Tech Brew Ride Home - Thu. 02/20 – Microsoft’s Quantum Breakthrough

Episode Date: February 20, 2025

Microsoft is claiming a historical quantum breakthrough with its first quantum processor, the Majorana 1. Google is going all in on the Gemini brand. The world’s thinnest foldable phone. Rabbit releases a demo it probably should have led with. And James Bond now works for Amazon. Sponsors: Freshbooks.com/pricing-offer Take your personal data back with Incogni! Use code RIDEHOME at the link below and get 60% off an annual plan: Incogni.com/ridehome Links: Microsoft announces quantum computing breakthrough with new Majorana 1 chip (The Verge) Microsoft’s Xbox AI era starts with a model that can generate gameplay (The Verge) Google app on iOS removing Gemini as it pushes users to full app (9to5Google) The world’s thinnest foldable phone doesn’t come cheap (The Verge) Rabbit shows off the AI agent it should have launched with (The Verge) 'Bond' Producers Stepping Back From Iconic Franchise As Amazon MGM Studios Takes Creative Control (Deadline) Join us for the livestream tonight here: https://riverside.fm/studio/2000th-episode Learn more about your ad choices. Visit megaphone.fm/adchoices

On April 4th, 2023, around 2 in the morning, a man was found stabbed multiple times on a sidewalk in downtown San Francisco. Hey, who did this to you? What happened next turned the story into a political firestorm. Reports have identified the victim as Bob Lee, the founder of Cash App. From Bloomberg Podcasts, this is Foundering, the Killing of Bob Lee, beginning April 16. Welcome to the Tech meme right home for Thursday, February 20th, 2025. I'm Brian McCullough today. Microsoft is claiming a historical quantum breakthrough with its first quantum processor, the Majorana One. Google is going all in on the Gemini brand, the world's thinnest foldable phone, Rabbit releases a demo it probably should have led with, and James Bond now works for Amazon. Here's what you missed today in the world of tech. As we were talking about the new iPhone yesterday, Microsoft made an announcement that might be orders of magnitude more important. They unveiled the Majorana 1, their first quantum processor that uses majorana particles instead of electrons with the potential to fit in a million cubits. It's complicated, quoting the verge.

Microsoft believes it has made a key breakthrough in quantum computing, unlocking the potential for quantum computers to solve industrial-scale problems. The software giant has spent 17 years working on a research project to create a new material and architecture for quantum computing, and it's unveiling the Majorana 1 processor, Microsoft's first quantum processor based on this new architecture. At the core of a quantum computer are qubits, a unit of information in quantum computing much like the binary bits that computers use today. Companies like IBM, Microsoft, and Google have all been trying to make qubits as reliable as binary bits for years now because they're a lot more delicate and sensitive to noise that can create errors or lead to loss of data. Majorana 1 can potentially fit a million cubits onto a single chip.

that's not much bigger than the CPUs inside desktop PCs and servers. Microsoft isn't using electrons for the compute in this new chip. It's using the Majorana particle that theoretical physicist, that Tori Majorana described in 1937. Microsoft has reached this milestone by creating what it calls the world's first topoconductor, a new type of material that cannot only observe but also control Majorana particles to create more reliable cubits.

Microsoft has outlined its research in a peer-reviewed paper, published today in nature explaining how its researchers were able to create the topological cubit. Microsoft has helped create a new material made from indium, arsenide, and aluminum, and it has placed eight topological qubits on a chip that it hopes can eventually scale to one million. A single chip with a million cubits could perform simulations that are a lot more accurate and help improve the understanding of the natural world and unlock breakthroughs in medicine and material science. That's been the promise of quantum computing for years now, and Microsoft believes it's topoconductor or topological superconductor is the next big breakthrough, end quote.

