Tech Brew Ride Home - Wed. 04/03 – Quantum Computing Breakthrough?

Episode Date: April 3, 2024

Have we just had a major breakthrough in Quantum Computing? The earthquake in Taiwan might lead to some chip issues. We don’t quite know yet. Amazon is retooling it’s Just Walk Out technology. Venture capitalists are having a hard time raising money. And in the last segment of the show, I actually break some news about Coinbase. Links: Microsoft and Quantinuum say they’ve ushered in the next era of quantum computing (TechCrunch) AI-generated songs are getting longer, not necessarily better (The Verge) Top musicians among hundreds warning against replacing human artists with AI (Axios) TSMC Facilities to Resume Production Overnight After Quake (Bloomberg) Amazon’s Grocery Stores to Drop Just Walk Out Checkout Tech (The Information) Venture capital reckons with the end of ‘megafund’ era (Financial Times) YouTube Of Our Interview With David Marcus 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 Wednesday, April 3rd, 24. I'm Brian McCullough today. Have we just had a major breakthrough in quantum computing? The earthquake in Taiwan might lead to some chip issues. We don't quite know yet. Amazon is retooling. It's just walkout technology. Venture capitalists are having a hard time raising money. And in the last segment of the show, I actually break some news about Coinbase. Here's what you missed today in the world of tech. Potentially historical news that is not AI related, Microsoft and Quintum this morning detailed a breakthrough in quantum error correction, helping run more than 14,000 experiments with no errors, an issue that has long been a problem for quantum computing.

Quoting TechCrunch, using quantumum's ion trap hardware and Microsoft's new qubit virtualization system, the team was able to run more than 14,000 experiments without a single error. This new system also allowed the team to check the logical cubits and correct any errors it encountered without destroying the logical cubits. This, the two companies say, has now moved the state-of-the-art of quantum computing out of what has typically been dubbed the era of noisy intermediate-scale quantum NISQ computers. Noisy, because even the smallest changes in the environment can lead a quantum system to essentially become random or decoher, and intermediate scale, because the current generation of quantum computers is still limited to just over 1,000

cubits at best. A cubit is the fundamental unit of computing in quantum systems, analogous to a bit in a classic computer, but each cubit can be in multiple states at the same time and doesn't fall into a specific position until measured, which underlies the potential of quantum to deliver a huge leap in computing power. It doesn't matter how many cubits you have, though, if you barely have time to run a basic algorithm before the system becomes too noisy to get a useful result or any result at all. Combining several different techniques, the team was able to run thousands of experiments with virtually no errors. That involved quite a bit of preparation and pre-selecting systems that already looked to be in good shape for a successful run, but still, that's a massive improvement

from where the industry was just a short while ago. It's a step in the right direction for quantum computing. There are still plenty of problems to be solved, and these results need to be replicated too, of course, but theoretically, a computer with 100 of these logical cubits could already be useful for solving some problems. The team used quantumum's H-2-trapped ion processor and was able to combine 30 physical cubits into four highly reliable logical cubits. Encoding multiple physical cubits into a single logical cubit helps protect the system from errors. The physical cubits are entangled together so that it becomes possible to detect an error in a physical cubit and fix it. This could, Microsoft says, unlock commercial advantage. It's this error-corrected. It's this error-corrected.

that has long vexed the industry. The lower the noise and the higher the quality of the physical qubits, the better, of course, but without sophisticated error correction, there is no way out of the NISQ era because these systems will all decoher sooner than later, end quote. This is AI-related. Stability AI has released stable audio 2.0, letting users create three-minute sound clips from copyright-free audio samples available for free via their website, quoting the Verge. The first version of stable audio was released in September 2023 and only offered up to 90 seconds for some paying users, which meant they could only make short sound clips to experiment with. Stable Audio 2.0 offers a full three-minute sound clip,

