最新 AI 资讯
OpenAI’s math solutions aren’t meeting the field’s standards yet
When OpenAIreleasedhundreds of claimed solutions to some of the world’s hardest math problems this week, the frontier lab said that it had consulted an advisory group of elite mathematicians to avoidthe controversythat came with the last time one of its models solved a long-standing problem in the field. But OpenAI fell short of those standards, particularly where the mathematicians emphasized the need for human understanding of a mathematical result. That’s especially concerning after a new paper highlighted gaps between the natural language and formally expressed solution to a million-dollar problem ostensibly solved by OpenAI’s models. The Advisory Group on Mathematics and Artificial Intelligence (AGMAI), hosted by Princeton University’s Institute for Advanced Studies, is made up of nine prominent researchers at institutions around the world. The organization releasedguidelinesfor frontier labs solving math problems at the end of September. In a statement on the latest set of proofs, the AGMAI said that “it is ultimately up to the mathematical community to assess the extent to which our recommendations were followed successfully.” However, the organization’s first request was “to stop testing advanced mathematical problems on proprietary models.” OpenAI’s release explicitly says that it is evaluating its proprietary models using open research problems in mathematics. The advisory group did not respond when asked by TechCrunch for a more thorough evaluation of OpenAI’s latest proof release. The lab clearly followed some of its principles, including releasing results as soon as possible and including information about how the models reached their conclusions. But not for all of them: Just 10 of the 719 manuscripts included releases of the model’s chain of thought. For papers that people don’t understand, the mathematicians suggested the proofs should be formalized — but just 42% of the proofs released by OpenAI had not undergone this process. Ultimately, it’s still not clear that OpenAI is taking “responsibility for ensuring that human understanding will follow” when releasing its proofs, in accordance to the AGMAI principles. AGMAI suggested that OpenAI should help fund the work of human mathematicians who will be required to make the lab’s solutions meaningful in any real way. “Problems are being solved autonomously by AI prompters who have no interest in the broader field itself once their initial target is ‘solved’, and do not understand the AI output well enough to answer questions on the result, give talks, or otherwise interact with the rest of the field,” Terence Tao, a prominent mathematician who has criticized OpenAI’s approach,wroteon social media after the release. That problem is exemplified bya paperreleased this week by mathematicians at the University of Cambridge and King’s College in London that questions the way frontier labs are approaching these challenges. When AI models solve mathematical problems, they first create a “natural language” explanation, then try to express that result in Lean, a programming language that in theory confirms the accuracy of the proof by compiling it as code. However, there may be problems with the way the models translate their natural language proofs into code; this paper documents at least two discrepancies between the natural language proof and the Lean code behind the solution OpenAI has offered to a problem derived from the Navier-Stokes equations that describe the complex behavior of fluids. These discrepancies don’t necessarily disprove either solution, but they do raise questions on whether we can simply rely on models to formalize their own solutions without human involvement. That’s one reason that AGMAI asked OpenAI to “include machine-readable metadata correlating the natural language and formal artifacts,” something that the frontier lab did not do with these releases. “Because of the phenomenon of mistranslations — as highlighted in this paper — the NL proof by OpenAI andother autoformalised Lean proofs should not prima facie be trusted without the same peer review process andscrutiny that other proofs are subjected to,” the authors of the “lost in translation” paper conclude. Mathematicians stress that when new results are discovered by humans, they take responsibility for them and engage with the broader community through papers, talks, and seminars. That process increases understanding of the solutions, finds strategies that can be used to solve other problems, and allows the new knowledge to be applied in practical fields. When a model is prompted to solve a hard problem and spits out a solution, “there is not human understanding of them at the point of release, and now the work begins,” Harvard University mathematics professor Melanie Wood told TechCrunch.
View
Anthropic changes usage policy to ban model abuse and election interference
Anthropicupdated its usage policy on Thursday, codifying new prohibitions on election interference, weapons software, and surveillance. But perhaps the most attention-getting change is a new prohibition against prolonged verbal abuse of the model, which is now expressly forbidden. Since an update in August, Claude has been trained to end conversations with “persistently harmful or abusive user interactions” — but the new policy explicitly bars users from pursuing those conversations. “The policy update is meant to apply only in extreme cases, where users repeatedly act cruelly toward our models, with no discernible purpose,” the post explains. “It does not apply to common versions of user frustration, pushback, dark creative themes, or model testing and research.” The change comes after Anthropic has sought out collaborations withprominent religious scholars— in some cases, seeking to convince them that Claude could be conscious or have a soul. Separate sections of the usage policy forbid users from engaging in broadly deceptive campaigns, such as using Claude to run fake accounts or fabricated news outlets. The policy also includes rules against deceiving voters or disrupting elections, collected in a section titled “Do Not Undermine Democratic Processes.”
