DeepMind has looked to developmental psychology to help AI gain a basic understanding of the physical world.

Real-world physics are for AIs to grasp when asked to start from scratch with only training data to guide them. But researchers have demonstrated babies as young as five months are surprised if they are shown a physically impossible event, such as a toy suddenly disappearing, implying they gain some intuitive physical understanding at an early age.

DeepMind researcher Luis Piloto and his colleagues developed an AI, dubbed PLATO, which adopts the thesis that objects play a central role in the representation and prediction of the physical world around us.

They then used video training data to showing it videos of many simple scenes to improve PLATO's performance. PLATO reacted similarly to how a baby expresses surprise when seeing an impossible event, and learning effects were seen after 28 hours of videos, according to a paper published in Nature today.

Piloto explained that PLATO uses objects at all stages of processing: representing visual inputs as a set of objects; reasoning about interactions between objects; and producing outputs that are predictions on a per-object basis.

"We found that PLATO passed the test in our physical concepts dataset," he said. "But when we trained flat models that were as big or even bigger than PLATO – but didn't have object-based representations – we found that they didn't actually pass all of our tests suggesting that objects really are a critical part of physical understanding."

In an accompanying article, Susan Hespos, professor of psychology at Northwestern University, and Apoorva Shivaram of Western Sydney University, Australia, said the finding confirmed understanding of the development of perception in humans as well as advancing AI.

"The findings indicate that visual animations can account for some intuitive physics learning, but not enough to account for what we see in infants. In other words, the computational models require some principled knowledge about how objects behave and interact to match the level of learning that is commonly seen in infants," they said.

While a wealth of AI applications might benefit from a bit of real-world physics – self-driving cars anyone? – the authors stressed that the finding was really about helping other studies in AI.

Piloto told journalists: "I think physical understanding is pervasive. It's hard to talk about specific applications because we think it's a little bit more general than that. It really kind of depends on what researchers want to do with it. The point of this work is to establish a benchmark so that people [realize] how well their models understand the physical world. We don't have a view on what they want to do beyond that point." ®

Computer-science departments across US universities do not have enough lecturers to teach increasing numbers of students interested in AI, a report from the Center for Security and Emerging Technology (CSET) this month suggested.

Interest in machine learning and artificial intelligence has risen and fallen since the field was formally founded in the 1950s. Neural networks have made a comeback in recent years, exploding in popularity with deep learning. Demand for machine-learning courses at universities has skyrocketed, we're told, and there aren't enough lecturers to support students' interest.

Data compiled by the Taulbee survey, and quoted in the report, showed that between 2011 and 2020, the number of students enrolled in computer-science programs in America tripled from 60,661 to 182,262. But the number of faculty in computer-science departments increased under 1.5X from 4,363 to 6,230. The aggregate student-to-faculty ratio across the surveyed departments doubled up from 14-to-1 to 29-to-1. 

Comment More than 250 mass shootings have occurred in the US so far this year, and AI advocates think they have the solution. Not gun control, but better tech, unsurprisingly.

Machine-learning biz Kogniz announced on Tuesday it was adding a ready-to-deploy gun detection model to its computer-vision platform. The system, we're told, can detect guns seen by security cameras and send notifications to those at risk, notifying police, locking down buildings, and performing other security tasks. 

In addition to spotting firearms, Kogniz uses its other computer-vision modules to notice unusual behavior, such as children sprinting down hallways or someone climbing in through a window, which could indicate an active shooter.

Opinion If the soap opera of Microsoft's relationship with open source had a theme tune, it'd be "The Long and Winding Goad".

To a company whose entire existence depended on market control, open source's radical freedoms were an existential, cancerous threat. In return, open source was only too happy to play the upstart punk movement to Microsoft's bloated prog rock.

In the end, both sides accepted the inevitable. Redmond wasn't going to control the cloud and mobile the way it controlled business IT, and the cloud and mobile loved open source. Interoperability was more profitable than insults. For its part, open source was, well, open. It couldn't stop Microsoft's newfound friendliness so wary acceptance became the new world order.

