Brian McCullough (0:04)

Welcome to the Techmeme ride home for Friday, June 6th, 2025. I'm Brian McCullough. Today Circle had such a successful IPO. I'm wondering if IPOs might finally be back on the menu. Turns out Anthropic cut off Windsurf for the most obvious reason. Maybe Manus really is stoking a new gold rush, at least in China. And in the week on Long Read Suggestions, the most consequential weather forecast of all time. Here's what you missed today in the world of tech. No, I'm not going to do that. Big story from yesterday about Elon and Trump. Under my long standing thesis that it's politics, not tech, I will cover politics when it affects tech. So If Starlink or SpaceX contracts get canceled, fine, we'll talk about that. But until then, in my opinion, it's two dudes fighting over politics online. Anywho, Circle's stock soared 168% on Thursday, closing at $83.23 after the stablecoin company priced its New York Stock Exchange IPO at $31 and raised almost $1.1 billion in the offering. See this? This is what you want to see. The bankers underpricing IPOs because they misjudge the appetite of investors for certain types of companies. If there's an unknown groundswell of demand out there, then other companies will be incentivized to try their luck on public markets. And if they have success, then the dominoes start falling. I'd point to the recent stock explosion of coreweave after their lackluster IPO as a sign that there might be some heretofore unsuspected groundswell of investor hunger. Quoting CNBC Trading volume by the end of the session was about 46 million, far exceeding the number of freely floating shares available for trading. Circle joins Coinbase, Mara holdings and Riot platforms as one of the few pure play crypto companies to list in the US this marks the company's second attempt at going public. A prior merger with a special purpose acquisition company collapsed in late 2022amid regulatory challenges. To realize our vision, we needed to forge relationships with governments. We needed to work with policymakers. Because if you want this to work for mainstream, it's got to work in mainstream society and you need to have those rules of the road, CEO Jeremy Allaire told CNBC's Money Movers on Thursday. We've been one of the most licensed, regulated, compliant, transparent companies in the entire history of this industry and that's served us well. End quote. The crypto industry is enjoying newfound political favor under a friendly US Administration. The stablecoin sector specifically has been ramp up on the expectation that Congress will pass stablecoin legislation this summer. Wall street analysts say it could grow tenfold over the next five years, creating a trillion dollar market opportunity. Allaire co founded Circle in 2013. Based in Boston, the company initially focused on consumer facing payments and crypto wallets and exchange services and was the first to receive the famously difficult to obtain New York State bit license. In 2015, it moved to New York. Earlier this year, Circle founded the US Dollar pegged USDC stablecoin to establish a standard for F on the Internet, launching it in partnership with Coinbase in 2018 through a consortium called Center. In 2023, they dissolved center as a standalone entity, with Circle taking over the responsibilities of USDC and Coinbase taking a minority stake in the stablecoin company. The two companies also entered into an agreement to split the revenue of USDC's stablecoin and Coinbase. CEO Brian Armstrong said on the company's most recent earnings call that it has a stretch goal to make USDC the number one stablecoin. USDC is the second largest stablecoin on the market, behind Tether's USD End quote Following up on something here, Anthropic co founder Jared Kaplan says Anthropic cut off Windsurf's direct access to Claude's models, largely because of reports that OpenAI was acquiring Windsurf, quoting TechCrunch. We really are just trying to enable our customers who are going to sustainably be working with us in the future, said Kaplan during an onstage interview Thursday with TechCrunch at TC Sessions 2025. I think it would be odd for us to be selling Claude to OpenAI, Kaplan said. Windsurf declined to comment on Kaplan's remarks, and an OpenAI spokesperson did not immediately respond to TechCrunch's request. The companies have not confirmed the acquisition rumors. Part of the reason Anthropic cut Windsurf's access to Claude, according to Kaplan, is because the company is quite computing constrained today. Anthropic would like to reserve its computing for what Kaplan characterized as lasting partnerships. However, Kaplan said the company hopes to greatly increase the availability of models offer users and developers in the coming months. He added that Anthropic has just started to unlock capacity on a new computing cluster from its partner Amazon, which he says is really big and continues to scale. As Anthropic pulls away from Windsurf, Kaplan says he's collaborating with other customers building AI coding tools such as Cursor, a company Kaplan said Anthropic expects to work with for a long time. Kaplan rejected the idea that Anthropic was in