Autonomous Language Agents [Link][GitHub]

Story illustration

Narraive - AI-Generated, Player Driven games [Link]

Story illustration

illa - Open source AI Agent community [Link]

Story illustration

Superflow - Open-source toolkit to build an AI assistant for SaaS products [Link] [GitHub]

Story illustration

EasyRAG - Dead simple API for Retrieval Augmented Generation [Link]

Story illustration

AI Jeopardy - GPT4 Powered Multiplayer Jeopardy show [Link]

Story illustration

Krea - Create an AI with your own style. [Link]

Story illustration

Astrocyte - Create and chat with 3D AI characters [Link]

Story illustration

Semantic Scholar - A free, AI-powered research tool for scientific literature [Link]

Story illustration

KBaseBot - From files & sites to chatbots: knowledge talks [Link]

MindPal 2.0 - Build AI agents for your tasks, with your data [Link]

OneAudio AI - Transform voice recordings into clean summarized notes [Link]

Story illustration

AI Survey Maker - Make or import a survey in seconds with ChatGPT [Link]

Story illustration

Auto Localize - Localize your projects automatically with the power of AI [Link]

Story illustration

StockInsights AI - Simplifying equity research with AI [Link]

Story illustration

Opinly.ai - One-click competitor research [Link]

Story illustration

Facia - Instant 3D face recognition for enhanced security [Link]

Omniscience - AI books - cover art, full text, and all - from a prompt [Link]

OpenPlugin - A ChatGPT-based chatbot that specializes in leveraging plugins [Link][GitHub]

Prompt of the Day 🍭: Learning using the 80/20 principle

Click on the image to load the full prompt in ChatGPT (requires Superpower ChatGPT)

Story illustration

Unclassified 🌀 

Story illustration