But let me come back to something. Microsoft says it has harnessed a completely new state of matter. Again, a new state of matter, not a solid, not a liquid, not gas, something else, using particles called Majorana Fermions, a significant breakthrough to create workable topological cubits. Again, I feel like this is a bit beyond my ken, but quoting the Financial Times, Microsoft's claim breakthrough follows years of research into a type of particle that comprises a fourth state of matter. Distinct from solids, liquids, and gases, the existence of the particles known as Majorana Fermions was first theorized in 1937, although scientists have struggled to demonstrate they actually exist. Microsoft Chief Sachin Adela said eight years ago that its work on the technology had put it on the cusp of a quantum revolution, but it was not until 2022 that the company's scientists were even able to record effects that they believed were caused by the particles.

The U.S. Tech Group placed its bet on the theoretical particles after deciding that they offered the best route to overcoming the biggest obstacles to building a practical quantum machine. While the bits in a traditional computer represent ones or zeros, the qubits used in quantum computers are able to represent both at once or any state in between. However, most types of cubit only hold their quantum states for tiny fractions of a second, meaning that any information they hold is quickly lost. To make up for this, fully functioning quantum computers will require many extra cubits to run the software needed. to correct for errors. By comparison, the so-called topological cubits Microsoft has been trying to make using major rana particles are more resistant to error. Information is held across the whole qubit, meaning that even when parts fail, a topological cubit as a whole should retain enough information to make it useful, said Sankar Das Sarma, a physics professor at the University of

Maryland. This greater stability means Microsoft is likely to need only about 100 extra cubits to correct the errors for every one, fully operational cubit, said Jason Zander, Executive Vice president of strategic missions and technologies at Microsoft. That is roughly a tenth as many as are expected to be needed in machines whose cubits are based on other materials. The latest data released by Microsoft, including in a paper published in Nature on Wednesday, represented a significant breakthrough towards creating workable topological cubits, said Das Sarma. However, he said there was still a small chance that the company's findings could be explained away by something other than the successful harnessing of the elusive particles. Speaking before the latest announcement,

one long-time investor in quantum computing compared its ambitious research to the pursuit of cold fusion, the effort to produce a nuclear reaction at room temperature that has become a byword for scientific over-promising. It also comes just weeks after Nvidia Chief Executive Jensen Huang, forecasts that useful quantum computers were still 20 years away, much longer than companies working in the field claim, end quote. So there you go, as I said. On the long arc of history may be a much bigger story than a new low-end iPhone, as Morning Brew tweeted, quote, Apple, here's another iPhone. Meta, we're doing more layoffs. Google, it's the Gulf of America now. Amazon, please come back to the office. Microsoft, we became God and created a new state of matter,

end quote. That wasn't all, actually, because Microsoft also announced Muse AI, a first of its kind generative AI model that can generate a game environment based on visuals, players' controller actions, or both, quoting the verge. Microsoft's new Muse AI model could help. help Xbox developers create parts of games in the future, and the company says it's even exploring the potential of using it to preserve classic games and optimize them for modern hardware. The Muse model was trained on a large amount of human gameplay data from the Xbox game Bleeding Edge, thanks to a collaboration between Microsoft Research and Xbox Studio Ninja Theory, both of which are based in Cambridge in the UK. This partnership allowed us to closely

collaborate with the game studio to understand what needs to be in place and how we can responsibly unlock access to a large amount of gameplay data, says Kacha Hoffman, head of Microsoft Research's game intelligence team. This allowed Mews to access the equivalent of seven years of human gameplay resulting in the model being trained on a billion image action pairs in total. It's a milestone research project that's being published in nature today, but as its early work, the model is limited to generating gameplay visuals at a resolution of just 300 by 180 pixels. That's up from the 128 by 128 resolution of Microsoft's earlier work on generative AI gaming models, but it's still far behind the 1080P resolution that is common among PC gamers.

