Tag: Large Language Models

Cybersecurity experts have discovered that ChatGPT, a large language model, can evade endpoint detection and response systems by creating mutating code. Hackers exploit the model's ability to generate dynamic versions of malicious code and bypass content filters using prompt engineering. Researchers have demonstrated the use of prompt engineering to build a polymorphic keylogger payload known as BlackMamba. Learn more about this dangerous vulnerability.

Elliptic, the crypto risk management firm, has partnered with ChatGPT to enhance its risk detection and intelligence gathering operations. With ChatGPT, Elliptic aims to improve its threat detection and analysis capabilities, without compromising its high standards. While there are still limitations in using AI for complex queries, leading firms like Solana are also planning to integrate ChatGPT, indicating a world of potential enabled by large language models. Discover how this integration could advance risk management in the world of cryptocurrency.

Discover how Large Language Models (LLMs), like ChatGPT and BERT, are revolutionizing productivity. But, with concerns over online privacy infringement, it’s important to find solutions. ISBInsight explores the ethical dilemmas of LLM technology and provides the necessary remedies to mitigate them. Learn how LLMs are shaping the future of technological and ethical advancements.

Qualcomm's new processors will allow AI models like GPT-4 and DALL-E2 to run on smartphones, reducing costs and increasing accessibility for users.

Chinese start-up Mobvoi is capitalizing on the popularity of generative AI technology in China, with its Xuliehouzi model boasting tens of billions of parameters, competing with OpenAI's 175 billion parameter GPT-3. The company has applied its technology to various products, including smartwatches, wireless earphones, and an AI voice dub assistant. Mobvoi is submitting its IPO plan to the Hong Kong stock exchange, aiming to raise up to US$300 million.