Dell has confirmed to The Verge that its new Arm-powered XPS 13 features a tandem pOLED display. It’s the first to market with such a display in a laptop, after Apple introduced tandem OLED on its latest iPad Pro. Dell’s new XPS 13 is part of a collection of Copilot Plus PCs that are launching on June 18th with Qualcomm’s Snapdragon X Elite processors inside.
That’s a wrap for the first Build keynote. We’re off to listen in to sessions and learn more about Microsoft’s big AI plans for 2024. Stay tuned for a lot more on that, soon.
It’s clear OpenAI Sam Altman isn’t at Microsoft Build to announce a new model, but he’s happy to tease that the next big one is on the way. Microsoft built an even bigger supercomputer for this work, and now Altman hints that new modalities and overall intelligence will be key to OpenAI’s next model. “The most important thing and it sounds like the most boring thing I can say... the models are just going to get smarter, generally across the board,” says Altman.
OpenAI CEO Sam Altman has just stepped onstage at Microsoft Build. He’s having a conversation with Microsoft CTO Kevin Scott about what’s next for OpenAI and Microsoft’s big supercomputer plans.
Sal Khan, CEO of Khan Academy, is up onstage at Microsoft Build to talk about the nonprofit’s use of AI in education. Microsoft is partnering with the Khan Academy for AI-powered tutoring tools that will be free for all US educators as of today.
“It’s 12x cheaper to make a call to GPT-4o than the original model,” says Microsoft CTO and EVP of AI Kevin Scott. “It’s also 6x faster in time to open response.” These speed increases and cost decreases are super important for OpenAI’s latest model, but things aren’t going to slow down. Things will get cheaper and more robust over time, says Scott.
“We’re riding an extraordinary platform wave,” says Microsoft CTO and EVP of AI Kevin Scott onstage at Build. He likens it to the PC evolution and Moore’s Law or even broadband internet. Microsoft has been contributing to this with the company’s Copilot stack, which has helped the company build AI products quickly. “We are nowhere near... how powerful we can make AI models,” says Scott.
