与「PROJECT_HORIZON」相关的搜索结果

📢Meet Qwen3.8-Max — our most capable model to date. Next week, the open weights of Qwen3.8-Max will be released, and Qwen3.8-27B is also going open-weights to meet you all!🎉 Qwen3.8-Max, a new bar for coding and cowork at 2.4T parameters: - Autonomous coding: 10+ days of self-evolving development, from empty folder to production without hand-holding, complete project trace in the GitHub： - Real work, real results: Production-quality deliverables across hundreds of professions. - Long-horizon mastery: System-level autonomous planning with closed-loop adaptive learning, driving 500+ turns of chip design optimization and 365 days of e-commerce strategy. - Native multimodal intelligence: Vision isn't just input — it's a continuous feedback loop for planning, execution, and self-correction. 💰Pricing: Input: $2.0 / M tokens Output: $6.0 / M tokens Implicit Caching: $0.25 / M tokens Start building with Qwen3.8-Max! 🚀 📖 Blog: ✅ Qwen Studio: ⚡ API:

Gold explorer @Horizongold_ definitive study has outlined a huge $1.85B cashflow, $1.3B net present value blueprint for its WA Gum Creek gold project, producing 880,000 ounces of gold over a 10-year mine life. $HRN @theage

Feels bad, -49.4% drawdown this month after the recent crash. My portfolio is mainly AI chokepoints and bottlenecks. In the memory, photonics, robotics, and upstream semis, (on margin) which all tend to be higher beta than others. But reduced leverage recently from the crash. I see a lot of people making fun of the drop or AI names, saying it’s obvious that: - “AI is a bubble” - “memory/kospi is a bubble” - “photonics is a bubble” - “humanoids won’t get anywhere” - “neoclouds will get replaced by hyperscalers like Meta” And a bunch of retail + bots saying “sell everything, it’s never going to recover”. But I have conviction that all these themes are backed by structural revenue growth or technological shifts. And I’ve had similar drawdowns back when there admin threatened global tariffs, before markets pulled off a recovery. I personally have a longer horizon + higher tolerance for volatility than others, to see how this plays out. Especially considering a lot of retail view things on a week to week basis: no, my thesis isn’t wrong yet if I project revenue inflection in H2 2027 and it’s 2026 now. Anyway, feels bad short term just wanted to share anyway for transparency.

This month, Alaya Lab is releasing a series of research projects and open-source initiatives. Today, we're excited to introduce Alaya World - an open-source interactive video world model. Project Page: Github: ✨ Highlights: - 720p 24 FPS streaming generation - Navigation and prompt-driven interactions (e.g., spell casting and summoning) - Stable long-horizon generation (>1 minute) - State-of-the-art performance - Inference code is available today. Training code and datasets are coming soon.

SpaceX 15-Year Investment @elonmusk My evaluation of a long-term bullish investment for SpaceX based on a 15-year investment horizon. The evaluation assumes SpaceX evolves beyond aerospace into a global infrastructure platform spanning communications, defense, transportation, AI infrastructure, and the emerging space economy. Key Value Drivers - Starlink broadband growth
- Direct-to-cell mobile connectivity
- Government and defense contracts
- Starship and launch-cost reduction
- Orbital infrastructure
- Lunar and Mars logistics
- Future industries enabled by low-cost access to space An extremely important development is the impact of large AI infrastructure contracts on the valuation of SpaceX. A key component of the bullish investment thesis is that the market may be underestimating the value of recurring AI compute revenue when compared with traditional launch and satellite communications businesses. Google AI Infrastructure Contract Recent reports indicate that Google entered into a computing capacity agreement with SpaceX valued at approximately $920 million per month. If maintained over the reported contract period, the agreement represents nearly $30 billion of contracted revenue. Anthropic Contract In addition to Google, reports indicate that Anthropic entered into a compute-capacity agreement valued at approximately $1.25 billion per month. Together, the Google and Anthropic contracts imply annualized revenue of approximately $26 billion. Why This Matters Many investors continue to value SpaceX primarily as a launch company and satellite communications provider. However, these AI-related contracts suggest the emergence of a third major business segment: AI infrastructure and compute services. Potential Strategic Position If SpaceX successfully integrates launch services, Starlink communications, AI infrastructure, and future orbital computing capabilities, the company could evolve into a hybrid of a cloud infrastructure provider, global telecommunications company, defense contractor, and transportation platform. Investment Implications Recurring infrastructure revenue generally receives higher valuation multiples than project-based revenue streams. Consequently, sustained growth in AI infrastructure revenue could materially alter the market’s perception of SpaceX and support valuation frameworks significantly above those used for traditional aerospace companies. Bull Case SpaceX becomes the dominant communications and transportation infrastructure provider for Earth and near-Earth economic activity. Starlink achieves global scale, Starship reaches full operational capability, and multiple new industries emerge around orbital infrastructure. Base Case SpaceX remains the global leader in launch services and satellite communications while generating strong growth from Starlink and defense-related revenue streams. Bear Case Growth continues but at a slower pace due to regulatory, competitive, technical, or capital allocation challenges. Comparison with Historic Winners Historically, Amazon, NVIDIA, Apple, Microsoft, and Tesla created extraordinary shareholder wealth by becoming platform companies rather than remaining confined to their original industries. The bullish SpaceX thesis assumes a similar transition. Key Risks The principal risks include execution risk, regulatory risk, geopolitical factors, competition, slower-than-expected adoption of space-based industries, and the possibility that SpaceX enables large industries without capturing most of the resulting value. Conclusion The bullish investment thesis is that SpaceX is transitioning from a launch and communications company into a foundational infrastructure platform serving communications, AI, defense, and future space-based industries. If successful, this transition could support valuation outcomes substantially above conventional aerospace benchmarks A bet on technological progress, and the emergence of a large space-based economy

