MIT Technology Review
The Download from MIT Tech Review
06.17
Good morning! Here are today's most important stories in emerging technology.
A prototype AI tool from Adobe automatically spots Photoshopped faces
Adobe Faces
Its algorithm correctly identified edited faces 99% of the time during tests, compared to a 53% success rate for humans.
The research: Researchers from Adobe and UC Berkeley have created a tool that uses machine learning to automatically identify when photos of people’s faces have been altered. The deep learning tool was trained on thousands of images scraped from the internet. In a series of experiments, it was nearly twice as good at identifying manipulated images compared to humans.
The context: There’s growing concern over the spread of fake images and ‘deepfake’ videos. However, machine learning can be a useful weapon in the detection (as well as the creation) of fakes, as this research demonstrates.
Some caveats: It’s understandable that Adobe wants to be seen to act on this issue, given its own products are used to alter pictures. The downside is that this tool only works on images that were made using Adobe Photoshop’s Face Aware Liquify feature. However, it’s just a prototype, and the company says it plans to take this research further, providing tools to identify and discourage the misuse of its products across the board.
The US claims it has inserted offensive malware into Russia’s power grid
It’s a step up from previous work which focused more on reconnaissance than the ability to attack.
The news: The US has deployed code inside Russia’s power grid and other targets as part of its work to push back against hacking and disinformation from Moscow, the New York Times reports. The revelation comes just three days after the news that a Russia-linked group called Xenotime has started laying the early groundwork for potential attacks on US power companies.
The details: Inevitably, there are few specifics given the sensitivity of the topic. However, the work is being led by the US Cyber Command, an arm of the Pentagon. It’s permitted by a new law passed last summer which approves “clandestine military activity” in cyberspace to “deter, safeguard or defend” against attacks.
A warning: The US hopes that these revelations will act as a deterrent against any future attacks by Russia on its systems. But there’s a risk that it could fuel an escalation, instead.
Facebook has built stunning virtual spaces for its AI programs to explore
Researchers at Facebook have created a number of extremely realistic virtual homes and offices so their AI algorithms can learn how the real world works.
Real deal: The team created 18 “sample spaces” using a program known as Replica. The idea is for AI agents to learn about real-world objects through exploration and practice. In theory, this could make chatbots and robots smarter, and improve VR in powerful ways.
Home alone: These virtual spaces can be loaded into a new environment called AI Habitat, inside which AI programs can explore and learn. The algorithms will first be trained to recognize objects in different settings. But over time they should build some common-sense understanding about the conventions of the physical world—like the fact that tables typically support other objects.
Uncommon sense: A lack of common sense is a glaring problem for today’s AI systems. Unlike a person, a chatbot or robot cannot rely on an understanding of the world—things like physics, logic, and social norms—to figure out the intent of an ambiguous command.
—Will Knight
Sponsor Message
102403121_m
The quiet revolution: Moving artificial intelligence from niche to everywhere
Arm_logo_blue_150MD
Today, machine learning is seeping into all kinds of applications, many of them mobile. It’s the golden age for this branch of AI, using sophisticated algorithms in models that can learn from data and identify important patterns. By uncovering connections, ML helps businesses make better decisions without the need for human input.
Read the full article today
The top ten must-reads
I’ve combed the internet to find you today’s most fun/important/scary/fascinating stories about technology.
1 Volvo’s new self-driving truck has started its first project in Sweden 🚚
Unusually, it’s fully autonomous—there’s nowhere for a human safety driver to sit. (TechCrunch)
+ A new law in Florida permits self-driving car tests without a human at the wheel. (Engadget)
+ Self-driving cars have a problem: tech is making human-driven vehicles safer. (WSJ)
2 How Hong Kong’s protesters are using tech to keep low profile
Wary of being tracked by China’s authorities, many are now communicating solely using secure messaging apps. (The Guardian)
+ Telegram’s boss blamed China for a cyberattack during the Hong Kong protests. (TR)
3 AI can tell when actors are kissing—and maybe when you are, too
This study shows how quickly the ability to analyze footage for specific actions has advanced. (TR)
+ Construction firms are developing an AI system that predicts worksite injuries. (TR)
+ What can AI tell us about fine art? That the most beautiful images aren’t always the most memorable. 🖼️ (IEEE Spectrum)
4 A lyrics site has accused Google of stealing its content
‘Information boxes’ at the top of searches might be handy for users, but mean fewer clicks (and $) for owners of sites like Genuis. (WSJ)
5 If GPS stopped working, it would cost the US economy $1 billion per day
This study shows just how much we’ve come to rely on the service. (Ars Technica)
6 Huawei has delayed the launch of its foldable phone until September
It says its doing extra testing after Samsung’s disastrous experience with the Galaxy Fold. (CNBC)
+ Huawei is bracing itself for a drop in phone sales of up to 60 million. (Bloomberg)
7 How Amazon cloned a neighborhood to test its delivery robots
Its digital copy mirrors the position of curbstones and driveways within inches. (Wired)
+ Amazon Spark, the company’s two-year-old Instagram competitor (ever heard of it?!) has closed down. (TechCrunch)
+ Does Amazon really pay no taxes? The answer is complicated. (WSJ)
8 Russia interfered in recent European elections, says the EU
The aim was partly to discourage people from voting. (The Verge)
9 How companies track your every move inside stores 👀🏪
Bluetooth beacons can detect how long you lingered in front of those potato chips. (NYT)
10 That video of a robot getting beaten up is fake—but didn’t you still feel sorry for it?
The clever parody raises some interesting questions. (The Verge)
Hear from technology and business leaders driving the new global economy on-stage at EmTech this September.
We're curating a program that brings our editorial coverage to life and highlights the executives, technologists, and innovators you need to know.
Download entire message
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment