Contents
Venture Capital
IPO
Apple
AI
Government Overreach
Politics & Technology
Interview of the Week
Startup of the Week
Editorial: 20 Years of TechCrunch
Josh Kopelman is the founder of First Round Capital. Way back in June 2005 he received an email from TechCrunch founder Mike Arrington. It reminded me not to let this day pass without noting TechCrunch. Here is Mike’s email to Josh:
---- Forwarded message ---
From: Michael Arrington <editor@techcrunch.com>
Date: Sat, Jun 11, 2005 at 10:56 PM
Subject: I need your help
To: <editor@techcrunch.com>
Hello, and please forgive this mass-email to my friends.
Keith Teare and I have been working on a new venture together. It's called Archimedes Ventures and it's focus is on web 2.0 - the two way or read-write web. We talk a bit about this on our website at www.archimedesventures.com.
We are investing in web 2.0 companies, consulting with others, and starting a couple of companies on our own. We have four employees now in addition to the two of us. Our first company will launch in the late summer/early fall and I'm pretty excited about it.
We also launched a new business weblog yesterday called TechCrunch, at www.techcrunch.com. We are hearing about new companies every day, and I realized that there is no one place to go to see profiles of all of these exciting new services. I just randomly come across them in weblogs, news sites or from tips from friends. So I am going to profile every new "web 2.0" company that I come across on this new site. It's sort of a cross between venturewire (which is for newly funded companies), and gizmodo, which is a blog that reviews new electronic gadgets.
TechCrunch is less about making money and more about contributing back to the Internet I also realized that since I am basically joining/checking out every new web 2.0 company anyway, I should go ahead and share my thoughts with others.
This was an incredibly open and honest email. Mike was on a mission to understand Web 2.0 and wanted to soak up every possible experience he could to better grasp it.
Josh points out that “…ever since TechCrunch has been an open tab in my browser. It is hard to underestimate the impact it has had on the ecosystem.”
Others in the X thread concur and others ask questions, like this from Jay Wong:
Archimedes Labs was our co-founded entity and TechCrunch was Mike’s way of documenting the profound changes going on in the Internet. It still exists and has gone through three major changes - The TechCrunch era; The post TechCrunch 2010-2016 era and the current era.
Today 2005 Archimedes would be called a Venture Studio. We were intent on creating new Web 2 companies that we would be owners in. Our founding document was also the basis for TechCrunch and summed up our approach to Web 2.
The full document is here:
The first writings from June 2005 are all stored in the WayBack Machine, which stores everything from that date forward.
The About TechCrunch section takes a lot from the Archimedes Newsletter. It reads:
Posted in TechCrunch,Web 2.0 on June 11th, 2005
TechCrunch is a weblog dedicated to obsessively profiling and reviewing every newly launched web 2.0 business and service. In addition to new companies, we will profile existing companies that are making an impact (commercial and/or cultural) on the web 2.0 space.
TechCrunch is edited by Michael Arrington and Keith Teare, with frequent input from guest editors. It is part of the Archimedes Ventures network of companies.
What is “web 2.0″? There are entire websites dedicated to trying to define it in a succinctly. For instance, here is Wikipedia’s entry on web 2.0. At Archimedes, we think of web 2.0 as the inevitable evolution of the web from a read-mostly medium to a read-write, or two-way medium (think geocities v. weblogs). Web 1.0 was static html pages. Web 2.0 is dynamic and interactive, and more fully exploits network effects. Web2.0 applications leverage key new web application frameworks like Ruby on Rails and AJAX.
We are seeing the separation of content from its old forms. Text is no longer necessarily embedded in a web page, it can be syndicated through RSS or ATOM. Audio is no longer tied to the Radio network. It can be Podcast or streamed or downloaded. TV shows are no longer necessarily tied to TV Networks. They can be delivered on demand across IP networks. And so on.
These trends throw many business models into question. New companies are being created to leverage these trends. We will profile them here.
Like all good web 2.0 services, this site will be a two-way communications medium. Comments, trackbacks and other feedback will be welcome. We will distribute this content in every way our readers want it – our website, RSS, Atom, email, and other web 2.0 distribution mechanisms that companies that we profile will think up!
If you’d like to contact TechCrunch with suggestions, comments, corrections, errors or new company announcements, please email us at editor@techcrunch.com.
As Mike said, TechCrunch was initially a utility project. There was no intention of it becoming an asset. But that changed quickly mainly due to Mike’s focus and dedication. I can’t do justice here to his work, but in short, he moved to the Valley from LA. Rented a home in Atherton that became the “TechCrunch House”. He leveraged a new technology called Wiki’s to launch events at the house. Chad Hurley pitched YouTube at the first. Published regular reviews of new Web 2 companies.
By the end of 2005 TechCrunch had become the center of Web 2 in the Valley. Mike was working crazy hours, mostly writing himself. I was running our other company edgeio. TechCrunch retained and grew its authority over the years, especially after Heather Harde took the lead operational role. I might credit my wife Gené McPherson too. She was a powerhouse behind the scenes at the early TechCrunch events leading up to Disrupt. She also became a founding team member at CrunchBase when it became independent, where she still works.
On a personal note, people often credit me for TechCrunch because I was there and got many mentions. On the 20th anniversary I have to repeat an often said point. The success of TechCrunch is 100% Mikes. I was the author of the Archimedes Newsletter, a co-owner, partner and co-thinker but the idea of TechCrunch, the inspiration, passion, work, writing, networking, product and events management was all his.
Thanks to Josh Kopelman for his X post. I might have passed over this anniversary were it not for that. TechCrunch’s role in the HBO series Silicon Valley is probably a testament to what it became and remains. I still read it daily.
Essays
Apple Researchers Publish Paper on the Limits of Reasoning Models (Showing That They’re Not Really ‘Reasoning’ at All)
Apple • John Gruber • June 8, 2025
Technology • AI • Machine Learning • Language Models • Reasoning Capabilities • Essays
Parshin Shojaee, Iman Mirzadeh, Keivan Alizadeh, Maxwell Horton, Samy Bengio, and Mehrdad Farajtabar, from Apple’s Machine Learning Research team:
Recent generations of frontier language models have introduced Large Reasoning Models (LRMs) that generate detailed thinking processes before providing answers. While these models demonstrate improved performance on reasoning benchmarks, their fundamental capabilities, scaling properties, and limitations remain insufficiently understood. [...] Through extensive experimentation across diverse puzzles, we show that frontier LRMs face a complete accuracy collapse beyond certain complexities. Moreover, they exhibit a counterintuitive scaling limit: their reasoning effort increases with problem complexity up to a point, then declines despite having an adequate token budget. By comparing LRMs with their standard LLM counterparts under equivalent inference compute, we identify three performance regimes: (1) low-complexity tasks where standard models surprisingly outperform LRMs, (2) medium-complexity tasks where additional thinking in LRMs demonstrates advantage, and (3) high-complexity tasks where both models experience complete collapse. We found that LRMs have limitations in exact computation: they fail to use explicit algorithms and reason inconsistently across puzzles. We also investigate the reasoning traces in more depth, studying the patterns of explored solutions and analyzing the models’ computational behavior, shedding light on their strengths, limitations, and ultimately raising crucial questions about their true reasoning capabilities.
The full paper is quite readable, but today was my travel day and I haven’t had time to dig in. And it’s a PDF so I couldn’t read it on my phone. (Coincidence or not that this dropped on the eve of WWDC?)
My basic understanding after a skim is that the paper shows, or at least strongly suggests, that LRMs don’t “reason” at all. They just use vastly more complex pattern-matching than LLMs. The result is that LRMs effectively overthink on simple problems, outperform LLMs on mid-complexity puzzles, and fail in the same exact way LLMs do on high-complexity tasks and puzzles.
Apple researchers' AI red flag
Youtube • CNBC Television • June 9, 2025
Technology•Ethics•Innovation•Privacy•Essays
Apple researchers are raising concerns about the rapid development and deployment of artificial intelligence (AI) technologies within the company. These internal voices highlight the potential risks associated with AI systems that evolve without sufficient oversight or understanding of their full implications.
According to insiders, some Apple AI specialists believe that the push for faster innovation may overlook critical safety measures and ethical considerations. They warn that hastily integrated AI could lead to unintended consequences, including privacy breaches, bias in AI decision-making, and vulnerabilities to exploitation by malicious actors.
The debate is reportedly ongoing inside Apple, with some researchers advocating for more cautious and methodical progress in AI development to ensure transparency and accountability. They stress the importance of thorough testing and the establishment of clear frameworks for responsible AI use.
This internal discourse at Apple reflects a broader trend in the tech industry, where companies grapple with balancing innovation speed and ethical responsibility. As AI technologies become more advanced and embedded in everyday applications, the need for comprehensive oversight and governance grows increasingly urgent.
The company has yet to publicly address these internal concerns, but the discussions underscore the complex challenges faced by major tech firms navigating AI's potential and pitfalls in a highly competitive landscape.
The Coming AI Clash: Utopia or Dystopia?
Khosla Ventures • Justine Sink • June 6, 2025
Technology•AI•Ethics•Innovation•FutureTech•Essays
In “The Coming AI Clash: Utopia or Dystopia?”, Vinod Khosla explores how AI could create a post-scarcity world of abundance—or lead to dystopia without proper safeguards.
The rapid advancement of AI technologies holds the promise of a transformative future, where automation and intelligent systems could dramatically increase productivity and quality of life. However, this future is not guaranteed to be positive. Khosla warns that without careful planning, regulation, and ethical considerations, AI could exacerbate inequalities, lead to mass unemployment, or even create new forms of existential risk.
Balancing innovation with responsibility, stakeholders must consider how AI development and deployment can foster equitable access to its benefits while minimizing potential harms. Discussions about AI governance, transparency, and long-term impacts are crucial to steer the technology towards a utopian outcome rather than a dystopian reality.
The debate captures a fundamental question about humanity's trajectory—will AI usher in an era of unparalleled human progress and abundance, or will it deepen social divides and create new challenges that threaten societal stability? Khosla’s insights call for a proactive approach to harness AI’s power wisely.
This A.I. Company Wants to Take Your Job
Mechanize, a San Francisco start-up, is building artificial intelligence tools to automate white-collar jobs “as fast as possible.”
By Kevin Roose, Reporting from San Francisco, June 11, 2025
Years ago, when I started writing about Silicon Valley’s efforts to replace workers with artificial intelligence, most tech executives at least had the decency to lie about it.
“We’re not automating workers, we’re augmenting them,” the executives would tell me. “Our A.I. tools won’t destroy jobs. They’ll be helpful assistants that will free workers from mundane drudgery.”
Of course, lines like those — which were often intended to reassure nervous workers and give cover to corporate automation plans — said more about the limitations of the technology than the motives of the executives. Back then, A.I. simply wasn’t good enough to automate most jobs, and it certainly wasn’t capable of replacing college-educated workers in white-collar industries like tech, consulting and finance.
That is starting to change. Some of today’s A.I. systems can write software, produce detailed research reports and solve complex math and science problems. Newer A.I. “agents” are capable of carrying out long sequences of tasks and checking their own work, the way a human would. And while these systems still fall short of humans in many areas, some experts are worried that a recent uptick in unemployment for college graduates is a sign that companies are already using A.I. as a substitute for some entry-level workers.
On Thursday, I got a glimpse of a post-labor future at an event held in San Francisco by Mechanize, a new A.I. start-up that has an audacious goal of automating all jobs — yours, mine, those of our doctors and lawyers, the people who write our software and design our buildings and care for our children.
“Our goal is to fully automate work,” said Tamay Besiroglu, 29, one of Mechanize’s founders. “We want to get to a fully automated economy, and make that happen as fast as possible.”
The dream of full automation isn’t new. John Maynard Keynes, the economist, predicted in the 1930s that machines would automate nearly all jobs, creating material abundance and leaving people free to pursue their passions.
That never happened, of course. But recent advances in A.I. have reignited the belief that technology capable of mass labor automation is near. Dario Amodei, the chief executive of Anthropic, recently warned that A.I. could displace as many as half of all entry-level white-collar jobs in the next five years.
Mechanize is one of a number of start-ups working to make that possible. The company was founded this year by Mr. Besiroglu, Ege Erdil and Matthew Barnett, who worked together at Epoch AI, a research firm that studies the capabilities of A.I. systems.
It has attracted investments from well-known tech leaders including Patrick Collison, a founder of Stripe, and Jeff Dean, Google’s chief A.I. scientist. It now has five employees, and is working with leading A.I. companies. (It declined to say which ones, citing confidentiality agreements.)
Mechanize’s approach to automating jobs using A.I. is focused on a technique known as reinforcement learning — the same method that was used to train a computer to play the board game Go at a superhuman level nearly a decade ago.
Disrupted or displaced? How AI is shaking up jobs
New technology is starting to have a profound effect on work and employment
Anjli Raval
During Ocado’s most recent earnings call, chief executive Tim Steiner said the group’s advances in artificial intelligence and robotics had allowed it to fulfil online grocery shops at an ever faster pace.
In 2012, it took 25 minutes of human labour to pick a 50-item order. That is now down to 10. But Ocado’s technological progress means the company requires 500 fewer workers this year, after it already announced 2,300 jobs would be at risk in 2023.
The UK company’s move over many years to phase down human labour where feasible exemplifies workers’ fears about generative AI: it may boost productivity, efficiency and profitability but it can also displace staff.
Some businesses are yet to embrace the shift but many have spent more than a year experimenting and engaging in workplace pilots.
“Companies are moving from asking, ‘What is our AI strategy?’ to experimenting . . . implementing generative AI into processes,” said Karin Kimbrough, chief economist at LinkedIn. “It is starting to change the landscape of work.”
Now employees, bosses and policymakers are trying to decipher what exactly the benefits of generative AI look like.
“This latest generation of AI could change every job. I don’t think that is too much of an exaggeration,” said Peter Cheese, chief executive of the Chartered Institute of Personnel and Development, the UK’s professional body for HR and people development. “Of course you can see examples where AI in different forms is already making a difference to their workforce, but it’s still early days for many companies.”
Many employers are cutting jobs under the guise of economic and political uncertainty. But high profile examples of AI-driven lay-offs in recent months, from technology company IBM to language learning app Duolingo, are fuelling questions about whether a slash and burn of white-collar roles is under way.
The 42-year-old billionaire Dario Amodei, who runs AI developer Anthropic, has warned the technology he and peers such as OpenAI are building could wipe out half of all entry-level office jobs in the next five years. Already, graduates account for just 7 per cent of hires across the 15 biggest technology companies, with the number of new recruits down a quarter compared with 2023, according to SignalFire, a venture capital firm.
“AI is starting to get better than humans at almost all intellectual tasks, and we’re going to collectively, as a society, grapple with it,” Amodei told television network CNN last month. “AI is going to get better at what everyone does, including what I do, including what other CEOs do.”
Academics, recruiters and management consultants are split on whether talk of a bloodbath is just scaremongering or a clear-eyed view of AI’s potential to shake up the labour market. But even if AI is not destroying jobs at scale today, it is certainly redesigning them and changing the equation between work, output and headcount.
“No sector is immune [to the impact of AI],” said Peter Brown, a global workforce expert at PwC. “But it is primarily changing roles not eliminating them, enabling humans to focus on more value-add elements of their jobs.”
Never Forget What They've Done
Wheres your ed • Edward Zitron • June 9, 2025
Technology•AI•TechIndustry•DigitalCulture•BigTechCriticism•Essays
I want my fucking tech industry back.
Maybe you think I sound insane, but technology means a lot to me. It’s the way that I speak to most of my friends. It’s my lifeline when I’m hurting or when those close to me hurt, and it’s the way I am able to make a living and be a creative — something I only was able to become because of technology. Social networks have been a huge part of me being able to become a functional human being, and you can judge me for that all you want, but you are a coward and a hypocrite for doing so, and you’re going to read to the end of this blog anyway.
Really, seriously, honestly — the Ed Zitron you know was and is only possible because of my deep connection to technology. This was how I made friends. This was how I got the confidence to meet real people. This was how I started my company. This was how I met the people closest to me, people I love with all my heart. I was only able to do any of this because I was able to get on the computer.
I am bombastic and frankly a little much today, and was the literal opposite less than 5 years ago, and I was even more reserved 10 years before that. Technology allowed me to find a way to be human on my terms, in ways that I don’t think are possible anymore because most of the interconnecting fabric that I used has been interfered with by bad actors and the rest with slop and SEO.
I think there are far more people out there like me than will admit to it. I think more people miss the past, or at least realize now what they lost.
There was a time this didn’t suck, when it wasn’t a struggle to do basic things, when my world was not a constant war with my god damn apps, when things weren’t necessarily turn-key but my phone wasn’t randomly burning through half of its battery life in an hour and a half because one app on the App Store is poorly configured. I swear to god, back in like, 2019, Zoom just fucking connected. I remember things being better, and on top of that, I see how much better things could be.
But that’s not the tech industry we’re allowed to have, because the people that run the tech industry do not give a shit.
It’s not enough to have your data, your work, your art, your posts, your friends, the things you’ve taken photos of, and the things you’ve searched for. The industry must have that of your children, and their children, as early as possible, even if it means helping them cheat on their homework so that they too can live a life where they’ve skipped having any responsibility or learning anything about the world other than how one can extract as much as possible without having to give anything in return.
Big tech is sociopathic and directionless, swinging wildly to try and find new ways to drag any kind of interaction out of a customer they’ve grown to loathe for their unwillingness to be more profitable. Decades of powerful Big Tech Business Idiots have chased out true value-creation in Silicon Valley in favour of growth economics, sending edict after edict down to the markets and the media about what’s going to be “hot” next, inventing business trends rather than actual solutions to problems. After all, that might involve — eugh! — experiencing the real world rather than authoring a new version of it every few years.
Apple barely escapes the void because its principle value proposition has, on some level, always been “our stuff works.” The problem is that Apple needs to grow, and thus its devices are slowly but surely becoming mired in sludge. The App Store is an abomination, your iPhone settings look like a fucking Escher painting, and in its desperation to follow the pack it shoved Apple Intelligence out the door — one of the most invasive and annoying pieces of software to ever grace a computer.
Apple’s willingness to do this shows that it’s rotten just like the rest of them — it's just better at hiding it. After all, look at the way in which it flaunted court orders telling it to open up third-party payments as a means of squeezing every penny out of the App Store. Loathsome. And it still ended up losing.
I adore tech. Tech made me who I am today. I use and love technology for hours a day, yet that experience is constantly mangled by the warring intentions of almost every product I use. I’m forced to log into the newspaper website and back into Google Calendar multiple times a week, my phone randomly resets — as every single iPhone has for multiple years — at least twice a week, my Apple Watch stops being willing to read my heart rate, websites I want to read sometimes simply do not load, and sometimes when I load websites on an iPad they just won’t scroll.
Everything feels like a fucking chore, but I love the actual things that technology does for me, like letting me take notes with ease, like building and maintaining my fitness through a series of connected products like Tonal and Fight Camp, like using Signal to talk to friends hundreds or thousands of miles away, like posting dumb stuff on Bluesky and interacting with my followers, like recording a podcast wherever I am in the world because USB-C mics are cheap and easy to use and sound great.
There are so many great things about technology, things I fucking love, and Large Language Models do not resemble their form or intention. There is nothing about an LLM that feels like it’s built to provide a real service, other than some sort-of fraudulent copy of something else lacking its soul or utility. Those that actually use them in their daily work talk about them as exciting tools that help them improve workflows - not like they're the next big thing.
Inside OpenAI’s Plan to Embed ChatGPT Into College Students’ Lives
Nytimes • June 7, 2025
Technology•AI•Education•ChatGPT•HigherEducation•Essays
OpenAI is actively integrating its AI chatbot, ChatGPT, into higher education to enhance learning experiences and administrative efficiency. In February 2025, OpenAI, supported by Microsoft, launched an education-specific version of ChatGPT, known as ChatGPT Edu, across 23 campuses of the California State University system, impacting approximately 500,000 students and faculty members. This initiative aims to provide personalized tutoring for students and assist faculty with administrative tasks. (reuters.com)
The deployment of ChatGPT Edu is part of OpenAI's broader strategy to embed AI into various facets of college life. By offering tailored AI tools, OpenAI seeks to support students in their academic endeavors and alleviate administrative burdens for educators. This approach reflects a growing trend in higher education to leverage AI technologies to improve learning outcomes and operational efficiency.
