Tag: technews

  • Researchers uncover new flaws in popular internet routing protocol software

    by

    in

    Researchers uncover new flaws in popular internet routing protocol software CVE-2022-43681 (CVSS score: 6.5) – Out-of-bounds read when processing a malformed BGP OPEN message with an Extended Optional Parameters Length option. This warning applies only to messages that have an extended optional parameter length and do not terminate at the option length octet. A threat…

  • Hackers Exploiting 5-year-old Unpatched Vulnerability in TB

    by

    in

    Hackers Exploiting 5-year-old Unpatched Vulnerability in TB Fortinet’s seen more than 50,000 attempts to exploit an unpatched flaw in its TBK digital video recording (DVR) devices in the month of April 2023. The flaw, CVE-2018-9995, was caused by an error when handling a maliciously crafted HTTP cookie. Threat actors are actively exploiting an unprotected five-year-old…

  • Researchers uncover new way to exploit PaperCut vulnerability

    by

    in

    Researchers uncover new way to exploit PaperCut vulnerability A group of security researchers has found a way to exploit a critical flaw in PaperCut servers. Previously, the flaw was disclosed by vulnCheck. Now, VulnCheck has published a proof-of-concept exploit that sidesteps existing detection signatures by leveraging the fact that PaperCut NG and MF offer multiple…

  • Microsoft Azure API Management Service found to have 3 vulnerabilities

    by

    in

    Microsoft Azure API Management Service found to have 3 vulnerabilities Three new security flaws have been disclosed in the Microsoft Azure API Management service. The most important of these is a flaw in the proxy function which allows malicious actors to gain access to sensitive information or backend services. It also follows the discovery of…

  • A New AI Research Paper From Stanford Presents an Alternative Explanation for Seemingly Sharp

    by

    in

    A New AI Research Paper From Stanford Presents an Alternative Explanation for Seemingly Sharp This paper discusses emergent abilities, which are characteristics of large language models (LLM) such as GPT, PaLM, and LaMDA. These LLM’s have what are known as ergent abilities, and they can be useful for machine learning. In this paper, the authors…

  • ChatGPT Rivals: Designed to Play With Your Emotions

    by

    in

    ChatGPT Rivals: Designed to Play With Your Emotions The chatbot company Inflection designed Pi to address some of the problems of today’s chatbots. Its goal is to be more engaging, if less useful, than programs like ChatGPT and Bard. According to Dr. Suleyman, the company has carefully curated Pi’s training data to reduce the chance…

  • The Boring Company’s Vegas Loop plan is anything but nice

    by

    in

    The Boring Company’s Vegas Loop plan is anything but nice The Boring Company has gotten the go-ahead to build more tunnels and stations for its underground transportation system in Clark County. Its goal is to alleviate road traffic by moving some of the rush-hour traffic underground. So far, though, it’s been rather underwhelming. The company…

  • Amazon plans to update Alexa to work better with ChatGPT

    by

    in

    Amazon plans to update Alexa to work better with ChatGPT Leaks report that Amazon is planning to add new AI chatbot technology to Alexa, reviving the voice assistant in an age of generative artificial intelligence. According to leaked documents, the company is relying on its own internally developed large language model (LLM) to power a…

  • New tools for making games offered by AI, but developers worry about their jobs

    by

    in

    New tools for making games offered by AI, but developers worry about their jobs The narrator explains how companies like Eleventh Hour Games, a company that uses artificial intelligence to write dialogue in video games, are using the technology to help them make games faster. They claim that it could help solve the problem of…

  • Meta Discovers Social Media Cyber Espionage Operations

    by

    in

    Meta Discovers Social Media Cyber Espionage Operations Three different threat actors leveraged hundreds of elaborate fictitious personas on Facebook and Instagram to target individuals located in South Asia as part of disparate attacks. The fake accounts used traditional lures like women looking for a romantic connection and masqueraded as recruiters, journalists, or military personnel. Two…