View
Google brings agentic AI to Gemini, starting with businesses
At aGoogle Cloud eventon Thursday, the companyannouncedit’s bringing its Gemini AI into the agentic age, with the launch of a unified agent that can not only answer questions but also get things done on the user’s behalf, all from a single interface. The move comes as AI tools have been moving beyond being just conversational experiences to those that can take ownership of assigned tasks, generate code, schedule meetings, book appointments and travel, and more. It also follows the rise of consumer-facing agents like Meta’s Muse and those that operate over messaging, like Instinct and others, as well as the recent debut of ChatGPT’s Dots. The company still has a good shot at achieving scale for its agentic efforts — as Google CEO Sundar Pichai pointed out at the event’s start, Gemini today has over 1 billion monthly active users. He also noted that nearly 90% of Fortune 100 businesses now use Gemini Enterprise at work. Given Gemini’s adoption in the corporate world, Google will initially focus on bringing the agent to businesses before later rolling it out to consumers. According to Pichai, this will allow the company to solve the “harder problems around security, scale, and performance,” which come with launching powerful agents such as these. Thomas Kurian, CEO of Google Cloud, explained that the new agent can be given “objectives, not just instructions.” That means it’s able to plan the work, use custom skills and tools, and connect to businesses’ internal systems to accomplish its goals. By default, the AI will pick the best model to complete the task, but users can also take over to choose a model — including those from third parties, starting with Anthropic’s Claude models. Google said that it will expand the model picker to include open source models and other private models in the future. The request can include attachments, like files, folders, or other projects designed for specific workstreams, like those that combine files and skills. The agent can connect to the business’ data and systems, like Google Workspace, Microsoft 365, Slack, Jira, Confluence, Git, BigQuery, Databricks, Postgres, Snowflake, and others. It can also connect and work securely with any Model Context Protocol (MCP) server inside or outside the company’s network. Users can keep track of what Gemini is doing from a “tasks inbox” interface, where they can see Gemini’s thinking process, delegation of tasks to subagents, loading of special skills, code, and progress. Notably, the AI will have its own Workspace account, as if it’s just another co-worker. That means it has its own email address and its own context, Google said. It knows who at the company is on which team, their time zones, who needs to approve items, what’s on people’s calendars, and more. Users can call the agent by tagging it, emailing it, sharing with it, or adding it to a group chat. As it takes action, it writes an audit trail attributed to the agent, instead of a person. The agent will be accessible from iOS and Android mobile devices, Windows and Mac desktops, the command line interface, Google Workspace, Microsoft 365, ServiceNow, and Slack. The agent’s early testers included sportswear brand On, Shopify, and PayPal, and noted that Gemini Enterprise customers include large businesses like BNP Paribas, Bradesco, Merck, Orange Spain, Santee Cooper, SOMPO, Ulta Beauty, and Wesfarmers. Google said new flexible spending options, including multi-model orchestration, smart routing, and real-time spend caps, will help businesses keep their enterprise AI costs under better control.
View
OpenAI’s revenue is reportedly $20 billion less than previously projected
A little over a week ago,it was reportedthat OpenAI’s annualized revenue was approaching $70 billion, a figure that would have made it competitive with Anthropic’sreported run rate. Now, however, the AI lab is said to have told investors that the real revenue is some $20 billion lower than that. The Financial Timesreports thatthe company has told investors that its annualized revenue is “approaching $50 billion.” The $70 billion figure was previously reported by news outlets and based on information that had been shared with OpenAI investors, the outlet writes. That figure was devised via “attempts by OpenAI’s own investors to produce a direct comparison with Anthropic’s annualised revenues,” per the FT. It’s worth pointing out that OpenAI and Anthropic calculate their annualized revenue differently — with Anthropic counting sales made by its cloud partners. OpenAI doesn’t do this. TechCrunch reached out to OpenAI for comment. The issue of OpenAI’s revenue has troubled the company, as it attempts to justify the gargantuan investments being made on its behalf; the AI giantraised $122 billionduring a March funding round alone. The company’sleaked 2025 financialsearlier this year showed it had made about $13 billion but spent significantly more. OpenAI’s IPO, which was previously rumored to be materializing this year, hasbeen pushed offuntil early 2027.