Despite initial concerns about potential misuse, such as cheating and plagiarism, OpenAI has been working towards integrating ChatGPT into classrooms since 2023. Universities like the Wharton School, the University of Texas, and the University of Oxford have already utilized ChatGPT Enterprise, leading to the release of ChatGPT Edu in May 2024. (reuters.com)
In response to the widespread adoption of AI tools, some educational institutions have lifted previous bans on ChatGPT. For instance, New York City public schools reversed their ban in May 2023, acknowledging the potential benefits of AI in education. Davis Banks, the head of New York City’s public schools at the time, stated that the initial decision to ban the tool was driven by a "knee-jerk fear [that] overlooked the potential of generative AI to support students and teachers." (en.wikipedia.org)
OpenAI's efforts to integrate ChatGPT into higher education signify a transformative shift in teaching and learning methodologies. By providing AI-driven tools, OpenAI aims to enhance educational experiences, promote personalized learning, and streamline administrative processes within academic institutions.
The Gentle Singularity
Sam Altman • June 10, 2025
Technology•AI•Superintelligence•Innovation•FutureTrends•Essays
We are past the event horizon; the takeoff has started. Humanity is close to building digital superintelligence, and at least so far it’s much less weird than it seems like it should be.
Robots are not yet walking the streets, nor are most of us talking to AI all day. People still die of disease, we still can’t easily go to space, and there is a lot about the universe we don’t understand.
And yet, we have recently built systems that are smarter than people in many ways, and are able to significantly amplify the output of people using them. The least-likely part of the work is behind us; the scientific insights that got us to systems like GPT-4 and o3 were hard-won, but will take us very far.
AI will contribute to the world in many ways, but the gains to quality of life from AI driving faster scientific progress and increased productivity will be enormous; the future can be vastly better than the present. Scientific progress is the biggest driver of overall progress; it’s hugely exciting to think about how much more we could have.
In some big sense, ChatGPT is already more powerful than any human who has ever lived. Hundreds of millions of people rely on it every day and for increasingly important tasks; a small new capability can create a hugely positive impact; a small misalignment multiplied by hundreds of millions of people can cause a great deal of negative impact.
2025 has seen the arrival of agents that can do real cognitive work; writing computer code will never be the same. 2026 will likely see the arrival of systems that can figure out novel insights. 2027 may see the arrival of robots that can do tasks in the real world.
A lot more people will be able to create software, and art. But the world wants a lot more of both, and experts will probably still be much better than novices, as long as they embrace the new tools. Generally speaking, the ability for one person to get much more done in 2030 than they could in 2020 will be a striking change, and one many people will figure out how to benefit from.
In the most important ways, the 2030s may not be wildly different. People will still love their families, express their creativity, play games, and swim in lakes.
But in still-very-important-ways, the 2030s are likely going to be wildly different from any time that has come before. We do not know how far beyond human-level intelligence we can go, but we are about to find out.
In the 2030s, intelligence and energy—ideas, and the ability to make ideas happen—are going to become wildly abundant. These two have been the fundamental limiters on human progress for a long time; with abundant intelligence and energy (and good governance), we can theoretically have anything else.
Already we live with incredible digital intelligence, and after some initial shock, most of us are pretty used to it. Very quickly we go from being amazed that AI can generate a beautifully-written paragraph to wondering when it can generate a beautifully-written novel; or from being amazed that it can make live-saving medical diagnoses to wondering when it can develop the cures; or from being amazed it can create a small computer program to wondering when it can create an entire new company. This is how the singularity goes: wonders become routine, and then table stakes.
We already hear from scientists that they are two or three times more productive than they were before AI. Advanced AI is interesting for many reasons, but perhaps nothing is quite as significant as the fact that we can use it to do faster AI research. We may be able to discover new computing substrates, better algorithms, and who knows what else. If we can do a decade’s worth of research in a year, or a month, then the rate of progress will obviously be quite different.
From here on, the tools we have already built will help us find further scientific insights and aid us in creating better AI systems. Of course this isn’t the same thing as an AI system completely autonomously updating its own code, but nevertheless this is a larval version of recursive self-improvement.
There are other self-reinforcing loops at play. The economic value creation has started a flywheel of compounding infrastructure buildout to run these increasingly-powerful AI systems. And robots that can build other robots (and in some sense, datacenters that can build other datacenters) aren’t that far off.
If we have to make the first million humanoid robots the old-fashioned way, but then they can operate the entire supply chain—digging and refining minerals, driving trucks, running factories, etc.—to build more robots, which can build more chip fabrication facilities, data centers, etc, then the rate of progress will obviously be quite different.
As datacenter production gets automated, the cost of intelligence should eventually converge to near the cost of electricity. (People are often curious about how much energy a ChatGPT query uses; the average query uses about 0.34 watt-hours, about what an oven would use in a little over one second, or a high-efficiency lightbulb would use in a couple of minutes. It also uses about 0.000085 gallons of water; roughly one fifteenth of a teaspoon.)
The rate of technological progress will keep accelerating, and it will continue to be the case that people are capable of adapting to almost anything. There will be very hard parts like whole classes of jobs going away, but on the other hand the world will be getting so much richer so quickly that we’ll be able to seriously entertain new policy ideas we never could before. We probably won’t adopt a new social contract all at once, but when we look back in a few decades, the gradual changes will have amounted to something big.
If history is any guide, we will figure out new things to do and new things to want, and assimilate new tools quickly (job change after the industrial revolution is a good recent example). Expectations will go up, but capabilities will go up equally quickly, and we’ll all get better stuff. We will build ever-more-wonderful things for each other. People have a long-term important and curious advantage over AI: we are hard-wired to care about other people and what they think and do, and we don’t care very much about machines.
A subsistence farmer from a thousand years ago would look at what many of us do and say we have fake jobs, and think that we are just playing games to entertain ourselves since we have plenty of food and unimaginable luxuries. I hope we will look at the jobs a thousand years in the future and think they are very fake jobs, and I have no doubt they will feel incredibly important and satisfying to the people doing them.
The rate of new wonders being achieved will be immense. It’s hard to even imagine today what we will have discovered by 2035; maybe we will go from solving high-energy physics one year to beginning space colonization the next year; or from a major materials science breakthrough one year to true high-bandwidth brain-computer interfaces the next year. Many people will choose to live their lives in much the same way, but at least some people will probably decide to “plug in”.
Looking forward, this sounds hard to wrap our heads around. But probably living through it will feel impressive but manageable. From a relativistic perspective, the singularity happens bit by bit, and the merge happens slowly. We are climbing the long arc of exponential technological progress; it always looks vertical looking forward and flat going backwards, but it’s one smooth curve. (Think back to 2020, and what it would have sounded like to have something close to AGI by 2025, versus what the last 5 years have actually been like.)
There are serious challenges to confront along with the huge upsides. We do need to solve the safety issues, technically and societally, but then it’s critically important to widely distribute access to superintelligence given the economic implications. The best path forward might be something like:
Solve the alignment problem, meaning that we can robustly guarantee that we get AI systems to learn and act towards what we collectively really want over the long-term (social media feeds are an example of misaligned AI; the algorithms that power those are incredible at getting you to keep scrolling and clearly understand your short-term preferences, but they do so by exploiting something in your brain that overrides your long-term preference).
Then focus on making superintelligence cheap, widely available, and not too concentrated with any person, company, or country. Society is resilient, creative, and adapts quickly. If we can harness the collective will and wisdom of people, then although we’ll make plenty of mistakes and some things will go really wrong, we will learn and adapt quickly and be able to use this technology to get maximum upside and minimal downside. Giving users a lot of freedom, within broad bounds society has to decide on, seems very important. The sooner the world can start a conversation about what these broad bounds are and how we define collective alignment, the better.
We (the whole industry, not just OpenAI) are building a brain for the world. It will be extremely personalized and easy for everyone to use; we will be limited by good ideas. For a long time, technical people in the startup industry have made fun of “the idea guys”; people who had an idea and were looking for a team to build it. It now looks to me like they are about to have their day in the sun.
OpenAI is a lot of things now, but before anything else, we are a superintelligence research company. We have a lot of work in front of us, but most of the path in front of us is now lit, and the dark areas are receding fast. We feel extraordinarily grateful to get to do what we do.
Intelligence too cheap to meter is well within grasp. This may sound crazy to say, but if we told you back in 2020 we were going to be where we are today, it probably sounded more crazy than our current predictions about 2030.
May we scale smoothly, exponentially and uneventfully through superintelligence.
Is Google about to destroy the web?
Thomas Germain,
Google says a new AI tool on its search engine will rejuvenate the internet. Others predict an apocalypse for websites. One thing is clear: the current chapter of online history is careening towards its end. Welcome to the "machine web".
The web is built on a simple bargain – websites let search engines like Google slurp up their content, free of charge, and Google Search sends people to websites in exchange, where they buy things and look at adverts. That's how most sites make money.
An estimated 68% of internet activity starts on search engines and about 90% of searches happen on Google. If the internet is a garden, Google is the Sun that lets the flowers grow.
This arrangement held strong for decades, but a seemingly minor change has some convinced that the system is crumbling. You'll soon see a new AI tool on Google Search. You may find it very useful. But if critics' predictions come true, it will also have seismic consequences for the internet. They paint a picture where quality information could grow scarcer online and large numbers of people might lose their jobs. Optimists say instead this could improve the web's business model and expand opportunities to find great content. But, for better or worse, your digital experiences may never be the same again.
On 20 May 2025, Google's chief executive Sundar Pichai walked on stage at the company's annual developer conference. It's been a year since the launch of AI Overviews, the AI-generated responses you've probably seen at the top of Google Search results. Now, Pichai said, Google is going further. "For those who want an end-to-end AI Search experience, we are introducing an all-new AI Mode," he said. "It's a total reimagining of Search."
You might be sceptical after years of AI hype, but this, for once, is the real deal.
If Google makes AI Mode the default in its current form, it's going to have a devastating impact on the internet – Lily Ray
People use Google Search five trillion times a year – it defines the shape of the internet. AI Mode is a radical departure. Unlike AI Overviews, AI Mode replaces traditional search results altogether. Instead, a chatbot effectively creates a miniature article to answer your question. As you read this, AI Mode is rolling out to users in the US, appearing as a button on the search engine and the company's app. It's optional for now, but Google's head of Search, Liz Reid, said it plainly when launching the tool: "This is the future of Google Search."
Here's the problem critics foresee – AI Overviews already sends much less traffic to the rest of the internet, and many fear AI Mode could supercharge that trend. If this comes to pass, it could crush the business model that's fuelled the digital content you've enjoyed for almost 30 years.
TechCrunch
Europe, we’re not leaving. Period.
Techcrunch • June 11, 2025
Business•Media•TechJournalism•Europe•Global Expansion•TechCrunch
Recent headlines suggesting that TechCrunch is "pulling out of Europe" and ceasing coverage of European startups are entirely incorrect. Such claims misrepresent our mission and values. The recent changes at TechCrunch are not about retreat but about realignment and reinforcement. This new chapter is fueled by our partnership with our sister company, Foundry, which was brought under the same ownership to create a tech media entity with unparalleled global scope.
