AI was the central focus of the opening keynote at CES on Monday as Dr. Lisa Su, Chair and CEO of Advanced Micro Devices (AMD), took the stage to unveil her company’s latest chips.
“You ain’t seen nothing yet,” said Su, during the opening keynote at the tech conference in Las Vegas. “We are just starting to realize the power of AI.”
Kicking off the event with a video that illustrates the possibilities of AI tech across lives from gaming to healthcare to connectivity to self-driving cars and even designing a renewable energy source as powerful as the sun itself” and “helping make travel time across the Atlantic, just another puddle jumper.”
A Nvidia competitor, AMD is on a mission “to create a world where the most advanced AI capabilities end up in the right hands: yours” with the tagline “together we advance AI.”
“AI is the most important technology of the last 50 years…and we are just scratching the surface,” said Su. “AI is for everyone, it makes us smarter. It makes us more capable.”
During the presentation, Su noted that AI has a billion active users, up from a million a couple of years ago, a growth in user base that even the Internet did not achieve. She predicts that AI usage will soon reach 5 billion active users and will be as indispensable to daily life, just as phones and the internet are today. To achieve this goal, we will need more computing power than ever before.
“We don’t have nearly enough compute for everything we can possibly do,” she said. “To enable AI everywhere we need to increase the world’s compute capacity another 100x over the next few years to more than 10 yottaflops over the next five years.” (A yottaflop =1 followed by 24 zeros)
“There has never been anything like this in the history of computing,” she said.
Su spent the next hour unveiling a number of the company’s AI chips including its advanced MI455 AI processors, which are used in data center server racks by customers including ChatGPT maker OpenAI. She also unveiled the MI440X, a version of the MI400 series chip designed for companies to use on-premise, designed to fit into infrastructure that is not specifically designed for AI.
OpenAI President Greg Brockman joined Su on stage and said chip advancements were critical to OpenAI’s vast computing needs. Brockman pushed the idea that ChatGPT is improving people’s lives, particularly when it comes to healthcare. He highlighted several examples of ChatGPT saving people’s lives. For instance, a colleague whose husband was sent home from the ER and told to wait out leg pain but returned based on the advice of the AI tool catching a potentially fatal blood clot.
To keep up with usage and to roll out new features for users, ChatGPT needs additional hardware and compute power.
“We are moving to a world where human intent becomes the most precious resource, there should be very low latency interaction anytime a human is involved but there should be an ocean of agentic compute that is constantly running,” said Brockman, noting that their partnership with AMD supports this growth, suggesting that the more compute power, the better the AI possibilities.
“The hardest problem for humanity will be how do we use the limited resources that we have to get the most benefit for everyone,” Brockman concluded.
Amit Jain, CEO of Luma AI, also joined Su on the stage where he discussed Ray3, an AI video tool that is designed to create studio level content in 4K, using reasoning that can think first in pixels. The company counts advertising, media and entertainment companies as clients. 2025 was a year of experimentation and now customers are asking for more control and precision.
“Control comes from intelligence, not just better prompts,” said Jain. To meet these needs, the company has created the latest version called Ray3 Modified that “allows you to edit the world,” ushering in what Jain described as “hybrid human AI productions.” The new AMD cards are supporting this growth.
Liquid AI co-founder Ramin Hasani also joined Su on stage to discuss how the two companies are partnering on LFM 3.0, designed natively multi-model to process text, visuals and audio as inputs and deliver text, visuals and audios as output in 10 different languages. This allows AI to be more proactive for PC users.
“Most AI assistants, co-pilots today are reactive agents,” said Hasani. “You open an app, you ask a question, it responds. But when the AI is running in the background on the device and is always on, it can be working on the tasks proactively for you. The tasks can be done in the background.”
Dr. Fei-Fei Li, the co-founder and CEO of World Labs, aka the godmother of AI also joined the stage discussing bringing AI models to spatial intelligence that connects perception with action.
“There is now a new wave of AI technology for both in-body AI and generative AI that we can finally give machines something closer to the human level of spatial intelligence,” said Li. “It is the ability to not only perceived but to create 3D and even 4D worlds, reason about objects and people and imagine entiley new environments that still obey the laws of physics and dynamics in worlds virtual or real.”
The World Labs tech, powered with AMD, can generate 3D navigable worlds, which are being applied in gaming, architecture and design, robotics and virtual production.
The keynote also hosted a panel on how AI tools (and AMD hardware) are involved in healthcare fro identifying issues in women health to male pattern baldness through a partnership with AstraZeneca and with an Illumina partnership, supporting tech that looks at the human genome and DNA sequencing to find disease (which uses more data in a day than is generated on YouTube).
The keynote also touched on physical AI, one of the “toughest challenges in technology,” which includes smart machines that take actions to address complex goals from factory robots to self-driving cars. AMD discussed partnerships with Generative Bionics and Blue Origin.
White House science advisor Michael Kratsios joined Su on stage to discuss how AMD was helping the U.S. “win” the “AI race” with the Genesis Mission, a public-private partnership that aims to use AI for scientific discoveries, which he compared in size to the Apollo Mission and the Manhattan project.
“This whole of government approach represents a historic mobilization of resources, tasking the department of energy to integrate it’s world class super computers and assets into a unified closed loop AI platform,” he said.
AMD is also investing $150 million in education programs that bring AI into classrooms and communities across the country.
“We are investing in the next generation of AI research and talent,” she said. The event closed with a look at this future talent, as teens from a recent AMD x Hack Club Hackathon took to the stage to show their work.
