In the News
Mojo Wang After Constance Gemson moved her mother to an assisted living facility, the 92-year-old became more confused, lonely and inarticulate.
So on a visit one day, Ms. Gemson brought her mom a new helper: a purring, nuzzling robot cat designed as a companion for older adults.
The AI Hardware Summit, Munich, 10-11 March 2020, welcomes hyperscalers and leading enterprises deploying AI hardware and design systems to the stage, to share their processing requirements, along with solutions to the challenges of accelerating machine learning at the edge.
In The News
Instead, while being clear-eyed about China’s aggressive pursuit of AI for military use and human rights-abusing technological surveillance, the United States and China must find their way to dialogue and cooperation on AI.
Trustworthy AI has to start with good engineering practices, mandated by laws and industry standards, both of which are currently largely absent.
Excerpted from “Rebooting AI: Building Artificial Intelligence We Can Trust”, by Gary Marcus and Ernest Davis.
To produce a better experience for hearing aid wearers, my lab at Ohio State University, in Columbus, recently applied machine learning based on deep neural networks to the task of segregating sounds.
Why it matters: This debate will define the future of the controversial AI systems that help determine people’s fates through hiring, underwriting, policing and bail-setting.
Bender argued in an essay last week that some technical research just shouldn’t be done.
The panel included Yoshua Bengio, MILA director and University of Montreal professor; Jeff Dean, Google’s AI chief; Andrew Ng, cofounder of Google Brain and founder of Landing.ai; and Cornell University professor and Institute for Computational Sustainability director Carla Gomes.
It has been an incredibly busy year for Yejin, presenting at an 13 conferences in 2019, discussing her latest work, which includes that of her primary research interest areas of Natural Language Processing, Machine Learning & Artificial Intelligence, alongside her broader interest fields of Computer…
Applied use cases
Our method uses these models to synthesise hypothetical behaviours, asks the user to label the behaviours with rewards, and trains a neural network to predict these rewards.
To encourage exploration during reward model training, ReQueST synthesises four different types of hypothetical behaviours…
Further Reading Cloudy with a chance of neurons: The tools that make neural networks work In an earlier deep learning article, we talked about how inference workloads—the use of already-trained neural networks to analyze data—can run on fairly cheap hardware, but running the training workload that…
Being branches of the same field, the terms artificial intelligence (AI), machine learning (ML), deep learning (DL), and natural language processing (NLP) are used interchangeably.
Several common themes have emerged such as cobots, emerging energy source, AI, and cybersecurity breaches.
Robots excel at carrying out specialized tasks in controlled environments, but put them in your average office and they’d be lost.
Alphabet wants to change that by developing what they call the Everyday Robot, which could learn to help us out with our daily chores.
Generative Teaching Networks: Accelerating Neural Architecture Search by Learning to Generate Synthetic Training Data
We in Uber AI Labs investigated the intriguing question of whether we can create learning algorithms that automatically generate training data, learning environments, and curricula to help AI agents rapidly learn.
Using a unique neurosymbolic approach that borrows a mathematical theory of how the brain can encode and process symbols, we at Microsoft Research are building new AI architectures in which neural networks learn to encode and internally process symbols—neural symbols.
Generative machine learning and machine creativity have continued to grow and attract a wider audience to machine learning.