Image for post
Image for post

Every week, my team at Invector Labs publishes a newsletter to track the most recent developments in AI research and technology. You can find this week’s issue below. You can sign up for it below. Please do so, our guys worked really hard on this:

Image for post
Image for post
Image for post
Image for post

From the Editor

What is curiosity? Can we build artificial intelligence(AI) agents that are inheritably curious? Curiosity is one of our marvelous cognitive abilities of humans that is hard to explain from a neuroscientific standpoint. In the context of AI, curiosity is often seen as an opposing force of actions. How much should an AI agent explore an environment instead of taking actions that yield immediate results? That’s the synthesis of the exploration-exploitation dilemma that constraints curiosity in areas such as reinforcement learning.

The magic of AI curiosity is based on differentiating effective from unproductive curiosity. Just this week, researchers from OpenAI and Google published separate papers about new method for building curiosity in reinforcement learning agents. While these efforts are mostly theoretical, they provide a glimpse of how curiosity can evolve in modern AI. Curiosity is a key pillar of human intelligence and, as a result, should be present in some form in the next generation of AI agents.

Now let’s take a look at the core developments in AI research and technology this week:

Image for post
Image for post

Research

OpenAI published a paper proposing a method called Random Network Distillation(RND) to encourage curiosity in reinforcement learning models.

>Read more in this blog post from the OpenAI team

Staying with the subject of curiosity, AI researcher from Google published a paper proposing a method called Episodic Curiosity to discriminate good from bad curiosity in reinforcement learning agents.

>Read more in this blog post from the Google AI team

IBM AI researchers took a step towards evolving the famous Word2Vec algorithm and creating universal text embedding representations.

>Read more in this blog post from IBM Research

Image for post
Image for post

Cool Tech Releases

Facebook open sources Horizon, a library for implementing large scale reinforcement learning solutions

>Read more in this blog post from the Facebook engineering team

Google released AdaNet, a TensorFlow-based library that uses neural networks to create ensemble learning models.

>Read more in this blog post from the Google AI Research team

Google also open sourced BERT, a bidirectional autoencoder library that enables the state-of-the-art training of question-answering systems

>Read more in this blog post from the Google AI Research team

Image for post
Image for post

AI in the Real World

Deep Learning for Detecting Glaucoma: AI researchers from New York University partnered with IBM in a new effort to use deep learning to detect glaucoma in images.

>Read more in this blog post from IBM Research

Malta Goes Decentralized AI: The stellar group of SingularityNet have been selected by the government of Malta for creating an AI strategy for the island.

>Read more in this blog post from the SingularityNet team

China AI Future: Chinese President Xi Jinping encouraged the country to advance AI research to secure the country’s technological future.

>Read more in this article by the South China Morning Post

Written by

CEO of IntoTheBlock, Chief Scientist at Invector Labs, Guest lecturer at Columbia University, Angel Investor, Author, Speaker.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store