Microsoft (MSFT) announced a slew of AI improvements at its Ignite conference on Tuesday. One message stood out: The tech giant is focused on making AI cheaper for businesses by offering specialized models for specific tasks. In some cases, the tasks that would've taken the company weeks were performed in minutes.
Microsoft AI CEO Mustafa Suleyman said he is hearing time and again from enterprises that "token spend is going through the roof."
"Our commitment is to give you the very best price performance out there, so that you can really scale your agents," Suleyman said. "And we also know that you want models that feel truly yours."
During the conference, executives highlighted Microsoft IQ, a new platform designed to help businesses organize and share employees' knowledge, expertise, and tools. The idea is to help AI systems learn how a company works and build on that knowledge over time.
CEO Satya Nadella highlighted the launch of new AI models for coding, cybersecurity, and decision making. The company's flagship coding model, MAI Code 2, is designed to handle complex, time-consuming programming tasks, work across multiple files, and understand entire software projects.
Nadella also underscored Copilot Fusion, which orchestrates "multiple models to deliver frontier-level results at significantly lower costs."
Microsoft highlighted a feature that selects the right model for the task, making it easier to maximize output quality while managing the cost of each request.
(MSFT)
Nadella kicked off the conference by acknowledging last week's outage, which spanned regions.
"We know the trust you place in our services and take that responsibility very seriously, and in that context, I know we have work to do," Nadella said.
"We are committed to ensuring that we learn and we continuously improve," he added.