the length of most radio-friendly songs. All uploaded audio must be copyright-free. Unlike Open AI's audio generation model, voice engine, which is only available to a select group of users, Stability AI made stable audio free and publicly available through its website and soon its API. One big difference between stable audio 2.0 and its earlier iteration is the ability to create songs that sound like songs complete with an intro, progression, and an outro, says Stability AI. Stability AI said in a press release that stable audio is trained on data from audio sparks, which has a library of more than 800,000 audio files. Stability AI maintains that artists under audio sparks were allowed to opt out

of their material being used to train the model. Training on copyrighted audio was one of the reason Stability AI's former vice president for audio Ed Newton Rex left the company shortly after the launch of stable audio. For this version, Stability AI says it partnered with audible magic to use its content recognition technology to track and block copyrighted material from entering the platform, end quote. Shot and then Chaser, Billy Eilish, Katie Perry, and more than 200 other artists have signed a letter to AI developers and tech firms decrying what they call predatory AI that steals voices replaces artists and more, quoting Axios. Unlike other advocacy efforts from creators around AI, this letter specifically

addresses tech firms about the concerns of musical artists such as replicating artist's voices, using their work to train AI models without compensation, and diluting royalty pools that are paid out to artists. Jen Jacobson, executive director at the Artist Rights Alliance, ARA, the trade group representing the artist signing the letter, told Axios, we're not thinking about legislation here. We're kind of calling on our technology and digital partners to work with us to make this a responsible marketplace and to keep the quality of the music sound and not to replace human artists, end quote. The letter penned by dozens of well-known musicians within ARA specifically calls on tech firms and AI developers to stop the, quote,

predatory use of AI to steal professional artists' voices and likenesses, violate creators' rights, and destroy the music ecosystem, end quote. Signatories include Elvis, Costello, Nora Jones, Nikki Minaj, Camilla Cabello, Casey Musgraves, John Batiste, Jha Rule, Jason Isbell, Pearl Jam, Sam Smith, and dozens more spanning every musical genre. We call on all AI developers, technology companies, platforms, and digital music services to pledge that they will not develop or deploy AI music generation technology, content, or tools that undermine or replace the human artistry of songwriters and artists or deny us fair compensation for our work. The letter reads, end quote. TSM has halted some chipmaking and evacuated some plants after that 7.4 magnitude earthquake in Taiwan,

the biggest in that country in 25 years. They also plan to assess the impact of the earthquake on their operations going forward. UMC made similar moves, quoting Bloomberg. Taiwanese firms from TSM to ASC technology make and assemble the vast majority of the semiconductors that go into devices from iPhones to cars, from factories vulnerable to even the slightest tremors. A single vibration can destroy entire batches of the precision-made semiconductors. The island's tech firms are still assessing impact from the earthquake, which leveled dozens of buildings on its eastern side and killed at least four people. On Wednesday, TSM said staff were beginning to return to evacuated sites, though it stressed it was examining impact.

Still, any halt in production threatens to upset a process that, especially for sophisticated semiconductors, can require uninterrupted seclusion in a vacuum for weeks on end, Barclay's analyst wrote. Some of the high-end chips need 24-7 seamless operations in a vacuum state for a few weeks. Analyst Bumkisson and Brian Tan said, Operation halts in Taiwan's northern industrial areas could mean some high-end chips in production may be spoiled, end quote. Taiwan is prone to quakes because it's near the convergence of two tectonic plates, yet it's also the source of an estimated 80 to 90 percent of the highest-end chips required

for advanced applications such as smartphones and AI. Industry executives and government officials have long called out the dangers of centering the world's advanced semiconductor production on an island that, apart from natural shocks, is considered a potential military flashpoint. That became particularly evident during the COVID era, which exacerbated a global shortage of the vital components, end quote. Yes, and I want to stress here, I'm not mitigating this situation by talking about this end of things. Any loss of life or property is tragic. I want to acknowledge that. But the reason, reasons. Everyone wants to diversify chip production globally are not just geopolitical, they're also just geo. Probably not a good idea to have 90% of the production of a key component of the

global supply chain in one place where a single hurricane or in this case, earthquake could take everything out. When I took the kids to the New York Auto Show this weekend, I got to use Amazon Just Walkout technology for the first time. Well, I guess I got that in under the wire, because Amazon apparently plans to remove Just Walkout from most of its grocery stores and add more grocery stores with dash carts, which instead let shoppers scan items while they shop, quoting the information. Amazon's pullback on deploying Just Walkout underscores how the company is still fine-tuning its broader grocery strategy. It also shows the limitations of using Just Walkout, which relies on cameras and sensors to track what people leave the store with in larger format stores.

