In the News
Cheap manual labor is one of the current requirements for large AI developments. China has a lot of inexpensive labor force and is making it one of its key advantages to become the AI world leader by 2030.
“News that tech giant Ma is a member of Communist party of China should set alarm bells ringing – and not just in China”
Yoshua Bengio wants to stop talk of an AI arms race and make the technology more accessible to the developing world.
Also in the news…
- US commerce department proposes new export restrictions on AI. Tweet, Report
- Fearful of gender bias, Google blocks gender based pronouns from Smart Compose. More
Lack time to read research papers? Have a look at this great summary of some of the main ideas found in recent research papers, along with comments from the community.
Multi-Task Learning is a general method in which a single architecture is trained towards learning several different tasks at the same time. Here’s an example with such a model (HMTL) trained to beat the state-of-the-art on several NLP tasks.
Looking at GANs less from a performance standpoint and more from an artistic one.
“If data issues have lasting consequences, why are we less sophisticated at testing than our software developing counterparts?”
How the NYT builds fast and reusable econometric models in-house
Software tools & code
A new state of the art in Image synthesis
Comparing prices to train models on GPUs and TPUs with AWS, Google, Paperspace, vast.ai and others.
Graphics pipeline, vector processing along with the other graphics operations. Useful to have in mind when wading into GPU-powered ML training.
Former Professor at Stanford University in Computer Vision, now Chief Scientist for Google Cloud AI — on her “quest to make AI better for humanity”.
This newsletter is a collection of AI news and resources curated by @dlissmyr. If you find it worthwhile, please forward to your friends and colleagues, or share on your favorite network!
Suggestions or comments are more than welcome, just reply to this email.