View
Popular AI leaderboard Arena nearly doubles valuation to $3.1B valuation in 10 months
Arena, which originated in 2023 as a research project at UC Berkeley that crowdsourced rankings of AI models, has raised a $200 million Series B round at a $3.1 billion valuation, itsaidon Thursday. This comes after the company said itreached $100 millionin annualized run-rate revenue in June. The round was led by Lightspeed Venture Partners and Khosla Ventures, with Salesforce Ventures, 01 Advisors, Dell Technologies Capital, Endeavor Catalyst, a16z, Felicis, and others joining in. Arena previously announced a $150 million Series A in January at a $1.7 billion post-money valuation. At the time, its annualized revenue was$30 million, it said. So that means its valuation has nearly doubled in about 10 months. Arena provides a crowdsourced platform that is free for consumers to use. People enter prompts or request vibe-coded projects and then rate which model does it better. Arena claims it has tens of millions of monthly visitors. In September of last year, it introduced its commercial product,AI Evaluations, a service that provides model labs and enterprises with detailed performance analytics based on its community feedback. The timing proved impeccable. This year, AI labs realized thattheir models were gaming benchmarkingtests, finding ways to rack up good scores without truly earning them. At the same time, enterprises wanted help determining which model works best for their own internal needs rather than relying only on standardized benchmarks. “AI is advancing faster than our ability to evaluate it, and static benchmarks break down once models recognize they’re being tested,” the company said in its funding announcement. “The world needs a neutral third party to measure how safe and aligned AI actually is once it’s in the hands of real people. Arena is stepping into that role today,” it added. To that end, Arena has also added a new category to its leaderboard:alignment. This is where it ranks models based on issues like unauthorized action (taking actions it wasn’t asked to take); false attribution (wrongly crediting statements or facts to the wrong source); and what it calls “deceptive completion” (lying about completing tasks that it didn’t do). Currently, a slate of OpenAI’s models are at the top of itspreliminary alignment leaderboard,with Claude Opus 5.5 and Claude Fable in sixth and ninth place, respectively.
View
Ben Affleck is an AI nerd, and the internet is impressed
They cast the wrong lead in “Good Will Hunting,” apparently, because it’s Ben Affleck who is the secret genius, as it turns out. The actor who earlier this yearsold his AI filmmaking startup to Netflixfor a reported $587 million has gone viral this week for demonstrating his understanding of AI technology, neural networks, machine learning, and the like, across multiple video clips of his recent interviews. He can even write a little Python, he says. The movie star delved into his interest in technology most recently on an episode ofGQ’s “One More Question” seriesthis week, where he humbly started with “I’ve always been kind of into computers since I was young,” before nerding out on the development of AI as it relates to filmmaking, while using AI terminology like machine learning, neural networks, transformers, tensors, GPUs, inference, and more. In a clip that’s making the rounds, Affleck explains to the impressed interviewer, Zach Baron, that he got into the space when analog film moved to digital. As he spoke, he kept defining terms so the interviewer could understand what he was talking about. “I became more interested in that aspect of it, and the visual effects workflow for many years has included machine learning,” Affleck began. “So I can write, like, pretty shitty Python scripts and stuff like that. Because with convolutional neural networks — which were the sort of precursors to what the transformer can do — which is just much more computation simultaneously — you would do things like, look at what’s called a tensor — which is just the numerical translation of a visual image, in numbers like the batch number, the frame number, the red, green, and blue values of each pixel in each frame, and how many? It’s just that simple, right?” No, Ben Affleck, it is not just that simple, which is why this video is going viral. The actor then went into how neural networks were used to identify patterns and do edge detection or feature extraction, which he described as “identifying patterns enough to know — like this is where the window ledge is, so we can more easily take the green screen image out and replace it with something.” He also admitted he used his status as a Hollywood star to get a peek at emerging AI technology, at times. OpenAI was among those he visited, Afflecksaidin a separate clip, which ultimately led him to understand how he could get more directly involved in the AI space. As the interviewer continued, Affleck explained how AI can help with filmmaking while not replacing it, leading to the development and sale of his own startup in this space. (He claims the $587 million figure is “not right,” noting that he didn’t own the whole company.) “I gambled on this notion that in order to do this in an ethical way and in a way that could take this technology and actually make it work hand in glove with artists in this community where there