Microsoft has released a number of examples of the Muse model in action, generating gameplay, and even allowing players to load visual elements into the game to prompt the model. This could be used for the early iteration stages of a game, but Microsoft is stressing this isn't designed to generate an entire game and replace creators. During a press briefing, Hoffman also demonstrated a real-time version of Muse that Microsoft is currently experimenting with to enable interactive AI-powered games. The demo generated game visuals on the fly in real time and even reacted to objects being dropped into the game to change the environment. While the real-time gameplay only ran at 10 frames per second and a 300 by 180 resolution, it was an early demonstration of what might be possible in the future.

Microsoft is now exploring how Muse could help improve classic games and bring them to modern hardware. You can imagine a world where from gameplay data and video that a model could learn old games and really make them portable to any platform. where these models could run, says Microsoft Gaming CEO Phil Spencer. We've talked about game preservation as an activity for us, and these models and their ability to learn completely how a game plays without the necessity of the original engine running on the original hardware, opens up a ton of opportunity, end quote. Google is removing Gemini support from the main Google app on iOS, pushing users to download the standalone Gemini app that launched in November 2024 instead. Quoting 9 to 5, Google. One of the first ways to assess.

Gemini on iOS was through the main Google app, which offered a switcher between Google search and Gemini for several months. However, as Google has built out the Gemini experience, new features have been missing from the Google app, a key example being Gemini Live. That's because Google has been focused on the full Gemini app for iOS, which launched in November. Now Google is letting iPhone and iPad users know that Gemini is no longer available in the Google app. In an email sent to users, Google says that Gemini is no longer available in the Google app, emphasizing that Gemini is now available in its own app, and that's now the best place to use Gemini. The email reads in part we're making some changes to create an even better Gemini experience on iOS.

Gemini is now available as its own app, and that's now the best place to use Gemini to continue using Gemini, download the new Gemini app from the App Store. With the Gemini app, you'll have access to all of the same features and more. Google removed the switcher interface earlier this month, as we previously reported, but the experience is currently still live in the app via the More tab with Google App version 356 on iOS. However, we previously noticed that a pop-up that directs users to instead download the full Gemini app for iOS via the App Store, end quote. Apo has launched the FindN5, the world's thinnest foldable phone at 8.93 millimeters closed,

and 4.21 millimeters went open with IPX6, X8, and X9 ratings, all for about 1,800 quoting the verge. That's more expensive than either the Samsung Galaxy Z Fold 6 or Google Pixel 9 Pro Fold in Singapore, though comparable once you adjust for storage, there won't be any U.S. release at all, as Sister Brand 1 Plus has already confirmed that it has no plans to launch a foldable phone this year. The Find N5 is slimmer than any rival foldable so long as you measure it when closed. At 8.93 millimeters, it's thinner than the previous record holder last year's Honor Magic V3 and less than a millimeter thicker than an iPhone 16 Pro. Open the phone up, and it's just 4.21 millimeters at its thinnest point, so thin that Oppo told me it had to customize the USBC port to

even fit, though measured this way, it is fractionally thicker than Huawei's 3.6mm Trifold Mate XT, which also got a global launch this week. The phone has IPX6X8 and X9 ratings, meaning it's protected against both submersion and spraying water, but not dust or dirt, arguably the bigger threats to folding phones. Opos has also made the display crease both narrower and shallower, making it harder to spot or feel when swiping. A 5,600-m-a-hour battery is surprisingly spacious for the phone's size and was good enough for two days of use at a time when I was testing the phone for my review. Appo has also managed to fit in wireless charging, one of the big emissions from the previous Find N3. There was no N-4 since it's considered an unlucky number

in China. The phone uses a seven-core version of the Qualcomm Snapdragon 8 Elite chip set and comes in just one configuration with 16 gigabytes of RAM and 512 gigabytes of storage. The Hasselblad branded cameras are impressive, too, with 50 megapixel main and telephoto lenses joined by a 12 megapixel ultra-wide, end quote. And remember Rabbit, makers of that Rabbit R1 AI gadget that made waves at CES over a year ago, they're demoing a generalist Android agent showing it controlling apps on a tablet, quoting The Verge. The company published a blog post and video today showing off a generalist Android agent,

