Clik here to view.
OpenAI Shouldn’t Release Their Full Language Model
In response to Dear OpenAI: Please Open Source Your Language Model In a world where researchers and corporations emphasize their goals of "democratizing AI" and "AI for everyone," there is an...
View ArticleClik here to view.
The Promise of Hierarchical Reinforcement Learning
Update: Jürgen Schmidhuber kindly suggested some corrections concerning the early work on intrinsic motivation, subgoal discovery and artificial curiosity since 1990, which I have incorporated and...
View ArticleClik here to view.
Call for Ideas - March 2019
We want to publish your thoughts! One year ago, we launched the Gradient alongside our first Call for Ideas. Since then, we've published overviews on the state-of-the-art in game theory, hierarchical...
View ArticleClik here to view.
Stanford's Human-Centered AI Launch Symposium
Will artificial intelligence replace or augment humanity? On Monday, Stanford gave its answer by launching the new Institute for Human-Centered Artificial Intelligence (HAI). The opening event featured...
View ArticleClik here to view.
When AI Plans Ahead
Recent advances in neural networks have generated considerable excitement about AI. But AI is not all about neural networks. Other avenues in AI research tackle problems such as building effective...
View ArticleClik here to view.
Goodhart’s Law: Are Academic Metrics Being Gamed?
Publishing a paper in academia is challenging, stimulating, and a bit baffling. Challenging because the research might fail. Stimulating because research may start assuming one outcome and finish with...
View ArticleClik here to view.
The Past, Present, and Future of AI Art
AI art, or more precisely art created with neural networks, has recently started to receive broad media coverage. Flattering coverage from outlets including the New York Times and The Atlantic,...
View ArticleClik here to view.
Why We Released Grover
This June, our research team at the University of Washington released Grover, a state-of-the-art detector of neural fake news. Neural Fake News is the threat of AI-generated news articles controlled by...
View ArticleClik here to view.
Is Deep Learning the Future of Medical Decision Making?
Healthcare is often spoken of as a field that is on the verge of an AI revolution. Big names in AI such as Google DeepMind, publicise their efforts in healthcare, claiming that “AI is poised to...
View ArticleClik here to view.
Leveraging Learning in Robotics: RSS 2019 Highlights
What's the current state of research in AI focusing on Robotics? This article will briefly summarize the technical work presented at RSS 2019 Conference. Let's start by knowing more about the...
View ArticleClik here to view.
Introducing Retrospectives: 'Real Talk' for your Past Papers
Researchers get credit for writing papers. If you’re a professor, the number of accepted papers determines whether you’ll get tenure. If you’re a student, it determines if and when you can graduate, as...
View ArticleClik here to view.
NLP's Clever Hans Moment has Arrived
It is now almost a cliché to find out that BERT (Devlin et al., 2019) performs "surprisingly well" on whatever dataset you throw at it. In their recent paper, Niven & Kao throw an argument...
View ArticleClik here to view.
The #BenderRule: On Naming the Languages We Study and Why It Matters
High Resource Languages vs Low Resource Languages #fnref1,#fnref2,#fnref3,#fnref5,#fnref4,#fnref11,#fnref10,#fnref14,#fnref15,#fnref16,#fnref24,#fnref21,#fnref32{ background-color: lightblue } Progress...
View ArticleClik here to view.
The State of Machine Learning Frameworks in 2019
Since deep learning regained prominence in 2012, many machine learning frameworks have clamored to become the new favorite among researchers and industry practitioners. From the early academic outputs...
View ArticleClik here to view.
Evaluation Metrics for Language Modeling
Recently, neural network trained language models, such as ULMFIT, BERT, and GPT-2, have been remarkably successful when transferred to other natural language processing tasks. As such, there's been...
View ArticleClik here to view.
Gaussian Processes, not quite for dummies
I recall always having this vague impression about Gaussian Processes (GPs) being a magical algorithm that is able to define probability distributions over sets of functions, but I had always...
View ArticleClik here to view.
How Machine Learning Can Help Unlock the World of Ancient Japan
Humanity’s rich history has left behind an enormous number of historical documents and artifacts. However, virtually none of these documents, containing stories and recorded experiences essential to...
View ArticleClik here to view.
Artificial Life for AI People
Artificial Life, often shortened as ALife. What is your first thought when reading those words? A brand of T-shirts? A Greg Egan novel? For me and hundreds of ALifers, ALife is the bottom-up scientific...
View ArticleClik here to view.
An Epidemic of AI Misinformation
Maybe every paper abstract should have a mandatory field of what the limitations of the proposed approach are. That way some of the science miscommunications and hypes could maybe be avoided.—...
View ArticleClik here to view.
Is NeurIPS Getting Too Big?
NeurIPS 2019, the latest incarnation of the Neural Information Processing Systems conference, wrapped up just over a week ago. Multiple great blog posts have already summarized various talks and key...
View Article