To be clear, Foundry is a powerhouse of international technology journalism. Its portfolio includes established and respected brands like PCWorld, Macworld, CIO, and TechAdvisor, with a vast network of journalists and deep-rooted expertise in local and regional tech ecosystems across Europe and the world. The suggestion that our new ownership believes international coverage is unessential is patently false. The entire purpose of bringing TechCrunch and Foundry together is to create a stronger, more globally focused media platform.
Europe is where fintech regulation is rewritten, where quantum startups spin out of Max Planck labs, where climate-tech pilots become the standard for the rest of the planet. In 2024 alone, European founders raised over €40 billion; many of the unicorns we covered last year were born on this continent. If you care about the future of technology, you have to be here. And we are.
As we integrate the strengths of both TechCrunch and Foundry, here is our promise to the founders, investors, and readers in Europe and beyond:
Radical presence. We will be on the ground — from demo days in Tallinn to hydrogen-hub unveilings in the Basque Country — because stories look different up close.
Relentless scrutiny. Hype dies in daylight. We’ll keep asking the annoying questions and digging for the real story so you don’t have to.
A true global megaphone. A breakthrough in Zagreb deserves the same volume as one in Silicon Valley. By merging the TechCrunch and Foundry networks, we will amplify European innovation to millions more readers in 190+ countries, providing deeper insights and a truly unified global vision.
To the startup community: Keep your tips, term sheets, and tantrums coming. Send them to tips@techcrunch.com or ping our encrypted channels. We’ll be listening — louder and more broadly than ever.
TechCrunch isn’t retreating from Europe. We’re doubling down.
Twenty years strong: a love letter to TechCrunch
Techcrunch • June 11, 2025
Media•Technology•TechJournalism•Startups•Innovation•TechCrunch
TechCrunch is celebrating its 20th anniversary, marking two decades of delivering in-depth technology news and analysis. Over the years, the publication has become a trusted source for insights into the tech industry, covering a wide range of topics from startups to major tech companies.
The journey began in June 2005 when TechCrunch was founded by Archimedes Ventures, led by Michael Arrington and Keith Teare. Their vision was to create a platform that provided timely and comprehensive coverage of the rapidly evolving tech landscape. This vision quickly resonated with readers, and TechCrunch grew to become a leading voice in technology journalism.
Throughout its history, TechCrunch has been known for its unique editorial approach. Writers often assign stories to themselves, allowing them to delve deeply into subjects they are passionate about. This method fosters a culture of curiosity and thoroughness, enabling the publication to produce insightful and engaging content. (techrookies.com)
In addition to its editorial endeavors, TechCrunch has played a pivotal role in the startup ecosystem through events like TechCrunch Disrupt. This annual conference brings together entrepreneurs, investors, and tech enthusiasts to showcase innovative startups and discuss industry trends. Over the years, Disrupt has featured numerous companies that have gone on to achieve significant success, underscoring TechCrunch's influence in the tech community. (techrookies.com)
The publication's commitment to quality journalism has also been evident in its investigative reporting. TechCrunch has consistently held companies accountable, uncovering stories that have had a lasting impact on the industry. This dedication to truth and transparency has solidified its reputation as a reliable and authoritative source of information.
As TechCrunch celebrates this significant milestone, it reflects on its journey from a startup-focused blog to a comprehensive tech news outlet. The publication's success is a testament to the hard work and passion of its team, as well as the trust and support of its readers. Looking ahead, TechCrunch remains committed to its mission of providing insightful and timely coverage of the ever-changing tech landscape.
Venture Capital
Marc Andreessen | The Future of Venture Capital
Youtube • Uncapped with Jack Altman • June 11, 2025
Business•VentureCapital•ArtificialIntelligence•HumanIntuition•InvestmentProcess•Venture Capital
In a recent discussion, Marc Andreessen, co-founder of Andreessen Horowitz, delved into the evolving landscape of venture capital, emphasizing the critical role of human intuition and relationships in the investment process. He highlighted that while artificial intelligence (AI) is increasingly integrated into various sectors, the essence of venture capital remains deeply human-centric.
Andreessen argued that venture capital is more art than science, particularly in investor relationships and mentoring founders—elements he believes AI cannot replicate. He noted that even with advancements in AI, the human touch in understanding a founder's psychology and the nuances of a startup's journey is irreplaceable. This perspective underscores the enduring importance of human judgment in the investment process.
However, the conversation also touched upon the paradox within the industry: venture capitalists are investing heavily in AI startups that could ultimately automate parts of their own industry. This raises questions about the future of venture capital and its ability to adapt to technological advancements. Andreessen acknowledged this challenge, suggesting that while AI may transform certain aspects of the industry, the foundational human elements of venture capital are likely to persist.
In summary, Andreessen's insights provide a nuanced view of the future of venture capital, balancing the potential of AI with the irreplaceable value of human expertise and relationships.
Top AI Series B lead investors
Signalrankupdate • Rob Hodgkinson • June 9, 2025
Technology•AI•VentureCapital•SeriesB•Investments•Venture Capital
In April, DST Global's Cole Rotman created a sheet to track Series A leads in every $5bn+ venture-backed company. This initiative aimed to identify the most active and successful investors in the Series A funding stage. The sheet provided valuable insights into investor behavior, highlighting those who consistently led significant funding rounds and demonstrated a strong track record in identifying promising startups.
Building upon this, a similar approach was applied to Series B funding rounds, focusing specifically on the artificial intelligence (AI) sector. By analyzing data from various AI startups that secured Series B funding, the goal was to pinpoint the leading investors who played a pivotal role in supporting and scaling AI innovations.
The analysis revealed several key investors who have been instrumental in driving the growth of AI companies through their Series B investments. Notable among them are:
Andreessen Horowitz (a16z): A prominent venture capital firm known for its significant investments in AI startups. They have led substantial Series B rounds for companies like SentiLink, a cybersecurity enterprise software firm specializing in identity verification technology. In 2021, Andreessen Horowitz led a $70 million Series B funding round for SentiLink. (topstartups.io)
Khosla Ventures: Another influential investor in the AI space, Khosla Ventures led a $74 million Series B funding round for Bitwise Asset Management in 2021, valuing the company at $500 million. Bitwise is a crypto index fund manager catering to individuals, investment managers, and institutions navigating the cryptocurrency landscape. (topstartups.io)
Y Combinator: A renowned startup accelerator, Y Combinator has been actively involved in funding AI companies. They led a $123 million Series B funding round for Observe.AI in 2020. Observe.AI develops a voice AI platform for contact centers, aiming to enhance agent performance through advanced AI solutions. (topstartups.io)
These investors have demonstrated a keen interest in the AI sector, recognizing its transformative potential across various industries. Their strategic investments have not only provided essential capital to AI startups but also offered valuable mentorship and industry connections, contributing significantly to the maturation and success of these companies.
The continued support from such investors underscores the growing confidence in AI technologies and their capacity to drive innovation and economic growth. As AI continues to evolve, the role of these investors will remain crucial in fostering the next generation of groundbreaking AI solutions.
Elon’s Empire: SpaceX, Tesla, Neuralink After the Storm & Anduril’s $2.6BN Power Move
Youtube • 20VC with Harry Stebbings • June 12, 2025
Technology•Innovation•SpaceExploration•ElectricVehicles•NeuralInterfaces•DefenseTech•Venture Capital
This discussion centers on Elon Musk's sprawling business empire and the latest developments involving his companies SpaceX, Tesla, Neuralink, and the defense technology firm Anduril. The video explores the current state and future prospects of these ventures after facing various internal and external challenges.
SpaceX remains at the forefront of space exploration innovation, continuing to push the boundaries of reusable rocket technology and expanding satellite internet services through Starlink. Despite facing regulatory and competitive pressures, SpaceX's rapid development cycle and ambitious launch cadence remain impressive.
Tesla, on the other hand, is highlighted for its ongoing advancements in electric vehicle technology, battery production, and energy solutions. The company is strategically navigating competitive pressures in the automotive sector and increasing efforts toward global market penetration, particularly in emerging economies.
Neuralink's progress is explored in terms of its ambitious goal to integrate human cognition with advanced neural interfaces. The potential implications for healthcare, artificial intelligence, and human-machine synergy are discussed, as well as the technical and ethical challenges that come with this frontier technology.
Finally, the conversation turns to Anduril, a defense startup with a recent $2.6 billion valuation boost. Anduril focuses on leveraging AI and autonomous systems for national security applications. This significant investment demonstrates rising confidence in cutting-edge defense technologies and the company's rapid growth trajectory amid an evolving geopolitical landscape.
Each company operates in distinct yet interconnected arenas, collectively illustrating Elon Musk's vision of shaping a future driven by technological breakthroughs across transportation, energy, biology, and defense.
Making Secondaries a Primary Thing
Therandomwalk • June 7, 2025
Finance•PrivateEquity•SecondaryMarket•VentureCapital•Liquidity•Venture Capital
The secondary market in private capital, particularly within venture capital, has experienced significant growth and transformation in recent years. This evolution is largely driven by the increasing liquidity demands of General Partners (GPs) and Limited Partners (LPs), prompting a shift towards secondary transactions as a primary strategy for portfolio management and capital distribution.
Ben Slome, a representative from New Vintage Partners, a dedicated secondary fund, provides insight into this dynamic landscape. He highlights the under-served liquidity needs of Venture GPs, emphasizing the challenges they face in managing aging portfolios and the necessity for innovative solutions to address these demands.
The surge in secondary market activity is underscored by several key developments:
Record Transaction Volumes: In the first half of 2024, venture fund secondaries saw a 151.3% increase in volume compared to the same period in 2023, reaching $2.7 billion. This surge was notably influenced by significant transactions, such as Lexington Partners' acquisition of a $1 billion venture portfolio. (axavp.com)
Expansion of Secondary Funds: The number and size of funds dedicated to venture capital secondaries have grown substantially. For instance, StepStone Group raised a $3.3 billion fund dedicated to VC secondaries, marking a 20% increase from its 2022 predecessor. (axavp.com)
Liquidity Constraints: Venture capital funds have faced challenges in generating liquidity for their investors, with 2023 marking the second consecutive year of negative net cash flow. This trend underscores the pressing need for effective secondary market solutions to provide timely capital returns to LPs. (alts.co)
The prominence of secondary transactions is further highlighted by the concentration of capital among a few dominant firms, which control the majority of the market share. This concentration influences pricing dynamics and access to high-quality assets within the secondary market. (londonvcnetwork.com)
In summary, the secondary market has become a pivotal component of the private capital ecosystem, offering essential liquidity solutions to address the evolving needs of Venture GPs and LPs. As the market continues to mature, secondary transactions are expected to play an increasingly central role in portfolio management and capital distribution strategies.