Amazon had previously tried to pitch big box retailers like Woolworth's on the technology and even Wade opening its own big box stores equipped with Just Walkout the information has reported. Originally, Amazon envisioned using the Just Walkout system to provide a fresh twist on grocery shopping that could speed checkouts, and it currently has the technology installed in 47 of its 64 fresh stores across the U.S. and UK. But Tony Hoggett, Amazon's grocery chief and a three-decade veteran of British retail giant Tesco, who joined Amazon in 2022, said, Just Walkout makes sense in smaller shops where customers are purchasing a few items, but isn't ideal for larger grocery stores like Fresh. With Just Walkout Amazon emails receipts to customers after they leave the store, a process that can take minutes or hours, potentially making budget-conscious shoppers uneasy.

With dash carts, by contrast, customers scan items as they shop and see their spending added up on a screen in real time. That can appeal to grocery shoppers, Hoggett said, giving the example of a family that wants to make sure it's getting the most out of a $130 weekly grocery, budget. Another likely factor is that Just Walkout can be expensive to install and has relied on humans to double-check transactions, a process that is particularly complicated in larger stores and with loose items, certain kinds of produce. In 2022, Amazon moved the team behind Just Walk out from Huggett's organization to Amazon Web Services, which has gained traction with selling the

technology to smaller retailers like convenience stores in airports and concession stands in arenas, like the one I used at the New York Auto Show. Amazon has already overhauled a handful of its fresh stores in California and Illinois into version two stores, which also feature brighter colors and crispy cream donut and coffee stands, among other changes, Hoggett said. While some of the remodeled stores never had just walk out in the first place, Amazon has removed the system from at least one store in Pasadena, a spokesperson confirmed. The spokesperson added that Amazon plans to keep just walk out in smaller format fresh stores in the UK. The technology also still features in Amazon's 22 Go

convenience stores, though the company closed nine of those stores last year, end quote. Some VC industry news for you, according to Pitchbook and NVCA, U.S. Venture Capital Investments in Q1 of this year hit $36.6 billion, the lowest since 2017, that comes across $2,882 deals, and is down from $51.6 billion across $4,026 deals in Q1 of last year. So the investing recession continues in 2020. tech. It's not just that IPOs are scarce, it's that private investment money has not yet bounce back either. Could a reason for that be that VCs themselves are having a hard time raising money for themselves? Pitchbook also says that VCs raised $30.4 billion globally

in Q1 of 2024, a marked slowdown from 2023, which itself was the worst year since 2016, suggesting the era of so-called megafuns is over. As ever, I like to share stuff like this with you not as inside baseball, but so that you founders out there know to what degree it is or is not relatively easy to raise money for your startups. And if VCs themselves are having a hard time raising money, then the knock-on effects could be real. Quoting the Financial Times, investors and venture funds known as limited partners, have reigned in spending over the past two years, taking a more cautious approach as interest rates have risen. Startup exits, including public listings and sales have slowed and returns from venture fund managers have cratered.