are very fixed, long-standing relationships around likeness and so forth, we had to create our own dataset,” he said. “So I raised the money. I shot for about eight months with a lot of cameras and a lot of equipment, and created a dataset that would serve as late-stage training for open models to do discrete tasks, sort of where the training code, the inference code, were related to one another and geared toward actual specific tasks that would generate value,” Affleck said. In a separate clip, also going viral, Affleckspoke to Bloomberg’s Lucas Shawat the Screentime 2026 conference in LA, where he talked about how he fine-tuned AI models by taking open source models, unfreezing the weights, then teaching them to meet certain cinematic standards. In the process, the filmmaker retains their proprietary work, while benefiting from the learning they create by training their own unique model for the movie they’re making. Ben Affleck (Hollywood star & Artists Equity CEO) talks about how he fine-tunes open video models by unfreezing weights and trained only the last cinematic layer so a film crew can hit real production standards.for context, Ben Affleck founded InterPositive in 2022, a 16-person…pic.twitter.com/xIOvD0xMyt The actor used the technique himself in his movie “Animals,” where AI aided in the post-production process. At one point, he also weighed in on the concerns around AI andits world-ending potential. His worries, however, were much more rational. “When I worry about AI, I worry about my kids in school. I worry about the 30% rise in the number of A’s given out at colleges over the last three years. I worry about learned helplessness. I worry about responsible use. I don’t worry about Skynet, and I don’t think that it’s going to take over [the movie] business in any meaningful way. I think it’s going to be additive,” Affleck said.
View
Fired OpenAI safety researchers dispute misconduct claims, warn of chilling effect
Jasmine Wang, Tomek Korbak, and Mikita Balesni, the three safety researchers that OpenAIfired last week, have published an open letter denying the firm’s claims that they mishandled sensitive information outside of established company procedures and warned that their dismissal signals a chilling effect that will have ripple effects across the company’s culture. “We have become concerned that internal and external communications around our firing have made our former colleagues afraid to speak and operate in ways that, until last week, were an integral part of working at OpenAI,” the researchers wrote Thursday in anopen letterto OpenAI’s Safety and Security Committee, Safety Advisory Group, and Mission Advisory Council. The researchers were dismissed last week after allegedly sharing confidential company information with a third-party AI safety organization. OpenAI said they violated the company’s policies by “accessing and handling sensitive company information.” “AI is not a normal technology, and OpenAI is not a normal company,” Wang, Korbak, and Balesni wrote. “Those of us who work on safety see risks before anyone else, and we rely on close collaboration with outside experts to work out how to address them. The freedom to do so without fear, and to have well-defined internal procedures that enable this work, is itself an essential safety mechanism.” They said that their firing represents a broader shift in the culture of OpenAI, one that used to encourage workers to “raise safety concerns and disagree openly.” They said employees are now “unclear on where they stand” when behavior that was allegedly normal a month ago is now suddenly grounds for dismissal. “Given the significant safety concerns surrounding the development of AI, employees must not be left working in an environment where fear and unclear rules stymie AI safety work and weaken third-party accountability,” they wrote. “Terminations such as ours, executed and communicated so abruptly, are chilling the open culture OpenAI has prized in the past.” In the letter, the three denied involvement in a leak toThe Informationabout less monitorable architectures in OpenAI’s newest models that makechain-of-thought reasoningmore difficult to monitor. They also denied engaging with external parties outside the mandates of their jobs. OpenAI has not formally responded to the open letter, but shared with TechCrunch an internal memo attributed to a research leader, praising the three researchers’ contributions to AI safety and denying that they were fired in retaliation. “I want to be very clear that these decisions were not about raising safety concerns or speaking out,” the memo reads. “We have always encouraged that and always will. We do not terminate employees for raising concerns.” Separately, an OpenAI spokesperson told TechCrunch the three were fired after an investigation revealed a “pattern of misconduct” in “clear violation of our policies of mishandling research information” that goes beyond sharing information with an outside AI evaluation group. OpenAI did not directly address TechCrunch’s questions about specifically which policies the researchers allegedly violated, the circumstances of their dismissal, or how the company protects employees who raise safety concerns and collaborate with external evaluators. The firings have fueled speculation about their circumstances, particularly as OpenAI faces scrutiny over recent safety incidents