GenerativeAI
AI in the World
The Model Makers
GenAI is premised on new kinds of models. Large Language Models (“LLMs”) have been the most prominent. But large multimodal models (“LMMs”) and diffusion models are also responsible for much of the Wow! emanating from recent advances in AI.
Early on, we penned articles like The Focus on ChatGPT Is Missing the Forest for the Tree and PSA: ChatGPT is the trailer not the movie because too many were treating ChatGPT as a stand-in for not only every model but every application of every model even though ChatGPT was only one application of one model. Thousands of models exist already. Many more are on the way. And there will be orders of magnitude more commercial and applications built on top of these models.
.jpeg)
Researchers have identified a vulnerability in mainstream GPUs, including those from Apple, AMD, and Qualcomm, that could allow attackers to steal data from a GPU's memory. The vulnerability, called LeftoverLocals, could expose sensitive information generated by large language models and other machine learning applications. While CPUs have been refined for security, GPUs have not prioritized data privacy to the same extent. The researchers emphasize the need for GPU security refinements and coordination among chipmakers, device manufacturers, and end users to address this vulnerability.
Researchers have identified a vulnerability in mainstream GPUs, including those from Apple, AMD, and Qualcomm, that could allow attackers to steal data from a GPU's memory. The vulnerability, called LeftoverLocals, could expose sensitive information generated by large language models and other machine learning applications. While CPUs have been refined for security, GPUs have not prioritized data privacy to the same extent. The researchers emphasize the need for GPU security refinements and coordination among chipmakers, device manufacturers, and end users to address this vulnerability.
Researchers have identified a vulnerability in mainstream GPUs, including those from Apple, AMD, and Qualcomm, that could allow attackers to steal data from a GPU's memory. The vulnerability, called LeftoverLocals, could expose sensitive information generated by large language models and other machine learning applications. While CPUs have been refined for security, GPUs have not prioritized data privacy to the same extent. The researchers emphasize the need for GPU security refinements and coordination among chipmakers, device manufacturers, and end users to address this vulnerability.
OpenAI's ChatGPT has reached 100 million weekly users and has over two million developers. The company also announced new features, including a platform for building custom versions of ChatGPT and GPT-4 Turbo, a model with knowledge of world events. The popularity of ChatGPT continues to grow rapidly.
OpenAI's ChatGPT, a chatbot powered by GPT-3.5, surprised the company with its viral success. Since its release, OpenAI has been updating and refining ChatGPT based on user feedback. The team is working on addressing issues such as bias and "jailbreaking" attempts. They are also exploring the challenges of deploying large language models in different contexts. Despite the progress, they acknowledge that there is still much work to be done and that language models are still in the early stages of development.
OpenAI has experienced a significant surge in revenue, reaching an annualized rate of $1.3 billion, marking a 30% increase since the summer. This financial growth highlights OpenAI's position as a frontrunner in the AI industry, especially when compared to its competitors. The broader AI landscape is also experiencing financial success, driven by the demand for AI solutions across various sectors. OpenAI's strategic collaborations and partnerships contribute to its dominant position, and there are speculations of a potential revaluation that could range between $80 to $90 billion.
Microsoft-backed OpenAI valued at over $80 billion in latest deal.
Google DeepMind has unveiled its new AI model called Gemini, which is positioned as a competitor to OpenAI's GPT-4. While Gemini is touted as a powerful and versatile model, some experts believe it may signal the peak of AI hype. The model combines text, images, and audio to answer a wide range of questions and perform tasks. However, it remains to be seen if Gemini represents a significant advancement over previous models. The full release of Gemini will be staggered over the coming months.
Apple Inc. is quietly working on artificial intelligence tools that could challenge those of OpenAI Inc., Alphabet Inc.’s Google and others, but the company has yet to devise a clear strategy for releasing the technology to consumers.