Why has there been such a sustained slowdown? At the core of the issue is exits, said Heidi Gao, a venture capital analyst at Pitchbook, a resurgence in initial public offerings or sales would allow LPs to recoup their invested capital and recycle it. Unless we see meaningful improvements in the exit market, we're expecting fundraising difficulties to linger, and that will put downward pressure on dealmaking, Gao said. Fundraising activity has slowed sharply since 2021 when VCs hauled in $555 billion. Last year, they raised a third of that total, and activity has continued to slow, putting venture investors on course for their worst fundraising year since 2015. In the U.S., just $9.3 billion in capital was raised in the first three months of this year, roughly a tenth

of the total raised in 2023. We want to be there for our partners, but we don't want to put ourselves in a hole, said the chief investment officer at one big U.S. Foundation. He pointed out that he and other LPs had written ever larger checks as the market boomed but are yet to see a return because exit activity has stalled. It's tough maths for a lot of investors, end quote. The sluggish pace of fundraising is an ominous sign for startups who rely on VC investments to grow in their earliest stages. It continues a sharp reversal in VC firm's fortune since 2021 when they spent a record $747.5 billion, according to Pitchbook. The splurge was made possible in large part by the advent of mega funds, vehicles of $5 to $10 billion, that created a period of

unprecedented abundance for startups. That era is now over, according to Pitchbook. Fundraising data for the first quarter quote shows there may be no appetite for such vehicles in today's market, end quote. And finally today, as promised, I have some news for you, in the sense that I can actually break some news instead of reading you other people's reporting on breaking news. Light Spark is a crypto startup founded by David Marcus, who you might have. might remember as the guy who did that whole Libra Diem project at Meta when Meta briefly flirted with making their own currency. Well, Lightspark this morning announced that Coinbase selected them to bring the Bitcoin Lightning Network to the Bitcoin platform. What does that mean exactly?

Well, here is David Marcus himself to tell you. Thanks for having me. Our news today is that Coinbase is going to implement Lightning using light spark stack, and that will mean faster, cheaper transactions on the Bitcoin network for all Coinbase users very soon. So let's assume that half of the audience maybe isn't very deep into crypto. Can you explain what the Lightning network is for those that might be unfamiliar? Sure. So Lightning is actually a layer two payments protocol on top of Bitcoin that enables Bitcoin to move in real time at a fraction of the cost of layer one.

So as you probably know, layer one Bitcoin is actually pretty slow. It's very secure, but slow. One new block every 10 minutes, which means that your transaction will take about 10 minutes to clear on layer one. Lightning makes it instantaneous, real-time, low-cost. And so when you want to deposit or withdraw Bitcoin from Coinbase, now you'll be able to do it in real-time at a much lower cost from any other wallet or any other exchange or any other institution that supports Lightning, which is almost everyone now. Okay, so explain that in greater detail, and we can come back to Lightning and Light Spark

in the technology. If I'm a Coinbase user, what am I going to see? What is the immediate benefit of this to me? So if you're a Coinbase user and you want to move the Bitcoin that you just purchased on Coinbase to another wallet, we've all had this experience, or many of us have had this experience where you have to copy and paste an address, a layer one address, and send a transaction, and then for about 10 minutes, you wait patiently with a lot of angst for your Bitcoin to arrive to the other side, never really knowing whether it's actually going to happen or not, and despite the fact that you know it will likely happen, it's always a stress generating. And now you'll be able to do the same, but just with a lightning address.

And when you do that, you paste a lightning invoice, that's the term, into your interface, then your Bitcoin will appear on your other wallet or other exchange or other institution, basically less than a second after you click the send button. So obviously, that's just a clip from a bonus episode interview, Chris and I recorded with David. Link to the YouTube of that interview is the final link in today's show notes. And also, of course, it will be a bonus episode on the podcast feed this weekend. Hey, PR people out there. As you can hear, I've been leaning more into interviewing actual newsmakers this year,

and that includes breaking news. If you have something newsworthy to announce, sure, brief tech crunch, brief the verge, but also consider coming here to break the news in your own words. I can clip it and include it on a daily show. If it's newsworthy enough, just have your founder or CEO jump on a Zoom with me. even have to do a full bonus episode interview. Just pop on for three minutes, share the news, and I can include it as a segment on the Daily Show. Open invitation. Brian at Techmeme.com. If interested, talk to you tomorrow.