We’ve trained a neural network called DALL·E that creates images from text captions for a wide range of concepts expressible in natural language.
We’ve trained a model called ChatGPT which interacts in a conversational way. The dialogue format makes it possible for ChatGPT to answer followup questions, admit its mistakes, challenge incorrect…
Starting today, developers can begin building apps with the DALL·E API.
We’re introducing Jukebox, a neural net that generates music, including rudimentary singing, as raw audio in a variety of genres and artist styles. We’re releasing the model weights and code, along…
An API for accessing new AI models developed by OpenAI
We’ve discovered neurons in CLIP that respond to the same concept whether presented literally, symbolically, or conceptually.
Large neural networks are at the core of many recent advances in AI, but training them is a difficult engineering and research challenge which requires orchestrating a cluster of GPUs to perform a…
We’ve trained a large-scale unsupervised language model which generates coherent paragraphs of text, achieves state-of-the-art performance on many language modeling benchmarks, and performs…
We are excited to announce a new embedding model which is significantly more capable, cost effective, and simpler to use. The new model, text-embedding-ada-002, replaces five separate models for text…
We are standardizing OpenAI’s deep learning framework on PyTorch. In the past, we implemented projects in many frameworks depending on their relative strengths. We’ve now chosen to standardize to make…
We've discovered that evolution strategies (ES), an optimization technique that's been known for decades, rivals the performance of standard reinforcement learning (RL) techniques on modern RL benchmarks (e.g. Atari/MuJoCo), while overcoming many of RL's inconveniences. In particular, ES is simpler to implement (there is no need for backpropagation)
We’ve developed Random Network Distillation (RND), a prediction-based method for encouraging reinforcement learning agents to explore their environments through curiosity, which for the first time exceeds average human performance on Montezuma’s Revenge.
We’ve trained and are open-sourcing a neural net called Whisper that approaches human level robustness and accuracy on English speech recognition. Read Paper View Code View Model Card Whisper…
DALL·E 2 is a new AI system that can create realistic images and art from a description in natural language.
OpenAI is committed to the safe deployment of AI. Since the launch of our API, we’ve made deploying applications faster and more streamlined while adding new safety features. Our progress with…
«developers in supported countries can sign up and start experimenting with our API right away.»
We trained a neural network to play Minecraft by Video PreTraining (VPT) on a massive unlabeled video dataset of human Minecraft play, while using only a small amount of labeled contractor data. With…
We’re releasing an analysis showing that since 2012, the amount of compute used in the largest AI training runs has been increasing exponentially with a 3.4-month doubling time (by comparison, Moore’s…
OpenAI researchers collaborated with Georgetown University’s Center for Security and Emerging Technology and the Stanford Internet Observatory to investigate how large language models might be misused…
How does Refind curate?
It’s a mix of human and algorithmic curation, following a number of steps:
- We monitor 10k+ sources and 1k+ thought leaders on hundreds of topics—publications, blogs, news sites, newsletters, Substack, Medium, Twitter, etc.
- In addition, our users save links from around the web using our Save buttons and our extensions.
- Our algorithm processes 100k+ new links every day and uses external signals to find the most relevant ones, focusing on timeless pieces.
- Our community of active users gets the most relevant links every day, tailored to their interests. They provide feedback via implicit and explicit signals: open, read, listen, share, mark as read, read later, «More/less like this», etc.
- Our algorithm uses these internal signals to refine the selection.
- In addition, we have expert curators who manually curate niche topics.
The result: lists of the best and most useful articles on hundreds of topics.
How does Refind detect «timeless» pieces?
We focus on pieces with long shelf-lives—not news. We determine «timelessness» via a number of metrics, for example, the consumption pattern of links over time.
How many sources does Refind monitor?
We monitor 10k+ content sources on hundreds of topics—publications, blogs, news sites, newsletters, Substack, Medium, Twitter, etc.
Can I submit a link?
Indirectly, by using Refind and saving links from outside (e.g., via our extensions).
How can I report a problem?
When you’re logged-in, you can flag any link via the «More» (...) menu. You can also report problems via email to email@example.com
Who uses Refind?
300k+ smart people start their day with Refind. To learn something new. To get inspired. To move forward. Our apps have a 4.9/5 rating.
Is Refind free?
Yes, it’s free!
How can I sign up?
Head over to our homepage and sign up by email or with your Twitter or Google account.