Elon Musk to Create a Social Network, Targeted Advertising on Facebook Fails, Yandex is Sanctioned and Other News From the AI World

More and more technologies use artificial intelligence for development. And accordingly, the level of quality of AI is growing depending on the needs and goals of the tech industry. Here are just a few of the latest AI news.

Yandex Felt the Power of Sanctions

As a result of the sanctions imposed against Russia due to its invasion of Ukraine, the tech giant Yandex may run out of semiconductors and other components needed to run servers by the end of 2022.

According to Bloomberg, the shortage of equipment at the Russian tech giant is growing every day. International sanctions against Russia in the field of dual-use technology, used in the military and civilian spheres, have hit the drone development unit the hardest.

In addition, they may affect other areas of Yandex’s activities, such as a search service, cloud business, music streaming and e-commerce segment.

In addition to the lack of raw materials in the technical field, Russia also began to experience a shortage of personnel. Many IT specialists simply fled the Russian Federation. Since the beginning of the war in Ukraine, up to 70,000 specialists have left Russia, and in April this number could reach 170,000 people.

China Has Developed a Roadmap For the Development of AI

100 Chinese researchers have developed a 200-page roadmap for AI in the “Big Model Era” with a bibliography of 1,637 titles.

In this research work, three main problems of modern artificial intelligence systems that need to be solved are considered. Firstly, the models are not portable – the algorithm is built and trained for a specific task, it cannot be used to solve another problem, therefore, you have to start training the neural network again.

Secondly, model training, including parameter tuning, requires the manual work of a large number of AI engineers. And, thirdly, the lack of high-quality data, which limits the application of algorithms in some areas.

The document also considers the possibility of eliminating the “main risk of AI algorithms”, since all their defects extend to subsequent neural networks and can be transferred to the entire community of basic models.

Elon Musk Plans to Create a Social Network Based On AI

American businessman Elon Musk thought about creating a social network with open source AI algorithms and respect for freedom of speech.

To one of the posts from a Twitter user about the development of a new platform, Musk replied that he was seriously thinking about it.

Musk also asked his followers if they believe that the social network (Twitter) adheres to the principles of free speech. More than 70% of users out of about 2 million who participated in the survey answered negatively.

Nvidia Converts 2D to 3D

Nvidia demonstrated the work of the Instant NeRF neural network, which turns several dozen 2D photos into a three-dimensional 3D image.

Developed in 2020 by American researchers in collaboration with Google Research, the NeRF method compares the colors and light intensity in photographs taken from different angles, taking into account information about camera angles, and then combines the collected data into a single 3D picture. For high-quality detailing of the final image, a larger number of two-dimensional images is needed.

According to Nvidia, Instant NeRF only takes a few seconds to train compared to other neural networks that take several hours. This technology could be used to train robots and drones in the future.

Boston Dynamics Launches Open Sale of Stretch Robot

Boston Dynamics is opening the sale of Stretch loader robots to everyone.

Stretch is a conveyor with a manipulator with seven degrees of freedom and equipped with suction cups. Thanks to built-in cameras and sensors, the robot can freely navigate in space and move cargo around the warehouse, as well as unload it from trucks.

The company has received orders for all of 2022, according to Boston Dynamics. But now the company is accepting orders for the supply of a new batch for 2023 and 2024.

Facebook Ads Ignore the Context of User Interests

According to a study by North Carolina State University scientists on how Facebook’s advertising algorithms work, interest patterns are created based on user actions, without regard to context.

For example, if a user writes that they hate cheese, the AI ​​will determine this as an interest in cheese, without taking into account the negative attitude towards it, and launch targeted ads about cheese.

The scientists conducted two experiments. In the first test, they registered 14 new accounts, where they controlled demographic data and tracked a generated list of preferences. The researchers found that 33.22% of the suggested interests were inaccurate or irrelevant.

For the second experiment, the researchers recruited 146 participants from around the world. They collected data from each account and analyzed the AI-generated preference list. As a result, 29.3% of interests did not correspond to the real preferences of users.

According to the researchers, the majority of respondents were not even aware of the existence of the Facebook Ads Preference Manager.

AI helps Google Maps Clean up Offensive Edits

In 2021, Google Maps blocked more than 100 million offensive edits on maps using artificial intelligence.

According to the company, thanks to the collaboration of human operators with AI, about 7 million fake profiles were removed, 12 million attempts of fraudsters to impersonate the company and 8 million cases of false claims claiming to be an enterprise account were prevented.

In addition, the Google Maps team removed almost 200 million low-quality or violating photos and videos.

AI Helps to Create Weather Forecasts

Camaliot, funded by the European Space Agency, developed an Android application that uses GPS receivers to collect information on time and satellite positions. The system then notes the time spent receiving messages and uses this data to determine the location of the smartphone.

According to the researchers, in this way it will be possible to make more accurate weather forecasts due to more information about the atmosphere, in particular, how the amount of water vapor affects the propagation of satellite signals through the air.

The BTC-Alpha team together with trusted volunteer communities created a charity fund ‘‘Alpha for Ukraine’. You can make a donation to help everyone in Ukraine who needs medicines, humanitarian aid and to support the defence of the country.

Post a Comment