Patterns Across 5 Years of YC Investing
Tomtunguz • June 8, 2025
Technology•Startups•InvestmentTrends•VentureCapital•Innovation•Venture Capital
Where venture capital flows, innovation follows. And for more than a decade, few faucets have been watched more closely than Y Combinator. An analysis of their investment patterns since 2020 doesn’t just reveal the accelerator’s strategy—it provides a map to the entire startup ecosystem’s next chapter.
With Demo Day approaching this week & inspired by Jamesin Seidel’s YC Series A analysis, I wondered how YC investment patterns have changed since 2020.
Cybersecurity and industrial/manufacturing are the two fastest growing categories. Education & life sciences are right behind. The Wiz acquisition and the overall growth rates of security companies as a durable budget within software spending has propelled security more broadly. Similarly manufacturing startups have seized on the tariff-induced reshoring opportunity.
B2B companies have increased their share from roughly 80 to 90% over the last five years, which is a parallel to the broader venture industry.
Crypto/web3 remains around 5% of investments. The 2022 spike followed the Coinbase IPO in 2021. It’s a steady but not a very large fraction of companies.
AI companies on average raise a little bit more, but the delta is not yet statistically significant - even though AI companies broadly do raise a premium.
Ultimately, YC’s portfolio mirrors the broader industry’s shift toward pragmatism. The significant growth isn’t in speculative tech, but in essential tools for manufacturing, security, and B2B. The takeaway is clear: the surest path to funding runs straight through solving a customer’s most expensive problems.
New Unicorns Add $22B In Value In May
Crunchbase • June 10, 2025
Business•Startups•Unicorns•Europe•Innovation•Venture Capital
Thirteen companies joined The Crunchbase Unicorn Board in May 2025, including five from Europe, Crunchbase data shows.
The five new unicorns from Europe mark the highest monthly count of new billion-dollar startups since 2023 for the continent. They included the first two from Germany and the first company from Portugal so far this year to be valued at $1 billion-plus. The U.K. also added two companies last month, marking three total this year.
Six companies joined from the U.S., adding up to 31 so far this year. And two companies joined from India, adding up to three companies in 2025 year to date.
Collectively, these 13 companies added $21.7 billion in value to the board in May.
Sales and marketing, and defense tech — sectors impacted by AI — led for new unicorn companies in May, with two each.
Exits
Six companies exited the board in May, removing $13.4 billion in value.
They include four unicorn companies that went public last month: Israel-based social trading platform eToro, San Francisco-based digital clinic Hinge Health, India-based electric scooter manufacturer Ather Energy, and Austin, Texas-based advertising platform MNTN. Each of these companies went public at or above their last known valuation, except for Hinge Health which was last valued at $6.2 billion and debuted at $2.6 billion.
Two unicorns were acquired. Coding startup Windsurf, last valued at $1.1 billion in 2024 was acquired by OpenAI for $3 billion. Daily Harvest, a direct to consumer snack company known for its frozen smoothies, valued at $1.1 billion in 2021, was acquired by Chobani for an undisclosed amount.
These developments underscore the dynamic growth and innovation within the startup landscape.
Defense Tech Unicorn Anduril Powers Up With $2.5B At $30.5B Valuation
Crunchbase • Judy Rider • June 5, 2025
Technology•AI•DefenseTech•Funding•Startups•Venture Capital
Defense tech startup Anduril Industries has raised $2.5 billion in a Series G round of funding, more than doubling its valuation to $30.5 billion post-money, the company’s chairman, Trae Stephens, told Bloomberg TV on Thursday.
Founders Fund, a venture firm where Stephens is also a partner, led the financing. Stephens told Bloomberg TV that as Anduril continued to grow, it was “really important to shore up the balance sheet and make sure” it had “the ability to deploy capital into these manufacturing and production problem sets that we’re working on.”
Since its 2017 inception, Costa Mesa, California-based Anduril has raised more than $6 billion in funding, per Crunchbase data. Other investors include General Catalyst, Andreessen Horowitz, Valor Equity Partners, BlackRock, and Lightspeed Venture Partners.
Founders Fund, which co-led Anduril’s $1.5 billion Series F in July 2024, contributed $1 billion to the latest financing, according to Stephens. He noted that it was the largest check the firm had ever written.
Palmer Luckey co-founded Anduril after being fired from Meta, the company to whom he’d sold his previous startup, Oculus. The two companies have since buried the hatchet, it seems, with Anduril announcing on May 29 that they are working together to build extended reality devices for the U.S. military.
Funding to VC-backed startups in defense — defined here as the industries of military, national security, and law enforcement — hit $3 billion in 102 deals last year, per Crunchbase data. That’s only an 11% uptick from 2023, which saw $2.7 billion raised in 100 announced rounds.
With this latest Anduril round alone totaling almost all of last year’s funding to the sector, expect this year to see a significant bump.
IPO
The Great IPO Awakening: What 2025’s Surprisingly Hot Market Means for B2B Companies
Saastr • Jason Lemkin • June 9, 2025
Business•Strategy•SaaS•IPO•MarketTrends
The IPO window isn't just cracked open—it's wide open. After years of drought, 2025 has delivered a scorching hot public market for tech companies so far, with some eye-popping returns that should have every SaaS founder and investor paying attention.
Let's start with the headline grabber: Circle's 247% gain from IPO price to current trading levels. That's not a typo. The company that filed at $24-26 per share and priced at $31 is now trading at levels that would make even the most seasoned growth investor do a double-take.
But Circle isn't alone. Looking at the last 20 tech IPOs, we're seeing an average first-day pop of 31% and current returns averaging 76.8% above IPO price. Even more telling? The most recent five IPOs are averaging 121.5% returns from their initial pricing.
The Morgan Stanley data reveals something fascinating: 10 of the last 20 tech IPOs happened in 2025, with the majority clustered in the past few months. This isn't coincidence—it's companies and bankers recognizing a fundamental shift in market appetite:
The average IPO size has been $887M, but here's the kicker: some of the best performers are actually smaller deals. Reddit ($860M) delivered 256% returns, while Arm's massive $5.2B deal "only" returned 161%. The lesson? You don't need to be a mega-round to capture mega-returns.
While not every company in this dataset is B2B or pure SaaS, the implications for software companies are massive:
With companies like ServiceTitan (filed at 52-57x revenue based on typical SaaS metrics) and others commanding premium valuations, we're seeing a return to growth-focused pricing. The market is once again rewarding recurring revenue models and predictable growth patterns.
Here's a nugget most people miss: companies are consistently pricing above their initial filing ranges. SailPoint filed at $19-21 but priced at $23. This "pricing power" indicates genuine demand, not just investment bank optimism.
The window is open, but it won't stay open forever. Market conditions this favorable historically last 12-18 months before either cooling off or becoming overheated. Companies with $100M+ ARR and 30%+ growth should seriously evaluate their IPO readiness.
Use this data to set expectations with investors. The public market appetite for tech validates higher private valuations, but don't get caught up in the hype. Focus on the fundamentals that make companies IPO-ready: predictable revenue growth, expanding margins, and clear path to profitability.
Not everything is roses. CoreWeave's -21.6% file-to-offer performance and SailPoint's negative current returns remind us that execution still matters. Even in hot markets, companies with weaker fundamentals or just poorer timing get punished.
Strip away the outliers, and the average first-day pop is 10.3%—healthy but not euphoric. This suggests the market is pricing most deals fairly, with exceptional companies (like Circle) earning exceptional returns.
We’re witnessing a fundamental reset in how the public markets value technology companies. The combination of AI excitement, economic stability, and pent-up demand from years of IPO drought has created conditions we haven't seen since 2021.
But here's the thing: unlike 2021's "everything rally," this market appears more discerning. Companies with real revenue, real growth, and real paths to profitability are getting rewarded. Those without are getting left behind.
For B2B companies, the message is clear: if you've been building a real business with strong unit economics and predictable growth, the public markets are ready to reward you. The question isn't whether you should consider going public—it's whether you can afford not to.
The IPO window is open. The question is: are you ready to walk through it?
The IPO Door is Swinging Open
Tomtunguz • June 5, 2025
Business•Startups•IPO•Cryptocurrency•DataCenters
Who could have predicted that crypto and data center real estate would be the categories swinging the IPO market doors open?
In late 2024, I predicted a thaw in the IPO market. We’re now seeing that forecast come to life with CoreWeave and Circle’s IPOs. Neither company is pure-play software, but their strong performances signal renewed investor appetite for the ragged edge of technology.
CoreWeave went public in March 2025, raising $1.5 billion at a $20 billion valuation. The GPU infrastructure company has since soared over 300% on 420% year-over-year revenue growth, reaching nearly $1 billion in Q1 alone.
Circle’s IPO on June 5th is a parallel story. Shares jumped 168% on the first day, closing at $83.23, up 168%, pushing the market cap over $18 billion.
The success reflects investor appetite for exposure to stablecoins - the fastest growing part of crypto and the first clear product market fit in the mass-market.
These IPOs highlight a broader trend: the reopening of the IPO window is not just a blip. Despite political uncertainties and mixed economic signals—the ongoing tussle between Donald Trump and Elon Musk, less-than-stellar jobs reports—appetite for high-growth, category-defining companies is unmistakable.
This resurgence dovetails with a significant M&A surge this year, pointing to robust liquidity in the back half of 2025.
The IPO window is swinging open.
Apple
Apple updates Spotlight to take actions on your Mac
Techcrunch • Maxwell Zeff • June 9, 2025
Technology•Software•macOS•Spotlight•Productivity•Apple
Apple has announced significant updates to Spotlight, the Mac's on-device search feature, during WWDC 2025. In macOS Tahoe, Spotlight now allows users to run shortcuts and perform hundreds of actions, such as composing emails, setting reminders, and playing podcasts.
This enhancement positions Spotlight as a central hub for various tasks, integrating functionalities that were previously spread across multiple applications. The visual redesign of Spotlight enables users to browse through apps, files, clipboard history, and more, with results intelligently ranked based on relevance.