slowly controlling apps on a tablet in much the same way that Rabbit claimed its R1 device would over a year ago. It couldn't and can't. The work builds on Lamb Playground, a generalist web agent Rabbit launched last year. The engineers don't use the Rabbit R1 at all for the demonstration. Instead, they type their request into a prompt box on a laptop, which translates them to actions on an Android tablet. They task it with things like finding a YouTube video or locating a whiskey cocktail recipe in a cocktail app, gathering the ingredients, and then adding them to a Google Keep grocery list. At one point, they asked it to download the puzzle game 2048 and figure out how to play it, which it does, albeit slowly. The model generally does the things they ask,

sometimes well and sometimes with quirks, like sending a poem over WhatsApp one message at a time instead of in a single block. One of the engineers wonders if they should have asked it to use line breaks in their prompt, but they don't go back and try again. Rabbit's AI agent is clearly still a work in progress, as it has been since R1 launched with almost none of the capabilities that founder of CEO Jesse Liu presented in January 2024. Rabbit has steadily rolled out updates, like the ability to train its AI agents to complete specific tasks or prompt it to remake its own interface.

The examples it presented today are only the core action loop and Android agent completes, according to Rabbit's blog post. The company promises to share more about its upcoming cross-platform multi-agent system in coming weeks, end quote. And finally today, Amazon is now in charge of James Bond. I'm being serious. Amazon MGM studios, Michael G. Wilson, and Barbara Broccoli have formed a joint venture to house James Bond intellectual property rights under which Amazon gains creative control of the franchise. Quoting deadline, financial details were not disclosed, but there is speculation that Broccoli and Wilson will have received a healthy payday from Amazon.

Since the most recent series starring Daniel Craig came to a close with the 2021 pick No Time to Die, 007 has appeared to be in no man's land. Despite wild speculation about who will replace Craig, there have been no substance. updates about the future of the franchise. In this information vacuum, there have been multiple deeply reported press stories about tensions between Amazon and the bond producers following the streamers $8.5 billion takeover of MGM in 2022. The Wall Street Journal reported in December that the relationship had all but collapsed, and Broccoli had characterized Amazon to friends as effing idiots. The tone in today's press release was not so caustic with Wilson describing Amazon as a trusted partner. Following the deal, Insiders said no decisions have been made on future bond

films, including who will star, direct, and write the next franchise. It means another installment is likely years away. This news is significant in the history of the iconic MI6 agent. As for decades, producers have gained significant leverage when it came to decisions regarding the franchise. Broccoli said, My Life has been dedicated to maintaining and building upon the extraordinary legacy that was handed to Michael and me by our father, producer Cubby Broccoli. I have had the honor of working closely with four of the tremendously talented actors who have played 007 and thousands of wonderful artists within the industry. With the conclusion of no time to die, and Michael retiring from the films, I feel it is time to focus on my other projects.

In 2022, Amazon acquired MGM, including a vast catalog with more than 4,000 films and 17,000 TV shows, since the MGM acquisition Amazon has held the rights to distribute all of the James Bond films, end quote. Okay, once again, the 2000th episode Spectacular is happening tonight, and I want you to be a part of it. As I say, we're using a new system that allows people to sit in the audience and be brought on stage to ask questions. Chris Messina will be there, so come say hi to him and bring us your questions about AI, about Sam Altman, about Intel, about anything in tech you want to talk about. You don't have to ask us questions for us to answer. You can bring up any topic and commiserate.

We'll talk about it. And if you just want to talk about the show in general, since it's the 2000th episode, please do. Just come and join us, 8 p.m. Eastern, 5 p.m. Pacific. Link is in the bottom of the show notes today. But remember, you have to be on a laptop or desktop to be able to. actually come on stage and go on camera. If all goes well, I will be posting this recording as tomorrow's show, so I can participate in the AI Engineer Summit tomorrow. So let's pray the recording goes well. Look for that tomorrow, and I'll talk to you on Monday.