Every week, Invector Labs publishes a newsletter that covers 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: An Interesting Message at AI’s Biggest Party
The NeurIPS conference is considered by many the top artificial intelligence(AI) academic event of the year. Last week, over 13000 AI researchers descended over the city of Vancouver to participate in the latest edition of NeurIPS. While previous editions of NeurIPS were filled with overly optimistic pictures about the future, this year saw a shift of focus towards addressing some of the most important limitations of the latest generation of AI systems.
To use some examples, deep learning legend Yoshua Bengio used his keynote to highlight the very narrow learning mechanisms of most AI models and the difficulty of reusing knowledge across AI systems. Similarly, Google’s Blaise Aguera y Arcas called AI researchers to pay attention to the “biological nature” of intelligence and to incorporate those mechanisms into deep learning systems. Overall, NeurIPS celebrated the recent advancements in AI research but emphasized a very sober and pragmatic view of the near term future of AI.
Now let’s take a look at the core developments in AI research and technology this week:
Microsoft researchers published a paper proposing a method that uses symbolic representations to understand the knowledge of neural networks.
Some amazing keynotes from the NeurIPS conference are now online including Turing Award winner’s Yoshua Bengio.
Deep Learning Training at Scale
Uber published a blog post describing a very innovative architecture for productionizing the XGBoost method used to train deep learning models at scale.
Cool AI Tech Releases
Google open sourced Fairness Indicators, a suite of tools that enable regular computation and visualization of fairness metrics.
Deep Fake Detection Challenge
Facebook announced a new Kaggle challenge and a new dataset to advance the research in deep fake detections.
A Better Dataset for Object Detection
Researchers from the Massachusetts Institute of Technology(MIT) released ObjectNet, a new dataset that attempts to bridge the gap between AI systems and humans when comes to object recognition.
AI in the Real World
Google Discusses AI Trends for 2020
Google AI chief Jeff Dean gave a vert insightful interview about some of the interesting AI trends that are likely to see traction next year.
AI R&D in 2019
The AI Index Report its one of the most anticipated documents in the field of AI. This year’s edition covered a lot of ground including AI research and adoption.
AI and the Search for Alien Life
NASA is using deep learning to interpret data that will be collected by future telescopes like the James Webb Space Telescope or the Transiting Exoplanet Survey Satellite mission.