Last Week in AI

Jesus Rodriguez
Aug 4 · 3 min read

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:

From the Editor: The Race for Artificial General Intelligence in On

Artificial general intelligence(AGI) is the term adopted within the AI community to describe agents that can perform multiple, diverse tasks at human level of superior. The current generation of AI systems most have shown proficiency on very specific tasks but have failed to generalized that knowledge to other areas. To many, AGI seems like the science fiction thing but the fact that is we might be one to two breakthroughs away from achieving some forms of AGI.

Last week, Microsoft announced a $1 billion investment in AI research firm OpenAI. Part of the investment is destined to the goal of building systems that can achieve generic levels of intelligence compared to humans. While most people believe that AI today is a bit more than sophisticated pattern recognition, the OpenAI team believes that huge computer power and some advancements in areas such as transfer learning might all we need. Imagine what could happen if we can assign $100 million of computer power to a multi-task neural network. What would it be able to do?

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

AI Research

Facebook AI Research(FAIR) published an incredibly detailed article about techniques, tools and datasets for tackling challenges in conversational AI.

>Read more in this blog post from the FAIR team

Google Research published an article proposing a technique that uses adversarial neural networks to improve the stability of machine translation models.

>Read more in this blog post from Google Research

IBM Research published a paper discussing HEIDL, a system in which humans and machine can collaborate to improve the effectiveness of natural language processing models.

>Read more in this blog post from IBM Research

Amazon AI researchers published a paper about new techniques used to teach assistants like Alexa to answer complex questions.

>Read more in this blog post from Amazon Research

Cool AI Tech Releases

Facebook unveiled EGG, a new toolkit for creating simulation in which two agents can develop their own form of communication in order to solve a specific task.

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

AI powerhouse DeepMind unveiled Streams, a research medical assistant that can give doctors up to 48 hours head start when treating critical medical conditions.

>Read more in this blog post from the DeepMind team

AI in the Real World

Airbnb disclosed that has developed AI models that can predict the prices of properties with up to 69% accuracy.

>Read more in this coverage from VentureBeat

An AI system can predict which movies will perform well on the box office by just analyzing plot summaries.

>Read more in this article in Science

Experts believe that some minor advancements in AI research and huge computing power might help us deliver the first iterations of artificial general intelligence(AGI) in five years.

>Read more in this article from the Financial Times

Jesus Rodriguez

Written by

Chief Scientist, Managing Partner at Invector Labs. CTO at IntoTheBlock. Angel Investor, Writer, Board Member of Several Software Companies

Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade