Before the public launch of GPT-4, we had law firms using our AI without knowing it was powered by GPT-4, and they were amazed by its capabilities, completing tasks that previously took a whole day in just minutes.

I saw the first version of GPT-4, and within 48 hours, we decided to shift the entire company to focus on building a new product called Co-Counsel based on that technology.

When we launched GPT-4.5, we were proud of the model, but users liked it much more than we anticipated, expressing that it was significantly better than GPT-4 in various ways.

After gaining early access to GPT-4, we developed a groundbreaking product that allowed us to add millions of dollars in revenue monthly, demonstrating the transformative power of AI in legal work.

In February, we launched the research preview of GPT-4.5, which is more intuitive and collaborative than previous models, making collaboration significantly easier.

We integrated some of the improvements from GPT-4.5 into our mainline GPT-4 model to enhance its collaborative capabilities.

We focused on alignment and safety for GPT-4, aiming to gather feedback from researchers to enhance its reliability and robustness.

The project for GPT-4.5 began about two years ago, anticipating a new cluster coming online, which led to extensive planning and risk management before the actual run.

The initial goal for GPT-4.5 was to create a model that is ten times smarter than GPT-4, which we ultimately achieved through a complex process.