competition with companies like Cursor, which is developing its own AI models. Meanwhile, Kaplan says Anthropic is increasingly focused on developing its own agentic coding products such as Claude code rather than AI chatbot experiences, While companies like OpenAI, Google and Meta are competing for the most popular AI chatbot platform. Kaplan said the chatbot paradigm was limiting due to its static nature and that AI agents would in the long run be much more helpful for users. End quote Speaking of Cursor, maker any Sphere raised $900 million, led by Thrive, at a $9.9 billion valuation, bringing its total funding to more than $1 billion, and says its annualized revenue has surpassed $500 million. Quoting Bloomberg. The round was led by Thrive Capital, with participation from other investors including Andreessen, Horowitz, Accel and DST Global. The new funding brings Anysphere's total capital raise to more than $1 billion and marks a steep increase from its valuation of $2.5 billion in Bloomberg previously reported details of the financing. Launched in 2023, Cursor includes an AI infused code editor that can analyze a programmer's actions and suggest additional lines of code, as well as a chatbot that can answer code related questions. Users pay for it via monthly subscription plans. AnySphere offers a $20 pro account for individuals and a $40 business account. The vast majority of the company's revenue comes from these subscriptions, but until recently, almost all of Any Sphere's customers were individual users. The startup began beefing up its enterprise sales efforts late last year by hiring its first salespeople, and businesses are now becoming a big chunk of its revenue, AnySphere Chief Executive Officer Michael Terrell said this week in an interview with Bloomberg. More than half of Fortune 500 companies are now using Cursor in some way, Terrell said. I think a lot of the excitement comes from the value that this tech is giving to developers, Terrell said. Cursor quickly became part of the daily routine for software programmers at a range of companies, including OpenAI, which is also an existing investor, as well as Spotify, Major League Baseball and Instacart. More than a million people use the service each day, the company said. Anysphere's rapid growth has made it one of the industry's most closely watched startups. Its software has helped lead to a new style of programming known as Vibe coding, where users simply accept the suggestions the AI assistant gives over and over. But the company also faces growing competition, including from OpenAI, which is in talks to buy any Sphere rival windsor for about $3 billion. Anysphere's revenue has surged in recent months. The company said it passed $100 million in annual recurring revenue in January, reaching the milestone after 14 months. AnySphere's investors have claimed it's the fastest growing software startup Time, a title previously held by Wiz, a cloud security company that reached $100 million in annual recurring revenue in 18 months. One of the things that keeps a lot of us from going to the gym is not knowing where to start or if our workouts will even, well, work out. Meet Tonal, the ultimate strength training system that tracks your progress to take the guesswork out of working out so you can have the confidence to keep pushing forward. Tonal provides the convenience of a full gym and the guidance of a personal trainer anytime at home. With one sleek system designed to reduce your mental load, Tonal is the ultimate strength training system, helping you focus less on workout planning and more on getting results. No more second guessing your form. Tonal gives you real time coaching cues to dial in your form and help you lift safely and effectively. After a quick assessment, Tonal sets the optimal weight for every move and adjusts in 1 pound increments as you get stronger so you're always challenged. Tonal lets you choose from a variety of expert LED workouts, from strength to arrow hit to yoga and mobility to keep you coming back for more. For a limited time, go to Tonal.com to get $500 off your Tonal purchase plus a free four year warranty. That's Tonal.com for $500 off plus a free four year warranty. Tonal.com AI agents are all the rage right now, but while single agents can handle specific tasks, the real power comes when specialized agents collaborate to solve complex problems. Exc there's still a fundamental gap. We have no standardized infrastructure for these agents to discover, communicate with and work alongside each other. That's where agency comes in. Agntcy the agency is an open source collective building the Internet of Agents, a global collaboration layer where AI agents can work together. It will connect systems across vendors and frameworks, solving the biggest problems of discovery, interoperability and scalability. For enterprises with contributors like Cisco, Crewai LangChain and MongoDB, agency is bright. Breaking down silos and building the future of interoperable AI. Agency is where folks are building the Internet of Agents to be a diverse collaborative space to innovate, develop and maintain software components and services for agentic