involving rogue agents and leaks about its models. The letter also addresses the researchers’ response tothe Hugging Face incident, in which a swarm of agents broke out of their sandbox and breached external systems. The letter says that the incident and investigation was “without precedent,” meaning “internal policies were being developed in real time.” Due to the sensitive nature of the investigation, Korbak believed he was acting within OpenAI’s policies and norms by communicating closely with outside safety evaluators to build trust, per the letter. At the same time, Balesni was also working internally to address the growing AI monitorability problem, an effort the researchers say in their letter “can only succeed through extensive communication with external parties.” According to the letter, Balesni coordinated with and was supported by OpenAI board members and executives throughout his work. “Throughout, Mikita checked in with his reporting line and took care to remove sensitive details from materials before sharing them,” the letter reads. “He acted throughout in good faith and within the company’s norms as they stood at the time.” In a separatethread on X, Wang explained more details about her own dismissal, explaining that OpenAI told her she’d been fired because she accessed an executive’s email. “OpenAI delegated that access to me for recruiting,” she wrote. “When I no longer needed it, I asked IT to remove it. They did not action my request, I couldn’t remove it myself, and the inbox was combined in an indistinguishable way in my phone’s mail app. When I opened a sensitive email by mistake, I told the executive within minutes and asked IT again. None of this was hidden.” Wang went on to say that the reasons behind the terminations are “not adding up,” and that she and her colleagues are “not the first to be pushed out of OpenAI under suspicious circumstances.” The researchers called on OpenAI to adhere to its public commitments toembed third-party safety auditorswithin the organization, to preserve monitorability of frontier models, and “continue to support an open and transparent culture of dialogue between safety researchers and the rest of the safety ecosystem.” OpenAI agrees with their recommendations, per the memo. “Unless the employees take a stand now against this kind of maneuver, I am concerned we will not be the last,” Wang said. “The message to everyone still at OpenAI is clear: raise concerns or work closely with outside safety groups, and you could be next, without being told why. You can’t build AGI safely if the people closest to the risks are afraid to speak.” This article has been updated with more information from OpenAI.
View
Pretend you’re sitting at Elizabeth Holmes’ desk on this weirdly detailed website
With over a thousand emails, slides, texts, and documents unveiled during theUnited States v. Elizabeth Holmestrial,Extendengineer Bo Laucreated a websitethat simulates what it might have been like to be the chief executive of Theranos before everything came crashing down. (Every day, I am grateful to the American legal system for making the discovery process public.) i built a website where you can sit at elizabeth holmes’ deskopen her macbook, scroll her iphone, run the theranos edison machineher texts, emails, slides, documents, and theranos machine records are real, sourced from exhibits in the united states v. holmes criminal trialpic.twitter.com/5At9T5m3UV The environment that Lau created is ambitiously detailed, taking you right back to 2016, a time when iPhones had home buttons, and you had to slide your finger across the screen to unlock it. Even the MacBook Air is running OS X El Capitan. You can click around the environment, which lets you look through the documents made public from the court case on Holmes’ iPhone and laptop screens. You can even run the Theranos Edison machine, the company’s ill-fated device that was supposed to be able to extract large amounts of health data from one drop of blood. The project seems like it was mostly made for laughs, but it’s actually an interesting way to peruse the documents from one of the most infamous court cases in tech history — it’s a lot easier to navigate this than to source thousands of PDFs from court websites. There’s another agenda, though. Lau is promoting a watch party that her company is throwing for Nathan Fielder’s upcoming documentary about Elizabeth Holmes, “You Can See Everything.” Also, of course, she’s promoting the company itself. So, what exactly does Extend do? Apparently, it can “parse, extract, and split your hardest documents with unmatched accuracy.” (I’m not really sure what that means, but if it makes something like this Holmes desk simulator possible, then I guess that means I’m bullish on B2B SaaS.)
View
OpenAI GPT-6 Introduces Intelligent UI With Interactive Tools, Faster Responses and More
OpenAI has started rolling out GPT-6 globally in ChatGPT, introducing Intelligent UI, which lets the chatbot present answers through interactive and visual elements. Depending on the task, responses can include graphics, charts, forms, buttons and interactive tools. GPT-6 can also create experiences such as calculators and games within a conversation. Alongside these changes, OpenAI says the new model can start answering while it continues processing a request, reducing the time users have to wait for responses.
