From the Editor: OpenAI Launches its First Product
OpenAI has been at the center of some of the biggest advancements in artificial inteligence(AI) in recent years. Created by industry luminaries such as Elon Musk and Sam Altman, OpenAI started as a non-profit organization with a focus of advancing AI research. After Altman took over as CEO last year, OpenAI transitioned to a capped profit structure and attracted $1 billion investment from Microsoft. The next step in the evolution of OpenAI seems to be to build up its commercial muscle and that’s what they seem to be doing.
Earlier this week, OpenAI unveiled an API product that exposes endpoints for some of its most sucessful language models including the controversial GPT-3. The API is available for Beta but already has dran interest from big technology companies such as Google and Microsoft. The current offer is mostly focused on text-generation but we can expect it to expand from there. OpenAI’s step have been seen as a controversial move. Personally, I like the approach as it might allow startups to have access to the kind of AI models that only big corporate labs can create these days.
Now let’s take a look at the core developments in AI research and technology this week:
Summarizing Long Texts
Google Research published a paper introducing PEGASUS, a transformer-based model for text summarization.
AI Agents Playing Diplomacy
DeepMind published a paper proposing a method to allow reinforcement learning agents to develop cooperative skills by playing the Diplomacy game.
Cross Language Transfer Learning
Microsoft Researchers published a paper proposing a method for transfer knowledge across different languages.
Cool AI Tech Releases:
Uber open sourced Neuropod, a framework that provides a common interface for executing deep learning models…