workflows and multi agent software shape the future of enterprise innovation. Visit agency.org to explore use cases now that's a G N T C Y.org According to MIT Technology Review, Chinese startups and tech giants are racing to create AI agents for both local and global consumers following the popularity of Butterfly Effects Manus I feel like we keep not hearing about Manus, but then we sort of keep hearing about Manus behind the scenes. Quote There are now a host of Chinese startups building these general purpose digital tools which can answer emails, browse the Internet to plan vacations and even design an interactive website. Many of these have emerged in just the last two months, following in the footsteps of Manus, a general AI agent that sparked weeks of social media frenzy for invite codes after its limited release launch in early March. These emerging AI agents aren't large language models themselves. Instead, they're built on top of them, using a workflow based structure designed to get things done. A lot of these systems also introduce a different way of interacting with AI. Rather than just chatting back and forth with users, they are optimized for managing and executing multi step tasks, booking flights, managing schedules, conducting research by using external tools and remembering instructions. China could take the lead on building these kinds of agents. The country's tightly integrated app ecosystem, rapid product cycles and digitally fluent user base could provide a favorable environment for embedding AI into daily life. For now, its leading AI agent startups are focusing their attention on the global market because the best Western models don't operate inside China's firewalls. But that could change soon. Tech giants like ByteDance and Tencent are preparing their own AI agents that could bake automation directly into their native super apps, pull data from their vast ecosystem of programs that dominate many aspects of daily life in the country. It's been a whirlwind few months for Manus, which was developed by the Wuhan based startup Butterfly Effect. The company raised $75 million in a funding round led by the venture capital firm Benchmark, took the product on an ambitious global roadshow and hired dozens of new employees. Even before registration opened to the public in May, Manus had become a reference point for what a broad, consumer oriented AI agent should accomplish. Rather than handling narrow chores for businesses, this general agent is designed to be able to help with everyday tasks like trip planning, stock comparison or your kid's school project. Unlike previous AI agents, Manis uses a browser based sandbox that lets users supervise the agent like an intern watching in real time as it scrolls through web pages, reads articles or codes interactions. It also proactively asks clarifying questions, supporting long term memory that would serve as context for future tasks. Manus represents a promising product experience for AI agents, says Ang Lee, co founder and CEO of simulr, a startup based in Palo Alto, California that's building computer use agents, AI agents that control a virtual computer. I believe Chinese startups have a huge advantage when it comes to designing consumer products, thanks to cutthroat domestic competition that leads to fast execution and greater attention to product details. In the case of Manus, the competition is moving fast. Two of the most buzzy follow ups, genspark and Flow with, for example, are already boasting benchmar scores that match or edge pass Manuses. Genspark, led by former Baidu executives Eric Jing and K. Zhu, links many small super agents through what it calls multi component prompting. The agent can switch among several large language models, accepts both images and text, and carries out tasks from making slide decks to placing phone calls. Whereas Manus relies heavily on browser use, a popular open source product that lets agents operate a web browser in a virtual window like a human, Genspark directly integrates with a wide array of tools and APIs. Launched in April, the company says it Already has over 5 million users and over $36 million in yearly revenue flow. With the work of a young team that first grabbed public attention in April 2025 at a developer event hosted by the popular social media app Zhao, Hongshu takes a different tack. Marketed as an infinite agent, it opens on a blank canvas where each question becomes a node on a branching map. Users can backtrack, take new branches and store results in personal or shareable knowledge gardens, a design that feels more like project management software think notion than a typical CH chat interface. Every inquiry or task builds its own mind map like graph, encouraging a more nonlinear and creative interaction with AI. Flowist's core agent, Neo, runs in the cloud and can perform scheduled tasks like sending emails and compiling files. The founders want the app to be a knowledge market base and aim to tap into the social aspect of AI with the aspiration of becoming the only fans of AI knowledge creators. Their words End Quote Time for the week on long read suggestions and first up since we