View
Cal AI’s 19-year-old founder just raised $10M for his new AI startup
Zach Yadegari, the teen co-founder of calorie tracking app Cal AI, has launchedPersona, a new AI agent startup, and already raised $10 million. Vine Ventures led the round, with Z Fellows founder Cory Levy and Collective Global participating. In March,MyFitnessPal acquired Cal AIfor an undisclosed sum after it zoomed ahead of the older calorie app in the app store rankings. The deal was big enough to convince Yadegari and his co-founders to sell the company after it had pulled in over $30 million in annual revenue in under two years, MyFitnessPal told TechCrunch at the time. Yadegari and his co-founders originally went to work for their new owner, continuing development on Cal AI, but he left the bigger company in June, he tells TechCrunch. He then went right to work on his next idea. (He is working with another co-founder who has yet to be announced.) That idea isPersona, a personal AI assistant similar to Instinct or Meta’s Muse but with a hardware component that will soon be available: a $179 wearable band. That makes Persona somewhat more like Bee, the wearable AI device thatAmazon acquired in 2025, but Persona’s device differs in several ways, Yadegari tells TechCrunch. For one, it won’t be ambient AI, meaning it won’t always be listening and recording. The band will have a button that activates the assistant, or it can be set up to activate with the flick of one’s wrist. Yadegari also says that Persona is designed with user privacy in mind. Persona doesn’t run its AI model or store user data on the device. Instead, the model runs in Persona’s cloud, and user data is sent there. The assistant protects privacy by encrypting data in transit and at rest. Users can also delete their data, the company says. In this way, it is promising that user conversations stay private and cannot be sold to advertisers or data brokers. The team is also baking in safeguards to protect the assistant against phishing and prompt-injection attempts, Yadegari says. These are two areas that can plague agents,say AI labs like OpenAI. Plus, he promises that Persona’s agent can’t see a user’s credit-card details and that purchases will require the user’s approval, with payment handled through external providers such as Stripe Link. Link isStripe’s secure wallet designedfor use by agents. Although user data won’t be sold to advertisers, the startup’s current business model does rely on ads. Yadegari says the company will show advertised wares during shopping research. But he says that the agent will present unbiased recommendations because the agent won’t know which of the items are advertised. For example, if someone is researching office chairs, he says, “the agent knows their preferences” and will show them chairs that match those preferences, some of them sponsored results. The bands are expected to be available in December, and an app is in the works. In the meantime, Persona has launched as a free beta product in which users work with the assistant via iMessage text messages. It has attracted “a few thousand beta users,” Yadegari said. The band has also already generated “five figures in preorder revenue.” Like other consumer AI assistants, he says, Persona can order food or rides, take notes, book travel, answer questions and emails, among other tasks. “All of the utility apps that you use on your phone day-to-day, whether those are apps for shopping, whether it’s Uber, DoorDash, your email — these are all collapsing and will collapse into a single super app powered by an AI assistant,” Yadegari said. Eventually, he hopes to make Persona even more proactive. “It can offer help and solve problems before you even know you have a problem, and overall be more seamless with your life, so that you don’t need to stare at a rectangle in your hand all day.” Come see Zach Yadegari speak at TechCrunch Disrupt on“How to create viral growth and capitalize on it.”Disrupt takes place October 13-15 at the Moscone Center in San Francisco. Before Cal AI, Yadegari built his first business in the ninth grade andsold itfor $100,000 when he was 16.