Apple has introduced "quick keys," short strings of characters that provide direct access to specific apps or actions. For instance, typing "SM" sends a message, and "AR" adds a new reminder. Users can also assign quick keys to custom functions within Spotlight. Developers can make their apps discoverable in Spotlight using the App Intents API.
In a demonstration, an Apple executive showcased how users could send an entire email from Spotlight, including specifying the recipient, subject line, and message, and even sending it without opening another app. This feature aims to streamline workflows by reducing the need to switch between applications.
Additionally, Spotlight is now integrated with Apple Intelligence, making it contextually aware. It suggests actions that users commonly take based on their current activities, enhancing productivity by anticipating needs.
These updates reflect Apple's commitment to enhancing user experience by consolidating functionalities and leveraging intelligent features within macOS Tahoe.
The Mind-Blowing Mac-ness of iPadOS 26
Spyglass • June 12, 2025
Technology•Software•iPadOS•MacIntegration•Multitasking•Apple
Apple embraces the blurring lines between iPad and Mac. Finally.
One slightly strange part of the WWDC Keynote came immediately after Craig Federighi introduced iPadOS 26, which beyond the fresh coat of 'Liquid Glass' paint that the other OSes received, also is being completely overhauled, at least when it comes to multitasking on the device. "Wow. More windows, a pointier pointer, and a menu bar? Who would've thought? We've truly pulled off a mind-blowing release!"
It was, of course, sarcasm. But he didn't really land it. Because it was also a sort of strange acknowledgement that perhaps Apple should have just been doing things this way all along. Which is to say, like a Mac.
So why didn't they until – checks calendar – some 15 years after Steve Jobs first sat down in the comfortable chair on stage with the device? Some of it, as Federighi talks about in this interview were technical limitations. The first several iterations of the iPad were certainly more akin, hardware-wise, to an iPhone and not a Mac – bust those "just a big iPhone" jokes out of cold storage. But clearly just as big of a part was that Apple really, really wanted the iPad to be a different type of device. Filling a space in between the iPhone and a Mac, just as Jobs envisioned.
I wrote about this issue a year ago, just after Apple unveiled the M4-powered iPad Pro – the first Apple product to get that chip, yes, even before the Mac. In a post entitled "Apple Invents a Laptop with a Touchscreen":
To me, perhaps the most interesting element of Apple's "Let Loose" iPad event yesterday was an almost throw-away line. "Feels just like using a MacBook," John Ternus, Apple's SVP of Hardware Engineering, said of the trackpad on Apple's new Magic Keyboard accessory for the iPad Pro.
That's great. But it's also a weird thing for an Apple executive to say...
For years at this point we've been told that the iPad and MacBook are two distinct Apple products and never the two shall meet. But increasingly, the two have been meeting. Trading notes back and forth, as it were, as they continue to bleed into one device. Really only one thing holds the two sides from fully coming together at this point: a touchscreen.
I think for a lot of people – myself included – the iPad did find a way to reside in this magical realm between the other two devices. But as time went on, and as iPhone screens got ever-larger and Macs got even more portable, the iPad was getting squeezed. At the same time, it remained in a unique position for both young people (my children) and older people (my mom):
As the years of using both devices go by, I find myself increasingly reaching up to touch my MacBook screen out of habit. And I grew up on PCs. My five year old simply doesn't understand that touching a MacBook screen does nothing. That's how computers work to her. Meaning, of course, the iPad.
But the realities of our world mean that you still can't really use the iPad computing paradigms for all of the computing you need to do. And Apple obviously should have morphed the iPad to meet these opportunities, but they dragged their feet:
So why has Apple waited so long? Well, first and foremost because Steve Jobs said so. Noting that touchscreens were "ergonomically terrible" at the launch of the iPad in 2010, he went on to say that: "Touch surfaces don't want to be vertical. It gives great demo but after a short period of time, you start to fatigue. And after an extended period of time, your arm wants to fall off. It doesn't work." And so Apple has gone out of their way in these past 14 years to ensure that they won't work on the Mac because they won't be included on the Mac.
Yes, Apple would have to do a bunch of work on the software side to make macOS touch-ready. But it's Apple. They have the resources and capabilities to do such things. And actually, what I really want first is something they can do right now, if they choose to: the ability to let your iPad Pro dual boot iPadOS and macOS. Even before the touchscreen conversion, if you have a Magic Keyboard attached, just let me use the iPad as if it's the non-touchscreen monitor of a MacBook.
But Apple won't do this because it not only blurs but basically erases the line between the iPad and the Mac. Users want this, Apple does not. Why? Again, legacy. These are two separate products, as told by Steve Jobs.
The platform has matured enough that people want to use it to do everything – including things you can do on both the iPhone and Mac. And the latter has been more or less impossible not because of the device itself any longer – that situation has inverted, where the hardware is now more akin to a Mac rather than an iPhone – but because of the software. And that has been the case for many years now. And rather than do the obvious thing – giving users an option to use their device more like a Mac – Apple tried to come up with all these convoluted ways to handle PC workflows with an iPad twist. Now they're removing the twist. Finally.
AI
What's Working for YC Companies Since the AI Boom.
Jamesin • Jamesin Seidel • May 30, 2025
Technology•AI•Startups•BusinessAutomation•VentureCapital
The goal of this analysis was to understand what’s working for YC companies by looking at which ones have raised Series A rounds.
I was really excited to take a look at the data for a number of reasons:
We finally have 2 years of batch data since ChatGPT started the AI boom, which gives us some meat to analyze.
YC is always interesting to dig into since it still has the cachet to attract top talent, and the batches are big enough to have a real population of companies that are representative of the market.
Demo day is coming up, and we’ve been tracking the latest batch.
First Round just announced Series A rounds for Reducto and David AI, and they have such an impressive hit rate - so I got curious whether there are actually tons of YC Series A rounds happening, or if First Round just crushes it (turns out they just crush it).
Before diving into the findings, I’ll say the data is skewed and has a lot of limitations. For example, companies that raised large initial rounds (like Worldware's $30M from Spark) aren't going back to market anytime soon. We’re only seeing what gets announced on Crunchbase. I am only looking at the four batches directly after ChatGPT launched. The goal here is a quick snapshot of “what's working” since the market is moving so quickly, rather than a full breakdown of all YC batches or getting any answer to the question, “Does the YC model still work?”
Final caveat - the topline number is: out of 998 companies from the first four batches since ChatGPT launched (Winter 2023, Summer 2023, Winter 2024, Summer 2024), only 24 have raised Series A rounds - just 2.4%.
But this isn’t a great number to lean into as a takeaway. Most of these companies haven't had enough time to mature, since the typical Seed-to-Series A timeline is around 18 months, and only the Winter 2023 batch fits that window.
So please take this post as surfacing directional trends rather than definitive statements.
Ok! Let’s get into it. Here is what I found looking at Series A data since the Winter 2023 batch —
Key takeaways from the 24 Series A companies include:
Business automation and operational tooling dominance. Probably the most surprising part of this analysis was how many of the winners were in internal business automation and operational platforms. The data looked different than the “what’s working in AI” analysis I did last year, which looked at Series A AI companies more broadly. There was less diversity here. This suggests the two obvious things:
Network advantages. The batches provide a built-in customer base for B2B operations and automation that can drive success (e.g, Deel, Brex). I don’t know who the customers are for the successful companies in this category, but I’m curious how many of them were accelerated by the YC batch advantage.
Technical talent. YC’s founder tends to be young, technical, and good at executing, which is an archetype that naturally gravitates towards automation infrastructure, developer tools, and general optimizations, perhaps rather than heavy vertically embedded software.
“AI for X” verticals are surprisingly narrow. Despite the hype around “AI for X” (e.g. AI for dentists), the only vertical AI categories that made it into the data are legal and patent-focused (e.g. Legora, Solve).
Platform/API-first success. 50% of successful companies are explicitly building platforms or APIs. This suggests YC’s breakout companies do lean into developer adoption and network effects for success and growth, and successful companies are not building one-off products.
There are notable absences from Series A success:
Zero LLM evaluation, observability, or tooling companies in the Series-A data.
Zero consumer, hardware, or deep tech companies in the Series-A data.
Top-tier lead investors matter. Most of the Seed rounds include top-tier lead investors — First Round Capital, General Catalyst, Uncork, Crat, Index, Soma, Greylock, Benchmark. Obviously, if one of these funds gives you a term sheet, it’s a no-brainer. However, we see party rounds at $2m on $20m (or $2m on $30m) with $200K left all the time pre-demo day, and this data suggests that having a reputable fund as a lead can be a better long-term setup for success.
Along with this analysis, I built a simple website to interact with the companies and get a better visual of the categories.
Check it out here:
https://yc-market-map-dot-chapter-one-340115.uc.r.appspot.com/
As mentioned above, the batches haven’t had much time to mature. Here is a table with the topline raise metrics:
Meta Is Creating a New A.I. Lab to Pursue ‘Superintelligence’
Nytimes • June 10, 2025
Technology•AI•Superintelligence•Investment•Regulation
Meta Platforms is embarking on an ambitious initiative to develop a superintelligent artificial intelligence (AI) system, aiming to create an AI platform that surpasses human cognitive abilities. This effort is spearheaded by CEO Mark Zuckerberg, who is personally assembling a team dedicated to this groundbreaking project.
To bolster this endeavor, Meta is aggressively recruiting top AI talent, offering compensation packages ranging from seven to nine figures. A notable recruit is Alexandr Wang, founder of Scale AI, who is expected to join the initiative. This strategic move underscores Meta's commitment to advancing AI technology and competing with other tech giants in the race for AI supremacy. (axios.com)
In addition to talent acquisition, Meta plans to invest approximately $15 billion to acquire a 49% stake in Scale AI, valuing the data-labeling startup at roughly $28 billion. This significant investment aims to enhance Meta's AI capabilities and position the company more competitively against rivals like OpenAI, Google, and Anthropic, all of whom are developing advanced reasoning AI models. (ft.com)
However, this substantial investment has attracted attention from U.S. regulators, raising potential antitrust concerns. The Federal Trade Commission (FTC) and the Department of Justice (DOJ) have the authority to investigate such minority investments, and Meta's deal structure may be crafted to avoid triggering antitrust scrutiny. (axios.com)
Meta's aggressive approach to AI development reflects a broader trend in the tech industry, where companies are making significant investments to secure leading positions in the rapidly evolving field of artificial intelligence.
Voice AI in a Box: The Future Is Talking Back
Medium • Astasia Myers • June 9, 2025
Technology•AI•VoiceAI•SpeechRecognition•ConversationalAI
By Astasia Myers and Dan Bartus
VC Astasia Myers’ perspectives on AI, cloud infrastructure, developer tools, open source, and security.