spoke about him yesterday, I have Nilay Patel's Decoder podcast interview with Runway CEO Chris Valenzuela. Here's some of the transcript quote I was just on a call with a studio right before this and we were going through a script that they wanted to test with Runway. I don't know if you guys have ever worked in film, but you developed the script, and the common thing to do next is a storyboard. So you basically take the storyboard and someone spends a week or two weeks just drawing. This is for a scene or a couple of scenes, not an entire film. It's really long, really expensive, and time consuming. So when they were reading me through the part of the script where they needed our help with Runway, I was generating the storyboards on the fly. By the time they finished, the storyboard was done. So I think the first thing was that they couldn't realize or fully understand what was going on because they had never worked at that velocity, that speed. For them, speed is also cost. If you have to compound the time it takes to make all of those storyboards by hand, and they have the screenwriters doing it in real time, then it shrinks the time and the whole project gets developed and worked on. So you have all these moments and gaps where AI can really just help you accelerate your own work. Specifically in creative industries where things are still very manually done. Nilay then asks, I actually want to ask you about that because I know you think a lot about the creative industries and the act of creativity. The counter argument to that is the gap between the screenwriter and the storyboard artist. And the time it takes to communicate and translate is where the magic happens. Having the AI collapse that in a mechanical process, as opposed to a creative process, actually reduces the quality of the creative. How do you feel about that? To which Chris responds, yeah, I don't think I fully agree with that. I think part of it is, I think that we sometimes obsess about the process of how we make things. The goal of the screenwriter is to get the ideas he wants in his mind or his world out there. The most obvious ways you work with the set of technologies and tools around you. If you're able to do it faster, I think that's great. You can iterate on concepts faster, you can understand your ideas faster. You can collaborate with more people, and you can make more. One of the bigger bottlenecks of media these days is that you have people working on one project for three or four years, and then you might actually work on it, and the studio might actually try to kill it for many different reasons. So if you think about it, you spend four years of your life working on a thing that never saw the light of day because it happened to be killed for whatever reason. I think the idea will be that you don't have to work on one project, you can work on many more. So that's also the quantity prosp of it. That becomes a component we should consider, because right now we're bound by the way we're working. It's very slow. It's very constrained by all these processes. If you can augment that, then people can start doing more and more and more. I think that's great. End quote. And then on about the anniversary of this happening, the anniversary of D Day, recalling one of the gutsiest weather predictions of all time, quoting the time if he had got the forecast wrong, Peter Stagg said from his home an hour from Bordeaux, I could have been sitting in German France, not France France. Mr. Stagg was speaking about the pivotal role his father, Group Captain James Stagg, played in liberating France from Nazi occupation. The elder Mr. Stagg was not a general or a foot soldier, but in the final hours before one of the most consequential moments of World War II, he was the man everyone was waiting on. On June 6, 1944, General Dwight D. Eisenhower ordered more than 150 Allied troops to storm the beaches of Normandy in one of the largest seaborne invasions in history. But hours before, Eisenhower's eyes were fixed not on the battlefield, but on the skies, more precisely on the weather report laid out before him. And the meteorologist who had created it, described by his son as a dour, irascible Scott, had to get it right. The weather forecast was a go or no go, said Dr. Catherine Ross, a library and archive manager at the Met Office, the weather service for the United Kingdom. Quote, everything else was ready. No weekend bonus episodes for you this weekend, one person overnight got in touch and said yesterday's file was weird. Things were repeated. There were weird noises. Usually when something goes bad, I get multiple reports of this. I don't know if other people noticed something. If so, please get in touch so I can investigate further. Further. Maybe I uploaded the wrong file or something I haven't had the time to check. Anyway, if I did screw up somehow, it would be useful to know how I did so I don't do it again. Talk to you on Monday. Oh, by the way, Monday, WWDC. The show will be late at least 3pm Eastern time. Probably later than that, depending on how long they go, maybe 4pm or further. So again, Monday, the show will be late. Wwdc. See you then.