Category: Shorts

OpenAI will let you erase your chat history in ChatGPT OpenAI has released a new feature that allows users to switch off their chat history when using its ChatGPT chatbot. This setting will not save conversations, nor will it be used to help train OpenAI’s AI models. OpenAI will still store new conversations for up…

Google’s search engine is feeling the heat Google’s search revenues increased slightly in the first quarter of 2023, which is much slower growth than the company saw in the previous two quarters. This is due to competition from Microsoft’s AI-powered Bing chatbot and OpenAI’s ChatGPT. Google is focusing on building out its AI capabilities and…

Chromebooks get new privacy features with camera and mic switch Google is introducing new system wide camera and microphone access toggles in ChromeOS settings that will block all apps and sites from using them. This is a software-based solution that replaces physical camera and mic kill switches on PC laptops. The new Privacy Controls panel…

Matter compatible WiZ smart lights WiZ, a smart lighting company, has updated all its products manufactured in 2021 or later to the Matter smart home standard. This allows users to add the products to any Matter-compatible platform, including Apple Home. WiZ products are also compatible with Amazon Alexa, Google Home, and Samsung SmartThings. WiZ products…

Microsoft Edge is sending the sites you visit to Bing Recent reports have revealed that Microsoft’s Edge browser is sending URLs visited by users to its Bing API website, and this feature is enabled by default. Reddit users were the first to spot this privacy issue, and Microsoft has since stated they are looking into…

“As an AI language model”: the phrase that shows how AI is AI language models have been predicted to cause a tidal wave of automated spam on the internet, but so far have not had a major impact. However, signs are emerging that AI is being used to power bots, generate fake reviews, and stuff…

A new vulnerability in the software used by speech-language pathologists (SLPs) A high-severity security vulnerability impacting Service Location Protocol (SLP) has been discovered, which could be weaponized to launch volumetric denial-of-service attacks against targets. This vulnerability is said to impact more than 2,000 global organizations and over 54,000 SLP instances that are accessible over the…

Iranian Hackers Use a PowerLess Backdoor to Attack Israel Check Point, a cybersecurity firm, has identified a new wave of phishing attacks targeting Israel that is linked to an Iranian nation-state threat actor. The attack chain begins with an ISO disk image file that drops a custom in-memory downloader which then launches a PowerLess backdoor.…

Attack ad created by AI responds to Biden reelection announcement The Republican National Committee (RNC) has released an attack ad featuring AI-generated images to oppose Joe Biden’s reelection campaign. The ad shows a series of stylistic images of potential crises if Biden is re-elected in 2024, such as explosions in Taiwan and military deployments on…

Text-to-Image Models with Zero Effort: Text2Video-Zero is Generative AI models have made it possible to obtain unique, photo-realistic images and videos by simply describing what we want to see. Diffusion models are a key contributor to this and they take a text prompt and generate an output that matches that description. However, constructing such datasets…