We’re excited for AI to expand the <$1T software TAM into the massive >$10T global services market. But to do that, AI needs to match more human capabilities, like answering calls, talking, negotiating, scheduling, and more. For AI to truly fulfill its TAM-expanding promise, it must be able to listen, understand, and speak more like a human.
The good news is it’s starting to happen. App builders are piecing together AI components to deliver a better conversational experience. The main components include:
Speech-to-text: Transcription platforms like Fireflies and Granola use models from Assembly and Deepgram to capture human voice accurately and turn it into text.
Text-to-text reasoning: Companies like OpenAI, Google, and Anthropic have made significant strides in conversational intelligence, and have made models capable of deep thinking.
Text-to-speech: Players like ElevenLabs and Cartesia have reached substantial valuations, driven largely by rapid adoption of their high-quality voice synthesis tools. We’ve also seen a vanguard of open-source text-to-speech projects like ChatTTS, Dia, and OpenVoice v2 that have offered developers incredible flexibility. Google recently announced native speech generation and live audio generation, which makes it possible for a model to sound even more natural (including whispering).
While individual models are promising, there’s still friction. Most developers don’t want to cobble together multiple models. Add in components like phone number provisioning, multilingual support, translation, and more, and the prevailing voice AI stack can be a messy web of tools.
This creates two big opportunities in voice AI: true speech-to-speech models, and “Voice in a Box.”
True Speech-To-Speech Models
Voice agents that lack prosody (that is, tone, timing, inflection, and pauses) feel one-size-fits-all, limiting their ability to deliver truly personalized experiences. While today’s models are increasingly accurate with words, they often miss the intent behind them. Prosody and emotion are the missing layers in voice AI. They’re what turn correct AI responses into “human” ones.
Without prosody modeling, voice agents struggle to interpret context and emotional nuance. With it, they can better understand the user’s emotional state and adjust their response accordingly; for example, speaking more quickly if the user sounds frustrated, or more calmly if they seem anxious.
Capturing and preserving prosody and emotion across languages remains one of the hardest unsolved challenges in voice AI. The next major breakthrough will come from large-scale, emotion- and prosody-labeled audio datasets across languages, enabling models to genuinely understand and convey tone, mood, and feeling.
We expect we’ll have voice models that natively understand and generate speech, language, and prosody in one loop. End-to-end models can preserve intonation, cadence, and timing better. Multimodal fusion lets the model interpret not only what is said, but also how and why. These models will generate speech directly with emotion, rather than stitching it on afterward. A fused model could do it all in one pass while reducing latency, which is crucial for real-time interaction.
Today, many conversational AI applications rely on modular three-step architectures that are controllable, auditable, and compliant for enterprise use. They allow for validation at each stage and are easier to debug. While multimodal fusion models will be a black box that will make it harder to isolate errors or easily verify outputs, they will also bring forth the future of more emotional and human-like experiences. Multimodal speech-to-speech models are when AI starts sounding (and acting) human.
Voice in a Box
Whether we move toward a multi-model stack or an end-to-end speech-to-speech future, we believe there will be value in the “Voice in a Box” layer — the packaging and API infrastructure that makes voice AI easy to deploy.
The key reason is that building effective voice AI applications requires much more than just a model. The real complexity begins after deployment: managing phone numbers, scaling calls, handling translation, enabling custom-branded voices, voice cloning, and more. These operational needs make a streamlined, developer-friendly voice layer essential. As models continue to rapidly outpace one another in performance, this packaging layer becomes the stable, enduring part of the stack.
There is also more to it than just packaging components. The real progress comes from improving latency and accuracy through fine-tuning, prompting, and feedback loops. These techniques are required to truly solve for areas where voice AI still struggles, such as:
Differentiating between multiple speakers
Appropriate tone and intonation
Management of crosstalk and cutoffs
Memory/context handling across calls
Correcting for background noise
Better understanding of unique business context
The massive voice AI opportunity exists in packaging and optimizing the voice stack into a streamlined, developer-friendly platform. Just as Twilio and Stripe turned complex communication and payments infrastructure into simple APIs, new startups like Vapi, Retell, and Bland are doing the same for voice AI. By packaging all of the necessary voice AI components into more developer-friendly APIs, they’re increasingly looking like a new “Twilio of voice AI.”
In the near term, we’re seeing insatiable demand for these developer kits to power high-growth AI apps that span customer support, healthcare, sales, accounts receivable, scheduling, debt collection, and more. The scope is massive. Longer-term, there is also a clear path laid by Twilio to follow: start with developer kits that abstract away communication complexity, and gradually move up the stack to offer more of the application layer.
Zooming out further, voice is one of the primary ways humans interact with each other, and will ultimately interact with AI. This spans both consumer and enterprise applications. Looking ahead, we also see real-time avatars as a natural evolution of the AI access layer. But again, low-latency, high-fidelity voice AI is also key to making avatars work.
The massive near-term opportunity to simplify voice complexity with developer kits, paired with the longer-term potential to move up or down the stack and power emerging technologies like avatars, signals that we’re at a true inflection point in voice AI. Taken together, it’s one of the most exciting and underexplored frontiers in AI infrastructure and is ripe for founders to build what’s next.
Thanks to Daniel Cahn of Slingshot for his feedback on a draft of this post.
Apple brings ChatGPT and other AI models to Xcode
Techcrunch • Ivan Mehta • June 9, 2025
Technology•AI•MachineLearning•Xcode•OpenAI
At the 2025 Worldwide Developers Conference (WWDC), Apple unveiled a new version of Xcode, its integrated development environment (IDE), featuring integration with OpenAI's ChatGPT. This enhancement aims to assist developers in tasks such as coding, documentation generation, and more. Additionally, Apple introduced the Foundation Models framework, enabling developers to incorporate AI models from various providers into Xcode, thereby offering AI-powered programming suggestions.
The integration of ChatGPT into Xcode allows developers to generate code previews, iterate on designs, and address errors more efficiently. Developers can access these AI tools directly within Xcode without the need to create a separate account. For those with paid ChatGPT subscriptions, connecting their accounts can provide increased rate limits.
In addition to the ChatGPT integration, Apple introduced the Foundation Models framework. This framework allows developers to access Apple's on-device foundational AI models with minimal code, facilitating the creation of offline, privacy-focused applications.
Previously, Apple had partnered with Anthropic, a startup backed by Amazon, to develop an AI-powered coding platform. This collaboration aimed to enhance Xcode by integrating Anthropic's Claude Sonnet AI model, enabling it to write, edit, and test code on behalf of programmers. However, the ChatGPT integration appears to be Apple's current focus for AI enhancements in Xcode. (reuters.com)
These developments signify Apple's commitment to integrating advanced AI capabilities into its development tools, aiming to streamline the coding process and enhance developer productivity.
Meta plans to invest $15bn in Scale AI in bid to catch up to rivals
Ft • June 10, 2025
Business•Investment•ArtificialIntelligence•Meta•ScaleAI•AI
Meta Platforms is set to invest approximately $15 billion to acquire a 49% stake in Scale AI, valuing the data-labeling startup at around $28 billion. This strategic move aims to bolster Meta's new "superintelligence" lab, enhancing its capabilities in developing advanced reasoning AI models to compete with industry leaders like OpenAI, Google, and Anthropic. (ft.com)
Scale AI specializes in preparing high-quality data and employs "reinforcement learning from human feedback" (RLHF), where humans help fine-tune AI models. The acquisition also includes bringing Scale AI's founder, Alexandr Wang, into Meta's new lab, potentially leading a division focused on integrating advanced human reasoning into AI systems. (ft.com)
This investment follows Meta's significant commitment to AI infrastructure, with plans to invest up to $65 billion in AI projects by 2025. CEO Mark Zuckerberg announced plans to build a data center large enough to cover a significant area of Manhattan, aiming to achieve approximately 1 gigawatt of online computing capacity by 2025 and expecting to have over 1.3 million graphics processing units (GPUs) by the end of this year. (gurufocus.com)
Meta's focus on human-assisted AI development reflects a shift from traditional reliance on computational resources and data. As freely available data becomes limited—a phenomenon likened to "peak data"—companies like Meta are exploring new methods to sustain AI progress. (ft.com)
The acquisition also casts uncertainty over Scale AI's IPO aspirations. Meta’s move follows similar strategies by other tech giants such as Microsoft's $650 million deal with Inflection and Google’s $2.7 billion arrangement with Character AI. These bespoke deal structures are reportedly designed to avoid regulatory scrutiny, though companies like Google and Microsoft have still faced antitrust examinations. (ft.com)
Vibe Coding, Windsurf and Anthropic, ChatGPT Connectors
Stratechery • Ben Thompson • June 9, 2025
Technology•AI•Software•Development•Innovation
Vibe coding is an emerging approach in software development that leverages artificial intelligence (AI) to streamline the coding process, allowing developers to focus more on conceptualization and less on the intricacies of syntax. Coined by OpenAI co-founder Andrej Karpathy, the term describes a workflow where developers interact with AI models using natural language prompts to generate code. As Karpathy aptly put it, "It’s not really coding — I just see stuff, say stuff, run stuff, and copy-paste stuff, and it mostly works." (wisprflow.ai)
This paradigm shift is facilitated by advanced large language models (LLMs) developed by organizations like OpenAI and Anthropic. These models are trained on extensive datasets of code and text, enabling them to comprehend human intent and produce functional software. The integration of LLMs into development tools has led to the creation of AI-powered integrated development environments (IDEs) such as Windsurf. Windsurf combines a modern IDE with Cascade, an intelligent AI coding assistant, to assist developers in writing code more efficiently. (anthropic.com)
Another significant advancement in this space is Wispr Flow, an AI voice tool that offers fast, accurate speech-to-text conversion, enabling hands-free interactions with development environments. When integrated with tools like Windsurf, Wispr Flow allows developers to dictate code, commands, and documentation, enhancing productivity and maintaining a seamless workflow. (wisprflow.ai)
The adoption of vibe coding is not limited to individual developers; it has also been transformative for teams and organizations. For instance, Codeium, a leading AI coding platform, utilizes Claude to power Windsurf and Cascade, reaching hundreds of thousands of users within weeks of launch and processing over 100 million tokens per minute. This rapid adoption underscores the growing reliance on AI-assisted development tools to accelerate the software creation process. (anthropic.com)
However, it's important to distinguish vibe coding from other forms of AI-assisted programming. Vibe coding specifically refers to the practice of building software with an LLM without reviewing the code it generates, embracing a more intuitive and less hands-on approach to development. This method contrasts with traditional software engineering practices, which emphasize thorough code review and understanding. (simonwillison.net)
In summary, vibe coding represents a significant evolution in software development, leveraging AI to simplify and expedite the coding process. By enabling developers to interact with AI through natural language and voice commands, tools like Windsurf and Wispr Flow are reshaping the development landscape, making it more accessible and efficient.
