Facebook owner Meta on Wednesday unveiled an artificial intelligence (AI) model that can select individual objects within an image, along with a dataset of image annotations it said was the largest ever of its kind. .
According to the company’s research division, Segment Anything Models (SAMs) can identify objects in images and videos, even if those items were not encountered during training.
SAM allows you to select an object by clicking on it or entering a text prompt. In one demonstration, I wrote the word “cat” and the tool now draws boxes around each of the multiple cats in the photo.
Meta hinted at several features that deploy the type of generative AI popularized by ChatGPT. ChatGPT creates entirely new content rather than simply identifying or classifying data like other AIs, but has yet to release a product.
Examples include tools that spin up surrealist videos from text prompts and tools that generate children’s book illustrations from prose.
Meta already uses SAM-like technology internally for activities such as tagging photos, moderating banned content, and deciding which posts to recommend to users on Facebook and Instagram.
Additionally, a New York startup called Runway AI generated a short video in less than two minutes of a quiet river in the woods after receiving the same short description.
Runway, which plans to roll out the service to a small group of testers this week, is developing an AI technique that can instantly generate a video by simply typing a few words into a box on a computer screen. is one of the company’s companies.
Google bets on speed
Google has released new details about the supercomputer it will use to train its AI models, saying the system is faster and more power efficient than comparable systems from Nvidia. Google designed its own custom chip called the Tensor Processing Unit (TPU). Google TPU is now in its fourth generation.
Google has published a paper detailing how it used a proprietary optical switch to string together more than 4,000 chips into a supercomputer.
Biden spots AI dangers, says tech companies need to make sure their products are safe
US President Joe Biden said he still doesn’t know if artificial intelligence (AI) is dangerous, but stressed that technology companies have a responsibility to ensure their products are safe before they go public.
Biden told science and technology advisers that AI could help tackle disease and climate change, but it would also be important to address potential risks to society, national security and the economy. “I think technology companies have a responsibility to make sure their products are safe before they go public,” he said at a meeting of the president’s Advisory Board on Science and Technology. The president said social media has already shown the harm powerful technology can do without proper safeguards.
He called for Congress to pass bipartisan privacy legislation to limit the personal data tech companies collect, ban advertising directed at children, and prioritize health and safety in product development. I repeated.
Australian mayor preparing world’s first defamation lawsuit over ChatGPT content
Australian regional mayor Brian Hood said he could sue OpenAI if false claims that ChatGPT has served time in prison for bribery are not corrected.– agency
