Algorithm Flags Biased Journalists, Israel Develops Combat Robot, and Other AI News

We aim to inform readers not only about events in the Bitcoin industry but also about developments in adjacent technology spheres — cybersecurity, and now the world of artificial intelligence (AI).

ForkLog has gathered the most important AI news from the past week.

Deputy Prime Minister Chernyshenko Heads the AI Funding Distribution Group

The Russian government appointed Deputy Prime Minister Dmitry Chernyshenko as head of the working group for AI research centers.

The group has been created to select grant recipients. The total funding from the federal budget amounts to 5.4 billion rubles.

In early October the group will select programs for at least six centers aimed at developing applied AI solutions for deployment and transformation of industries.

The EU Will Develop Its Own “Chip Act”

The president of the European Commission, Ursula von der Leyen, announced the creation of a “European Chips Act” to bolster resilience and sovereignty of regional semiconductor supply chains.

The project will include research in the field, the development of a joint plan to increase production capacity, and a framework for international cooperation and partnerships. The law will also draw on other digital initiatives, including curbing the influence of tech giants and increasing platform accountability, regulating AI and fighting misinformation online.

Biden Announced Creation of a Technology Alliance Between the United States, Britain and Australia

US President Joe Biden announced the creation of the AUKUS defence pact with the United Kingdom and Australia for the mutual exchange of advanced technologies.

The collaboration will simplify the sharing of information across areas such as AI, cybernetics, quantum technologies and undersea systems.

According to Biden, the partnership will protect the participants’ interests in the Indo-Pacific region and will enable Australia to build its own nuclear-powered submarines.

Chinese authorities reacted very negatively to the creation of the new bloc. Although leaders did not name specific states, “threatening stability in the region,” Beijing took the statements as directed at it.

In Kyiv, AI Cameras Deployed to Optimise Traffic

Kyiv City State Administration installed 31 AI-powered cameras at complex intersections to monitor traffic congestion.

The AI analyses data from the devices and adjusts traffic flows to ease congestion and optimise municipal transport. In the event of an accident or unforeseen obstacle, the system automatically notifies the police or municipal services.

By the end of 2021 Kyiv authorities plan to connect another 43 AI cameras. They also plan to optimise traffic-signal control in the future.

Apple Unveils the A15 Bionic

Apple, at its product-launch event, introduced the iPhone 13 smartphones and iPad tablets powered by the 5-nanometer A15 Bionic system-on-a-chip with 15 billion transistors.

The chip features two high-performance and four efficiency cores, plus a 4-core GPU and a 16-core Neural Engine designed to accelerate AI tasks. It increases AI operations to 15.8 trillion, up from 11 trillion in the A14 Bionic.

The Neural Engine could prove effective for Siri voice synthesis, image recognition, facial focus in photos, and unlocking the phone with Face ID.

Apple says the A15 Bionic is 50% faster and its graphics are 30% more powerful than the nearest competitor. It is presumably referring to Qualcomm’s Snapdragon 888.

Tech Giants Developing AI Technologies for the Pentagon

US tech corporations are partnering with the US Department of Defense under multimillion-dollar contracts to develop object-recognition algorithms for imagery from military drones, researchers from Tech Inquiry reported.

Microsoft won a $30 million contract and will provide the Pentagon with tools for analyzing data on “full-motion video.”

Amazon signed a $20 million contract to develop algorithms for detecting and classifying objects in “moving video.”

IBM won a $5.6 million contract to build “statistical reasoning models in conjunction with AI.”

Google was also involved in government contracts, but in 2018 stopped cooperating with the Pentagon due to employee pressure. Its task was to develop computer-vision algorithms to analyse drone footage.

Artificial Intelligence Exposes Bias Among Major Media Journalists

Researchers from MIT used AI to identify bias among journalists at the world’s leading English-language news outlets.

The authors analysed over 3 million articles. They removed direct quotes, changed British spelling to American, and converted uppercase to lowercase. Using natural-language processing, the researchers identified the first 100,000 most common phrases, ranked them, de-duplicated and combined into a list.

Using the Black Lives Matter topic as an example, researchers found that “protesters” shift from “anarchists” to “rioters” as they move through the political stance of the outlet in question.

WhatsApp to Add Voice-Message Transcription

Facebook will add to WhatsApp on iOS a voice-message transcription feature. The tool will operate locally on the device, with no data sent to the company’s servers.

The pop-up notes that the information will be sent to Apple. The voice messages will assist it in developing speech-recognition technology, but will not be directly linked to the user’s identity.

Israel Aerospace Industries Develops Combat Robot to Defend Infantry in Battle

Israel Aerospace Industries has developed the Rex MK II unmanned robot to protect ground troops in field conditions.

The device is controlled via an electronic tablet and equipped with two machine guns, cameras and various sensors. It collects reconnaissance data, carries wounded soldiers and delivers ammunition. The robot can also patrol combat zones, track intruders and engage nearby targets.

Russia Launches Pet-Finding Service

The Goznak company launched an online service, “Tail of Luck” (Tail of Luck), which uses AI to search for lost domestic animals. It is free and available to anyone after registration.

Users must upload a photo of the pet and indicate the date and time of its disappearance or discovery. The algorithm will then analyse hundreds of thousands of images and compare them with other listings.

AI recognition accuracy is 97% for dogs and 96.5% for cats.

Also on ForkLog:

• Researchers explained, how to tell a deepfake from a real photo.
• Biden nominated an opponent of facial recognition to the FTC.
• US authorities requested data from 12 automakers in the case against Tesla’s autopilot.
• The UN called for a ban on certain kinds of AI.
• Researchers fooled the facial recognition system with natural makeup.

Subscribe to ForkLog news on Telegram: ForkLog AI — all the AI news!