OpenAI expects subscription revenue to nearly double to $10bn
Ft • June 9, 2025
Business•Technology•AI•OpenAI•RevenueGrowth
OpenAI's annual recurring revenue has surged to $10 billion, nearly doubling from $5.5 billion at the end of the previous year. This growth is driven by high demand for its AI tool, ChatGPT, which has reached over 500 million users. The company, though not yet profitable and not expecting to be until 2029, currently generates revenue through consumer subscriptions, approximately 3 million business and education clients, and API sales. (ft.com)
Other AI companies like Cursor and Anthropic are also seeing major revenue increases, though all remain lossmaking. Venture capital and strategic investments, such as a $40 billion fundraising effort led by SoftBank for OpenAI, continue to fuel this expansion. Despite these gains, overall adoption of AI tools may be slowing, particularly in the U.S. business sector. (ft.com)
OpenAI is diversifying its operations through acquisitions, including Jony Ive’s hardware start-up and possibly the code editing company Windsurf. It is also collaborating with the Trump administration on Stargate, a large-scale data center aimed at supporting future AI advancements. (ft.com)
Government Overreach
Judge Denies Apple’s Appeal; Ordered to Keep Allowing Link-Outs to the Web in the U.S. App Store
Theverge • John Gruber • June 5, 2025
Technology•Mobile•AppStore•LegalChallenges•EpicGames•Government Overreach
In April, a federal judge demanded that Apple begin allowing web links, cease restricting how links are formatted, and enable developers to offer external payment options without giving the company a cut of their revenue. Apple promptly appealed and requested that the order be put on hold until the legal proceedings were finished.
But an appeals court has now denied Apple’s emergency request to block the order. The court said it was “not persuaded” that blocking the order was appropriate after weighing Apple’s chances to succeed on appeal, whether Apple would be irreparably harmed, whether other parties would be hurt if the order is halted, and what supports the public interest.
The rejection bodes poorly for Apple’s chance of overturning the order, which stems from a lawsuit by Epic Games.
Here’s the denial from the United States Court of Appeals for the Ninth Circuit, nearly in its entirety, omitting only legal citations:
Apple’s Emergency Motion Under Circuit Rule 27-3 for a Partial Stay Pending Appeal is DENIED. In deciding whether to impose a stay, we consider: “(1) whether the stay applicant has made a strong showing that he is likely to succeed on the merits; (2) whether the applicant will be irreparably injured absent a stay; (3) whether issuance of the stay will substantially injure the other parties interested in the proceeding; and (4) where the public interest lies.” Apple “bears the burden of showing that the circumstances justify an exercise of [our] discretion.” After reviewing the relevant factors, we are not persuaded that a stay is appropriate.
The earliest this might get overturned, it seems to me, is the end of this year, but I get the feeling this injunction is here to stay.
Politics & Technology
The Democrats’ Problems Are Bigger Than You Think
Nytimes • David Brooks • June 5, 2025
Politics•DemocraticParty•Strategy•ElectoralPolitics•VoterEngagement
David Brooks, a New York Times columnist, has been critical of the Democratic Party's strategies in countering former President Donald Trump. He argues that the party's focus on Trump's personal flaws has been ineffective, as it reinforces his supporters' perception of him as a victim of elite attacks. Brooks suggests that Democrats should instead address the concerns of working-class voters and present a compelling vision for the future. (foxnews.com)
Brooks also criticizes the Democratic Party for supporting far-right candidates in Republican primaries, believing that this strategy could backfire and lead to a Republican takeover. He emphasizes the importance of promoting the best candidates in both parties to ensure a healthy democracy. (foxnews.com)
Furthermore, Brooks highlights the party's disconnect with working-class voters, noting that Democrats have lost ground with white voters without college degrees, which has cost them crucial Rust Belt states in the 2024 elections. He points out that the party's elite image and focus on issues like climate change and cancel culture have alienated many voters. (en.wikipedia.org)
The Bluesky bubble hurts liberals and their causes
Washingtonpost • Megan McArdle • June 8, 2025
Technology•SocialMedia•Decentralization•ContentModeration•PoliticalCommunication•Politics & Technology
Ever since Elon Musk bought Twitter, changed the social media site’s name to X and altered its moderation policies, progressives have been hunting for a substitute. To judge how their search is going, consider a recent item from Politico’s Playbook, which notes that “a number of prominent commentators, experts and groups” are pledging to post on other platforms before X.
Enter Bluesky, the high-profile Twitter alternative launched in part with funding from Twitter itself. It’s a decentralized social network, designed to give users more control over moderation and censorship by allowing them to choose their own moderation algorithms. In theory, it’s an appealing idea to those unhappy with the increasingly permissive and chaotic environment at X under Musk’s leadership.
In practice, though, Bluesky has proven to be a very niche product. Although it boasts some big names in tech and journalism, its user base remains tiny compared with those of mainstream platforms. Even among progressives, it’s becoming clear that Bluesky doesn’t quite satisfy the need for a robust, engaging, and politically friendly platform.
What’s more, because Bluesky is decentralization-driven, its content moderation policies vary widely depending on which servers or communities users join. This patchwork of moderation complicates the user experience and limits the platform’s ability to foster a unified community or consistent political discourse. For those seeking a vibrant, active substitute for Twitter, this inconsistency can be a dealbreaker.
Meanwhile, X’s enormous scale and network effects continue to make it the dominant platform for political dialogue, even as controversial policies and posts proliferate. The network effect—the fact that everyone important is on X—means most activists, journalists, and politicians still must use it to reach large audiences, regardless of its flaws.
In short, Bluesky’s bubble is helping neither liberals nor their causes. Its exclusivity and fragmentation serve to isolate voices rather than amplify them, while X remains the indispensable battleground for political communication, requiring difficult compromises from its users.
Interview of the Week
The Empire Strikes Back: Karen Hao on OpenAI as a Classic Colonial Power
Five Key Takeaways
1. OpenAI is a Modern Corporate Empire Hao argues OpenAI operates like the British East India Company—a private corporation wrapped in a "civilizing mission" that extracts resources globally while externalizing costs to vulnerable communities. The company's stated goal of "benefiting all humanity" serves as ideological cover for profit-driven expansion.
2. AI Development Didn't Have to Be This Destructive Before OpenAI's "scaling at all costs" approach, researchers were developing smaller, more efficient AI models using curated datasets. OpenAI deliberately chose quantity over quality, leading to massive computational requirements and environmental damage that could have been avoided.
3. The Climate and Social Costs Are Staggering McKinsey estimates global energy grids need to add 2-6 times California's annual consumption to support AI infrastructure expansion. This means retired coal plants staying online, new methane turbines in working-class communities, and data centers consuming public drinking water in drought-prone areas.
4. The Business Model May Be Unsustainable Despite raising $40 billion (Silicon Valley's largest private investment), OpenAI hasn't demonstrated how to monetize at that scale. Subscriptions don't cover operational costs, leading to considerations of thousand-dollar monthly fees or surveillance-based advertising models.
5. Resistance is Possible and Already Happening Communities worldwide are successfully pushing back—from Chilean residents stalling Google data centers for five years to artists suing over intellectual property theft. Hao argues collective action across AI's supply chain can force a shift toward more democratic, community-centered development.
Startup of the Week
Cursor’s Anysphere nabs $9.9B valuation, soars past $500M ARR
Techcrunch • Marina Temkin • June 5, 2025
Technology•AI•MachineLearning•Productivity•Startup•Startup of the Week
Anysphere, the developer behind the AI-powered coding assistant Cursor, has secured $900 million in funding, elevating its valuation to $9.9 billion. This funding round was led by Thrive Capital, with participation from Andreessen Horowitz, Accel, and DST Global. This marks Anysphere's third fundraising effort in less than a year, following a $100 million raise at a $2.5 billion valuation in late 2024.
Founded in 2022 by MIT alumni Sualeh Asif, Arvid Lunnemark, Aman Sanger, and Michael Truell, Anysphere has experienced rapid growth. By April 2025, the company reported annual recurring revenue (ARR) exceeding $500 million, a significant increase from $300 million in mid-April.
Cursor has gained popularity among developers for its ability to write nearly 1 billion lines of code daily using natural language commands, significantly boosting productivity. The tool is utilized by major tech companies, including Stripe, OpenAI, and Spotify. Despite competition from tools like GitHub Copilot, Cursor has carved out a substantial user base. (ft.com)
The surge in Anysphere's valuation reflects a broader trend of investors focusing on AI application developers amid a booming AI startup scene. In 2024, AI application startups raised $8.2 billion, more than doubling the previous year's total. (ft.com)
Facebook made social networks. @TechCrunch made startup networks. Twenty years
X • joshk • June 11, 2025
X•Post of the Week
How TechCrunch Created the Startup Networks We Live In Today
Key takeaway: Over the past two decades, TechCrunch has fundamentally shaped how entrepreneurs, investors, and the wider tech ecosystem connect and operate — serving as the definitive startup network that parallels how Facebook revolutionized social networks.
Context & Reflection: Josh K recounts receiving an early email from Mike Arrington, founder of TechCrunch, around twenty years ago. From that moment, TechCrunch became an indispensable and constant presence in his browser — a testament to its enduring influence.
He draws a powerful analogy:
Facebook made social networks.
TechCrunch made startup networks.
This highlights how TechCrunch not only reported on startups but also actively connected entrepreneurs, investors, and enthusiasts, fostering a vibrant and interactive tech ecosystem around the globe.
Overall Impact: JoshK emphasizes that it is difficult to overstate TechCrunch’s influence in shaping startup culture, investment trends, and networking opportunities. It remains an open tab for many in the tech ecosystem, symbolizing its role as a vital, enduring resource for innovation and community building.
A reminder for new readers. Each week, That Was The Week, includes a collection of selected essays on critical issues in tech, startups, and venture capital.
I choose the articles based on their interest to me. The selections often include viewpoints I can't entirely agree with. I include them if they make me think or add to my knowledge. Click on the headline, the contents section link, or the ‘Read More’ link at the bottom of each piece to go to the original.
I express my point of view in the editorial and the weekly video.
Share this post