“Liquid” Machine-Learning System Adapts To Changing Conditions
MIT researchers have developed a type of neural network that learns on the job, not just during its training phase. These flexible algorithms, dubbed “liquid” networks, change their underlying equations to continuously adapt to new data inputs. The advance could aid decision making based on data streams that change over time, including those involved in …
MIT Continues AI Collaboration With Toyota Research Institute
The Toyota Research Institute (TRI) announced that it will be extending its AI research collaboration with MIT. TRI, which also has existing relationships with Stanford and the University of Michigan, has selected 13 additional academic institutions to participate in the next five-year phase of its research initiative. TRI’s next five-year phase involves investing more than …
AI Goes Anonymous During Training To Boost Privacy Protection
Privacy is vital – even more so in the modern era of AI. But AI trained on personal data can be hacked. Even if the hacker doesn’t access the training data, there’s still a risk of leaking sensitive information from the models themselves. For example, it may be possible to reveal if someone’s data is …
Darktrace Version 5 Released For General Availability To Protect Cloud And Remote Workforce
Darktrace, the world’s leading cyber AI company, launched Version 5 of its autonomous, self-learning Darktrace Immune System for general availability, empowering organizations to embrace the advantages of the cloud, while protecting against novel cyber-threats. Powered by self-learning Cyber AI, the Darktrace Immune System works by learning the normal ‘pattern of life’ of an organization, and …
Lab Will Co-Lead NASA Mission To Study Stars, Planets
NASA has selected Lawrence Livermore National Laboratory (LLNL) and Goddard Space Flight Center (GSFC) to serve as lead institutions for the Pandora scientific mission that will study 20 stars and their 39 exoplanets. The goal of the Pandora mission is to learn about starspots (akin to sunspots) and identify which of these exoplanets are …
A New Open Data Set For Multilingual Speech Research
What it is: Facebook AI is releasing Multilingual LibriSpeech (MLS), a large-scale, open source data set designed to help advance research in automatic speech recognition (ASR). MLS is designed to help the speech research community’s work in languages beyond just English so people around the world can benefit from improvements in a wide range of …
Objectively Measuring Chronic Pain With AI
Pain. We will all experience it at some point, and some of us suffer from it chronically. Still, measuring and treating pain is one of the most difficult and complex healthcare issues. Together with Boston Scientific, a leading medical device technology company, we have developed a new method to measure pain. Our team of neuroscientists, …
Model Analyzes How Viruses Escape The Immune System
One reason it’s so difficult to produce effective vaccines against some viruses, including influenza and HIV, is that these viruses mutate very rapidly. This allows them to evade the antibodies generated by a particular vaccine, through a process known as “viral escape.” MIT researchers have now devised a new way to computationally model viral escape, …
Work At Warp-Speed In The BigQuery UI
Data analysts can spend hours writing SQL each day to get the right insights. So it’s crucial that the tools in the Google Cloud Console make that job as easy and as fast as possible. Now, we’re excited to show you how BigQuery’s Cloud Console UI has been updated with radical usability improvements for more …
How To Get More Electric Cars On The Road
A new study from researchers at MIT uncovers the kinds of infrastructure improvements that would make the biggest difference in increasing the number of electric cars on the road, a key step toward reducing greenhouse gas emissions from transportation. The researchers found that installing charging stations on residential streets, rather than just in central …