PyTorch now has an independent foundation

Meta-created framework PyTorch now has an independent foundation to drive the adoption of AI tooling.

The PyTorch Foundation will operate under the Linux Foundation umbrella and boasts a governing board with representatives from industry giants including AMD, Microsoft Azure, NVIDIA, Amazon Web Services, Google Cloud, and Meta.

“The new PyTorch Foundation board will include many of the AI leaders who’ve helped get the community where it is today, including Meta and...

Xbox’s Matt Booty wants to see QA testers replaced by AI

Xbox Game Studios head Matt Booty told a live audience at PAX that he’d like to see QA (Quality Assurance) testers replaced by AI.

Gaming QA testers regularly come under fire as bug-riddled games reach the hands of disappointed gamers with seemingly increasing prevalence.

The backlash QA testers receive from frustrated gamers is often misplaced. Managers are often aware that games are full of bugs but decide to release unfinished games anyway in the hope of starting to...

AI discovers over 20K taxable French swimming pools

The French government is successfully using AI to discover taxable swimming pools.

IT firm Capgemini worked with Google to create the AI which analyses aerial photos taken by France’s National Institute of Geographic and Forest Information, identifies pools, and then cross-references them with national tax and property registries.

So far, the project has spotted 20,356 undeclared swimming pools. The government has used the data to collect almost €10 million in...

DARPA seeks AI solutions for sourcing critical minerals

DARPA has launched a competition to find AI solutions that can help with sourcing critical minerals.

Critical minerals are raw, non-fuel materials that are vital for manufacturing products that are essential to national security.

DARPA is teaming up with the US Geological Survey (USGS) to explore how machine learning and AI can accelerate critical mineral assessments.

“The USGS’s critical mineral resource assessments are at the heart of our domestic supply...

Jason Steer, Recorded Future: On building a ‘digital twin’ of global threats

Recorded Future combines over a decade (and counting) of global threat data with machine learning and human expertise to provide actionable insights to security analysts.

AI News caught up with Jason Steer, Chief Information Security Officer at Recorded Future, to learn how the company provides enterprises with critical decision advantages.

AI News: What is Recorded Future’s Intelligence Graph?

Jason Steer: Recorded Future has been capturing information...

Appen: 93% say ‘responsible AI’ is the foundation of all projects

Appen’s latest annual State of AI and Machine Learning report has some interesting findings about the pursuit of safe and ethical solutions.

“This year’s State of AI report finds that 93 percent of respondents believe responsible AI is the foundation of all AI projects,” said Mark Brayan, CEO at Appen.

“The problem is, many are facing the challenges of trying to build great AI with poor datasets, and it’s creating a significant roadblock to reaching their...

Vodafone and Google launch AI Booster platform

A new platform launched by Vodafone and Google called AI Booster aims to handle thousands of ML models a day across 18+ countries.

AI Booster is the result of 18 months of development and is built upon Google’s Vertex AI and integrates with Vodafone’s Neuron platform.

Vertex AI, among other Google technologies, had not been officially announced when Vodafone started development on AI Booster.

Cornelia Schaurecker, Global Group Director for Big Data & AI...

AI learns how to play Minecraft by watching videos

Minecraft characters.

Open AI has trained a neural network to play Minecraft by Video PreTraining (VPT) on a massive unlabeled video dataset of human Minecraft play, while using just a small amount of labeled contractor data.

With a bit of fine-tuning, the AI research and deployment company is confident that its model can learn to craft diamond tools, a task that usually takes proficient humans over 20 minutes (24,000 actions). Its model uses the native human interface of keypresses and mouse...

Google employs ML to make Chrome more secure and enjoyable

Google has explained how machine learning is helping to make Chrome more secure and enjoyable.

Starting with security, Google says that its latest machine learning (ML) model has enabled Chrome to detect over twice as many phishing attacks and malicious sites.

The new on-device machine learning model was rolled out in March. Since its rollout, Google claims that Chrome has detected 2.5x more threats.

Beyond security, Google is also preparing to use machine...

Apple’s former ML director reportedly joins Google DeepMind

A machine learning exec who left Apple due to its return-to-office policy has reportedly joined Google DeepMind. 

Ian Goodfellow is a renowned machine learning researcher. Goodfellow invented generative adversarial networks (GANs), developed a system for Google Maps that transcribes addresses from Street View car photos, and more.

In a departure note to his team at Apple, Goodfellow cited the company’s much-criticised lack of flexibility in its work...