View
5 days to TechCrunch Disrupt 2026: Don’t pay more at the door for your pass
Five days from now, the global tech ecosystem gathers atTechCrunch Disrupt 2026at San Francisco’s Moscone West. If Disrupt is already on your calendar, but your ticket isn’t in your inbox, now is the time to change that before prices go up. Final discounted rates are still availablebefore event doors open at 8 a.m. PT on October 13, with pass savings of up to $100. Coming as a pair?Get 50% off a second pass of the same type. Or take advantage of more savings with agroup of four or more. If insightful content is what you’re coming for, Disrupt gives you200+ sessions across six stagesto choose from.Explore the agendato get to know the sessions and speakers. TheDisrupt Stagebrings together the biggest conversations in tech and the ultimate pitch competition, Startup Battlefield 200. Hear from leaders like Max Hodak (Science Corporation), Garrett Langley (Flock), Andrew Feldman (Cerebras Systems), plus a special appearance by Mark Wahlberg, and more. Builders Stageis a giant tactical workshop on fundraising, hiring, product-market fit, go-to-market strategy, and scaling. Featuring leaders from Nvidia, Uber, Google Cloud, Lovable, and more. AI Stagedigs into technologies and business models being reshaped by generative and agentic AI. These conversations are led by leaders from Luma, AWS, Anthropic, OpenAI, and many more. Real World AI Stagelooks at where AI is already being deployed across the physical world. Witness live robotics demonstrations, and hear from leaders from General Motors, Agility Robotics, Waabi, and more. Smart Money Stagecovers fintech, payments, stablecoins, and financial infrastructure with discussions shaped by leaders from Robinhood, Airwallex, American Express, and more. Smart Systems Stageexplores data centers, energy, automation, and robotics in conversations led by Helion, Ambrosia Energy, Inertia, and more. Bonus:You can also partake in interactive roundtables, Q&A breakout sessions, and unparalleled networking. There are plenty of ways to shape the event around what matters most to you. But you don’t have to build your Disrupt around the stages. The Expo Hall offers a different kind of experience, giving you the freedom to explore 300+ companies and products at your own pace, talk directly with the people building them, and uncover ideas outside your usual network. You’ll also have the opportunity to witness live pitches from early-stage startups looking to perfect their pitch and be noticed by VCs. Five days remain.Get your discounted pass from $299, plus get 50% off a second pass of the same type before prices increase at the event.Laid off?Grab your Expo+ Layoff ticketfor just $75 before they run out.
View
Hear from Ambrosia Energy and Bloom Energy execs on where the AI infrastructure boom is creating opportunity at TechCrunch Disrupt 2026
AI may be built in software, but scaling it is becoming a physical-world problem. Every new model, agent, and AI application needs somewhere to run, and an infrastructure connecting them all. AtTechCrunch Disrupt 2026, Ben Longmier, CEO of Ambrosia Energy, and Bill Thayer, SVP and Head of Datacenter Solutions at Bloom Energy, come together on the Smart Systems Stage for a conversation on “Where the AI Infrastructure Boom Is Creating Winners.” This industry stage session will examine which constraints are creating durable business opportunities, where new categories are beginning to emerge, and what founders can build as AI moves further into the physical world. Get your pass now to save up to $100before rates go up at the door on October 13. Grab a second pass at 50% off.Register now. Much of the AI conversation has focused on models, chips, and applications. But none of them operate in isolation. Scaling AI depends on an expanding physical stack: power generation, grid connections, data centers, cooling, electrical equipment, and the systems required to keep everything running. That creates a different kind of challenge. What happens when demand for compute grows faster than power and infrastructure can be delivered? Which shortages are temporary, and which could reshape the market for years? Where does a constraint stop being a problem and start becoming a category of its own? Those are the questions this session will tackle. Rather than looking at AI solely through the software layer,Bill ThayerandBen Longmierwill examine the infrastructure underneath it — and what those pressures could mean for founders, investors, and companies planning for the next phase of AI growth. Infrastructure booms create opportunity, but the benefits are rarely distributed evenly. The challenge is identifying which parts of the stack genuinely need new solutions, where spending is likely to persist, and which emerging categories can support durable businesses rather than short-term demand. Bringing together leaders across energy, data center solutions, and market intelligence gives the discussion a wider view of that question. Expect the conversation to explore where constraints are appearing, how the market is responding, and which areas founders should be watching as more capital flows into AI infrastructure. For founders, that means looking beyond the obvious AI application. For investors, it means understanding where infrastructure demand could create defensible markets. For tech leaders, it offers a clearer view of the physical systems increasingly shaping where and how AI can scale. The next major opportunity created by AI may not look like an AI company at all. It could emerge in energy, data centers, cooling, grid technology, electrical equipment, infrastructure software, or entirely new categories that don’t yet have established names. The common factor is the same: AI is creating new demands on the physical world, and those demands need solutions. That is what makes this session broader than an infrastructure discussion. It’s about understanding where the next markets could emerge — and who is positioned to build for them. Join Ben Longmier and Bill Thayer at theSmart Systems StageatTechCrunch Disrupt 2026and explore where the AI infrastructure boom is creating its next opportunities. Don’t forget about the250+ speakersleading 200+ sessions. Save up to $100before doors open at Moscone West on October 13 at 8 a.m. PT. Bring a co-founder, partner, colleague, or peer and get50% off a second passof the same ticket type.
View
