Last week's tech developments saw significant stories amidst the Catholic holiday, with the likes of Google DeepMind's new AI fact checker for LLMs, SAFE, announced to the world.
This also saw a massive rise in Canada's car thefts, which took advantage of Apple's AirTags Bluetooth trackers and X's plan to move NSFW content to specific groups.
The latest developments bring insight into how technology evolves in the present, one that centers on improving the daily lives of its users globally.
Google DeepMind: AI Fact Checker for LLMs, SAFE, Is Here
Google DeepMind wants to bring more accurate and legitimate experiences with large-language models (LLMs) with a new AI fact checker for the renowned technology called 'SAFE.'
The company released its latest study on the pre-print server, arXiv, detailing more of what the technology brings for the current setup of AI companies. The Search-Augmented Factuality Evaluator, a.k.a. SAFE, uses the LLM's responses and scrutinizes them while cross-referencing the results it brings with search engine results for verification.
It was revealed that SAFE is more accurate to as much as 76 percent in the study's cases compared to a human evaluator when presented with the LLM's performance. The team applied their studies to present LLMs including Gemini, GPT, Claude, and PaLM-2.
Apple's AirTags Exploited: Canada Car Theft Rises
Apple AirTags are now exploited by car thieves in Canada, as the rising cases of theft reportedly use the renowned technology to track and monitor these vehicles before criminals make their move. This was amidst the massive campaign of the Canadian government to fight against these crimes, with various regions in the Great North seeing a rise in the tech's use.
Reports share that victims report finding AirTags in several inconspicuous areas in their cars, with car owner Ethan Yang claiming that he found one on his car's front grill in Montreal.
Vermont owners claim that other forms of GPS monitoring devices were attached to their vehicles last March.
X is Planning to Move NSFW Content to Groups
Instead of increased content filtering on the platform, X is looking to move NSFW (not safe for work) content to new groups, centering on adult and explicit content available on the social media. Other platforms either ban or allow NSFW content but is only accessible for accounts that are of legal age, a.k.a. age filters, which do not necessarily remove it.
This new feature on X was discovered recently, allowing users to create or join groups that cater to NSFW content on Elon Musk's social media, designated by creators and members as adult-service content.
Failing to label these groups as NSFW may lead to content removal, filtering, and other actions by the platform.
X's known verification may also come into play for this new NSFW groups, as it could help determine the real age of the user and only allow those who are of the right age.
Related Article : DrugGPT: Oxford's New AI Revolutionizing Medication Prescriptions