We strive to inform readers not only about the events in the bitcoin industry, but also to talk about what is happening in related technological areas – cybersecurity, and now in the world of artificial intelligence (AI).
Cryplogger AI collected the most important news from the world of AI over the past week.
- The government of the Russian Federation approved the plan of development of AI in regions.
- Human rights activists have discovered in New York 25,000 cameras watching the townspeople in public space.
- In South Korea, they created a deepfake presidential candidate to draw the attention of young people to the elections.
- A member of the Idaho House of Representatives drafted a bill to ban AI from taking on a “personality.”
- Eric Schmidt has launched a $125 million fund for “problem AI” research.
- Study: People trust synthesized faces more than real ones.
- Fast food restaurant chain White Castle will hire robo-cooks in 100 establishments.
- An enthusiast has created a Pokemon neurogenerator.
- The most important AI deals of the last week.
Russian regions and the Ministry of Digital Development have developed an AI development plan
Regional authorities of the Russian Federation together with the Ministry of Digital Development developed roadmap for AI development in 2022. The document was preliminarily approved by the office of Deputy Prime Minister Dmitry Chernyshenko.
The action plan contains 17 points. It includes the collection of anonymized data by the regions for training algorithms and providing access to them to local authorities. The document also provides for the development of computing resources and the development of mechanisms for creating research and production consortiums for testing AI.
Among the performers in the roadmap are Sberbank, the Ministry of Digital Development, the Ministry of Economy and the Ministry of Finance.
By December 2022, the regional government needs to select the best AI projects.
The plan’s budget has not yet been approved.
Study: New York authorities monitor citizens with 25,000 cameras
In New York revealed about 25,000 facial recognition cameras, most of them in ethnic minority areas. Experts said it “disproportionately threatens the rights of the city’s non-white residents.”
In January 2021, Amnesty International launched the Ban the Scan campaign, in which 7,000 volunteers checked 45,000 intersections and discovered 25,500 recorders. According to the report of the researchers, about 3300 of them are used by the authorities and law enforcement officers.
Human rights activists have found that the Bronx, Brooklyn and Queens have the largest number of devices in areas with a high concentration of ethnic minorities.
Concentration of CCTV cameras in all five New York City boroughs. Data: Amnesty International.
Activists also explored the relationship between the number of cameras and police searches. They assigned each precinct a “surveillance level” according to the number of state cameras per 1,000 inhabitants. Areas with high levels were more likely to be subjected to stop-and-search police practices.
Idaho bans artificial intelligence from taking on a “personality”
State Committee of the Idaho House of Representatives (USA) approved a bill banning artificial intelligence from gaining “personality” status. The document has already been put to the vote.
The project aims to prevent any future efforts to give technology, inanimate objects or animals the same legal rights as humans.
According to Rep. Tammy Nichols, some organizations want to give non-human objects the status of “personality” to restrict access to certain natural resources.
“Children cannot be inferior to artificial intelligence or become on a par with water bodies and trees,” she added.
South Korea created a deepfake presidential candidate
In South Korea created a deepfake of a presidential candidate to attract young voters to the March 9, 2021 elections.
The developers used about 20 hours of audio and video content with over 3,000 recorded sentences from politician Yoon Seok Yeol. They created a digital AI Yoon avatar that looks like a real candidate and knows how to joke.
In a month and a half, the video has collected over a million views and raised many political issues. One user asked the AI who it would save if the incumbent president of the country and a rival candidate were drowning. AI Yoon replied that he would “wish both of them good luck”.
Eric Schmidt Launches $125M AI Research Fund
Former Google CEO Eric Schmidt will launch a $125 million charitable foundation to fund AI research.
Called AI2050, the project aims to address “hard problems” in the field, including algorithm bias, misuse, societal harm, geopolitical conflicts, and the scientific limitations of technology. The goal of the initiative is to develop AI by 2050 that is generally agreed to benefit humanity.
The fund, chaired by Schmidt and Google’s new head of technology and society, James Manica, will provide funds for five years to individual non-corporate scientists.
In New York, artificial intelligence issued the first fines to owners of noisy cars
New Yorkers started receive notifications from the Department of Environmental Conservation (DEC) about exceeding the permissible level of vehicle noise.
Cameras and microphones installed in the city automatically record the volume of passing vehicles. If the level of noise emitted by the car exceeds the permissible value, the algorithm recognizes the license plate of the car and sends a notification to the driver by mail.
The document contains an instruction to the owner to bring the car to a place designated by the department for inspection and installation of a muffler. In case of refusal, the violator faces a fine of up to $875 and additional penalties for ignoring the subpoena.
According to DEC, on June 30, 2022, New York authorities will reconsider the feasibility of using the system, as a result of which it will be expanded or discontinued.
Google Docs introduces AI annotator
Google added in the service for working with text documents, the function of automatically creating summaries.
The tool uses artificial intelligence to analyze multi-page text, highlight important information and generate a summary. The user can also create a summary manually.
Study: People trust synthetic faces more than real ones
Scientists from Lancaster University and the University of California at Berkeley conducted a study and found that synthesized faces cause people have more trust than real ones.
The researchers asked respondents to distinguish the faces generated by the StyleGAN2 algorithm from photos of real people.
In the first experiment, 315 participants classified 128 images out of 800 available. Their accuracy was 48%. In the second, 219 new subjects, trained in advance to recognize fakes, checked 128 pictures from the same dataset. Their accuracy increased to 59%.
In the third experiment, 223 respondents were asked to rate 128 faces on “reliability” from 1 (very unreliable) to 7 (very reliable). As a result, the level of confidence in the synthesized images was 7.7% higher.
Burger chain White Castle has hired robot chefs
In 2022 Miso Robotics will launch Flippy 2 robot chefs in 100 White Castle burgers.
The device, using AI, determines the type of food, places it in the right fry basket, cooks it and puts it in the hot storage area.
According to the company, the Flippy 2 is capable of handling 60 baskets an hour on its own, or about 300 hamburgers a day.
Enthusiast created a Pokemon generator
Machine learning engineer developed Nokémon neuropokemon generator.
He used the model to create pictures from the DALL-E text description from OpenAI. The engineer trained the algorithm on images of all Pokémon, indicating their attribute types.
In the web application, users can specify a specific character or element, after which the AI will generate a suitable option. We at Cryplogger AI chose Slowpoke as the basis, and this is what Nokémon came up with.
The most important AI deals of the last week
From February 13 to February 19, 2022, artificial intelligence startups raised over $166 million. Here are the most interesting transactions.
- Overview attracted $10 million to support an AI platform for visual defect detection on a production line.
- Exodigo collected $29 million to expand the team of software developers to create three-dimensional maps of underground spaces and structures.
- Creators of the service for collecting and processing production data Instrumental received $50 million
- Raydiant collected $30 million to accelerate the creation of digital signage for stores.
- Nanonets attracted $10 million to scale document processing software.
- Creators of digital AI showcases for online creators Voila received $6 million
- DuploCloud collected $15 million for the development and marketing of IaC products.
- atommerce attracted $16.7 million to scale the mental health platform.
Also on Cryplogger:
What to read on the weekend?
In the new Cryplogger AI educational cards, we tell you what recommender systems are, how they work, where they are used, what advantages and disadvantages they have.
Subscribe to Cryplogger news in Telegram: Cryplogger AI – all the news from the world of AI!
Found a mistake in the text? Select it and press CTRL+ENTER