Comment Future AI could be a challenge for US Patent and Trademark Office (USPTO) officials, who need to wrap their heads around complex technology that's perhaps not quite compatible with today's laws.

Under the Department of Commerce, the USPTO's core mission is to protect intellectual property, or IP. Creators file patent applications in hope of keeping competitors from copying their inventions without permission, and patents are supposed to allow businesses to thrive with their own novel designs while not stifling wider innovation.

Fast evolving technologies, such as deep learning, are pushing the limits of today's IP policies and rules. Clerks are trying to apply traditional patent approval rules to non-trivial machine-learning inventions, and bad decisions could result in a stranglehold on competition among public and private AI creators. We all know how overly broad patents on software and other technology can make it past USPTO, causing headaches for years to come.

Meta's quest to translate underserved languages is marking its first victory with the open source release of a language model able to decipher 202 languages.

Named after Meta's No Language Left Behind initiative and dubbed NLLB-200, the model is the first able to translate so many languages, according to its makers, all with the goal to improve translation for languages overlooked by similar projects. 

"The vast majority of improvements made in machine translation in the last decades have been for high-resource languages," Meta researchers wrote in a paper [PDF]. "While machine translation continues to grow, the fruits it bears are unevenly distributed," they said. 

In brief US hardware startup Cerebras claims to have trained the largest AI model on a single device powered by the world's largest Wafer Scale Engine 2 chip the size of a plate.

"Using the Cerebras Software Platform (CSoft), our customers can easily train state-of-the-art GPT language models (such as GPT-3 and GPT-J) with up to 20 billion parameters on a single CS-2 system," the company claimed this week. "Running on a single CS-2, these models take minutes to set up and users can quickly move between models with just a few keystrokes."

The CS-2 packs a whopping 850,000 cores, and has 40GB of on-chip memory capable of reaching 20 PB/sec memory bandwidth. The specs on other types of AI accelerators and GPUs pale in comparison, meaning machine learning engineers have to train huge AI models with billions of parameters across more servers.

Kakao has backed down from what appeared to be a standoff with Google regarding external payment methods following the suspension of updates to its popular messaging app KakaoTalk this month.

As of last month, Google instituted a new payments policy requiring developers selling digital goods and services to use its first-party billing system. The company said apps using an in-app billing system as an alternative to Google Play's would need to remove it.

"Starting June 1, 2022, any app that is still not compliant will be removed from Google Play," said Google.

UK telecoms giant Vodafone is deploying an MLOps service within the organization, with the help of Google Cloud.

Called AI Booster, the project is designed to automate and standardize building and distributing machine-learning models within the carrier.

Over the past 18 months Voda has employed Google tools BQML, AutoML, and Vertex AI as well as working with Google Cloud partner Datatonic.

Google has placed one of its software engineers on paid administrative leave for violating the company's confidentiality policies.

Since 2021, Blake Lemoine, 41, had been tasked with talking to LaMDA, or Language Model for Dialogue Applications, as part of his job on Google's Responsible AI team, looking for whether the bot used discriminatory or hate speech.

LaMDA is "built by fine-tuning a family of Transformer-based neural language models specialized for dialog, with up to 137 billion model parameters, and teaching the models to leverage external knowledge sources," according to Google.

In brief Numerous people start to believe they're interacting with something sentient when they talk to AI chatbots, according to the CEO of Replika, an app that allows users to design their own virtual companions.

People can customize how their chatbots look and pay for extra features like certain personality traits on Replika. Millions have downloaded the app and many chat regularly to their made-up bots. Some even begin to think their digital pals are real entities that are sentient.

"We're not talking about crazy people or people who are hallucinating or having delusions," the company's founder and CEO, Eugenia Kuyda, told Reuters. "They talk to AI and that's the experience they have."

The Register - Independent news and views for the tech community. Part of Situation Publishing

Biting the hand that feeds IT © 1998–2022