Meta-owned Facebook has introduced a "kill switch" that lets users decide whether or not they want their data to be used in the company's artificial intelligence technology.
Within Facebook's Help Center, users can access a form asking Meta to delete, inspect, or edit information from an external source (third party) about themselves that can feed the artificial intelligence (AI) for training.
Meta defines third-party information as data that is publicly available on the Internet or obtained through licensed sources. According to the company, this category of information forms part of the vast “billions of data points” used to train generative AI models, which use predictive capabilities and patterns to create new content.
Amid the rise of machine intelligence, technology companies aim to create the most advanced and world-leading artificial intelligence, and are therefore searching for "treasures" of training data.
This decision to opt out comes in the wake of a joint statement issued by a group of global data protection agencies last week, in which they urged against data theft and called for greater protection of user privacy with regard to the companies “Meta” and “Alphabet”, which owns “Google” and... Microsoft".
Meta acknowledged that blog posts may contain personal details such as names, contact information and work details, which were collected by the company to train its generative AI technology.
It is important to note that the data used is already "publicly available on the Internet or licensed sources," according to the Meta blog post.
The company said this type of information could represent some of the "billions of data" used to train generative AI models that "use predictions and patterns to create new content."
Users can now limit the spread of that data with a section titled Generative AI Data Subject Rights, which enables users to "submit objections to the use of their third-party information to train generative AI models."
It is still unclear whether users' personal data includes their activities on other Meta-owned platforms, including their public comments and photos on Instagram.
A Meta spokesperson said: “Depending on where people live, they may be able to exercise their data subject rights and object to the use of certain data to train our AI models. They can submit an objection form to us through the Privacy Center link.”
“X”(twitter) intends to collect biometric data, education and work history of users for artificial intelligence
New York: The “X” social networking platform owned by businessman Elon Musk , which was formerly known as “Twitter”, has expanded its user data collection operations to include biometric information.
What kind of information is being collected - and how the platform intends to use it - is of interest to tech watchers.
The company recently added to its online privacy policy: “We may use the information we collect to help train machine learning or artificial intelligence models for the purpose specified in this policy.”
But these purposes are not entirely clear - except for authentication and service operation - and X's definition of biometric data is incomplete in terms of fine details. The US Department of Homeland Security defines biometrics as “distinctive physical characteristics, such as fingerprints, that can be used for automatic identification.”
X intends to record information about its users' education level and work history. This data will be used, according to the advanced company, Mask, in order to make recommendations for jobs. The change in methods of collecting information on the platform will begin at the end of this September.
Switzerland : A drone controlled by artificial intelligence outperforms human heroes for the first time
Paris: A drone controlled by artificial intelligence was able to outperform the champions of drone races from a distance for the first time, according to a study published in the journal Nature, paving the way for improving the systems used in self-driving cars or industrial robots.
“I was so close to the self-driving parade that I felt its turbulence as I tried not to let go,” said Alex Vanover, one of three champions in the field drawn from the Robotics and Cognition group at the University of Zurich in Switzerland to confront the winning rally.
The race was held on a 75-meter circuit consisting of seven gates, which are large blue tires that must be crossed in a specific order, over three laps. The race took place with machines that easily reached speeds of 100 kilometers per hour, an ability to increase speeds that exceed even what happens in Formula One races, and turns on 180-degree turns.
The three men, including a former world champion drone racer, had a week of training, equipped with helmets that transmitted images from their drones to drive.
The car won the majority of its races against each of them, and completed the fastest lap on the track. This is the first time that "an autonomous mobile robot has achieved world champion-like performance in a competitive sport in the real world," according to the study.