We've created GPT-4, the latest milestone in OpenAI's effort in scaling up deep learning. GPT-4 is a large multimodal model (accepting image and text inputs, emitting text outputs) that, while less capable than humans in many real-world scenarios, exhibits human-level performance on various professional and academic benchmarks. For example, it passes a simulated bar exam with a score around the top 10% of test takers; in contrast, GPT-3.5's score was around the bottom 10%. We've spent 6 months iteratively aligning GPT-4 using lessons from our adversarial testing program as well as ChatGPT, resulting in our best-ever results (though far from perfect) on factuality, steerability, and refusing to go outside of guardrails.

Over the past two years, we rebuilt our entire deep learning stack and, together with Azure, co-designed a supercomputer from the ground up for our workload. A year ago, we trained GPT-3.5 as a first "test run" of the system. We found and fixed some bugs and improved our theoretical foundations. As a result, our GPT-4 training run was (for us at least!) unprecedentedly stable, becoming our first large model whose training performance we were able to accurately predict ahead of time. As we continue to focus on reliable scaling, we aim to hone our methodology to help us predict and prepare for future capabilities increasingly far in advance—something we view as critical for safety.

During a round of layoffs totaling some 10,000 employees Microsoft has kicked one of its AI ethics teams to the curb. This development comes off the heels of high profile launches for generative AI systems, extended partnerships, and further integrations of AI services into Edge and Bing. The team was not large, having been cut down in size back in October during a reorganization, but still made an impact within Microsoft as a mediator for AI policy and implementations. While the larger Office of Responsible AI that dictates the rules of play for the use of AI within the company remains intact, sources from within stated that the ethics and society team played a more pivotal role in ensuring the rules were applied during design and development of products.

These events ring familiar tones to those which occurred at Google in 2020, when ethical AI researcher Timnit Gebru was removed after publishing a critical overview of the large language models now popular in products such as OpenAI's ChatGPT. This cascaded through the company as several other leading members left amidst the turmoil. These events marred Google's reputation and raised concerns over the company's credibility regarding responsible use of AI, and as we're still mentioning it have had lasting impacts on their brand.

According to TrendForce's latest survey of the server market, many cloud service providers (CSPs) have begun large-scale investments in the kinds of equipment that support artificial intelligence (AI) technologies. This development is in response to the emergence of new applications such as self-driving cars, artificial intelligence of things (AIoT), and edge computing since 2018. TrendForce estimates that in 2022, AI servers that are equipped with general-purpose GPUs (GPGPUs) accounted for almost 1% of annual global server shipments. Moving into 2023, shipments of AI servers are projected to grow by 8% YoY thanks to ChatBot and similar applications generating demand across AI-related fields. Furthermore, shipments of AI servers are forecasted to increase at a CAGR of 10.8% from 2022 to 2026.

Yesterday, Microsoft held a conference showcasing its next-generation AI solutions in partnership with OpenAI. Among them, the company showed that it is making steady progress toward integrating Large Language Models (LLMs) in the Bing search engine, which will appear soon. What is interesting about the future release is the new functionality and capability of LLM, which Microsoft says is much more potent than an existing solution like ChatGPT. OpenAI's ChatGPT is not suited particularly well for search engines and its capabilities, which is why Microsoft's next-generation Bing will use a custom LLM called Prometheus Model, envisioned for advanced search capabilities.

As the company notes, Prometheus will show users relevant search results and allow up to 1000 character input in the chat. Even when the model doesn't know the answer, it will not hallucinate and produce false results. Instead, Prometheus will point the user to further reading, allowing for a safer and more accurate search experience. Currently, anyone can go to Bing.com and join the waitlist to try out the preview of the upcoming Bing Chat. Microsoft Consumer Chief Marketing Officer Yusuf Mehdi said, "We're going to scale the preview to millions in the coming weeks." A mobile-tailored preview is coming soon as well.

OpenAI's ChatGTP has reportedly reached an astonishing 100 million monthly active users in the heating wars of AI. This figure is achieved after a few months of availability, and big tech companies are pressured to respond. Today, we have information that Google will release a model similar to ChatGTP called Bard AI. Based on Language Model for Dialogue Applications (LaMDA) that Google introduced over two years ago, the Bard AI solution will integrate with Google search to access the latest information around the web. Currently in preview for private testers, Bard AI will roll out to the public in the coming weeks as the demand for Large Language Models that are in chat format soars.

Google CEO Sundar PichaiBard seeks to combine the breadth of the world's knowledge with the power, intelligence and creativity of our large language models. It draws on information from the web to provide fresh, high-quality responses. Bard can be an outlet for creativity, and a launchpad for curiosity, helping you to explain new discoveries from NASA's James Webb Space Telescope to a 9-year-old, or learn more about the best strikers in football right now, and then get drills to build your skills.

Today, we are announcing the third phase of our long-term partnership with OpenAI through a multiyear, multibillion dollar investment to accelerate AI breakthroughs to ensure these benefits are broadly shared with the world.

This agreement follows our previous investments in 2019 and 2021. It extends our ongoing collaboration across AI supercomputing and research and enables each of us to independently commercialize the resulting advanced AI technologies.