Dear ICP community, the Internet Computer has now been running strong for 5 years 👏👏👏 Here is a celebratory preview of ICP "cloud engines," the sovereign frontier cloud technology the network shall soon provide from Main points: — Cloud engines enable anyone to spin up their own sovereign frontier cloud. The technology involves an extraordinary inventive step, in which cloud is created from a mathematically secure network of nodes. The nodes run as part of the Internet Computer network ( but are selected and configured by the cloud engine's owner. — The frontier cloud provided by engines is strongly focused on enabling AI agents to build and update online applications and services for us. The world is changing fast, and nearly all new online apps and services are already being built with the help of AI, and thus cloud engines target the future of cloud. — Software hosted on cloud engines is tamperproof, which means that it is immune to infrastructure hacks, because it runs inside a mathematically secure network protocol, rather than on computers directly. This means that AI agents, and those building with them, don't need to have a security team in the loop, or to trust someone else's security team. This is crucial, because in the future, non technical people will demand the freedom to build with full automation — where they just need to issue instructions to AI about what to build, and don't need to worry about anything or anyone else. Of course, apps and services running on engines are also vastly safer from the new breed of hacker being enabled by frontier AI. (The cloud engines themselves are also "tamperproof." Even if a hacker gains physical access to some portion of a cloud engine's nodes, and can make arbitrary changes, the computations and data of the hosted apps and services cannot be corrupted or interrupted so long as the network's fault bounds aren't exceeded. The recent hack of Vercel, a major cloud platform, which gave hackers access to the apps it hosted, provides additional perspective on the importance of this advantage.) — Software hosted on cloud engines is guaranteed to run, so long as a sufficient number of the engine's nodes are running. This means that AI can build applications and services without the need to have a human systems admin team constantly tinkering with the underlying platform to keep it running, which is again crucial, because in the future, non technical people will expect the freedom to use AI to build without the support of others. — New frontier programming language technology, in the form of the Motoko language developed by Caffeine Labs, leverages seminal "orthogonal persistence" technology that unifies program logic and data to deliver further unlocks for AI (Motoko is the first computer language being developed that targets agents that are writing software rather than humans engineers per se). Nowadays, AI can build and update production apps at a prodigious rate, even at the speed of conversation. But it can also make mistakes, and there's a risk that an update it creates might be "lossy" in the sense it causes some transformed data to be lost. Again, in this new world, it's both undesirable and impractical for everyone to have to have a systems admin team on-hand to detect lossy updates and roll them back, but Motoko provides a solution: it can detect new software updates are lossy before they are applied, reducing potentially catastrophic errors by AI to harmless coding retries. — Software hosted on cloud engines is "serverless" but unlike traditional serverless software, directly it directly incorporates data through "orthogonal persistence." Another key purpose is simplify backend software logic and fuel the modeling power of AI by increasing abstraction (sorry for the technical language!!!). Put simply, this enables AI to produce more sophisticated backends, faster, and at dramatically lower costs, as measured by the number AI API tokens consumed during coding. (Tip for the technical: orthogonal persistence is a new paradigm where "the program is the database," and data lives inside program variables, which is possible because it's as if hosted software runs forever in persistent memory). — An expanding database of skills at shall make it possible to develop and directly deploy apps and services to your cloud engines directly from Claude Code, Perplexity, Codex and other AI platforms. Further, your account on can be connected, so that new apps and updates created through conversation automatically appear hosted from your cloud engine. In the future, R&D is going to be very seamless. You converse with AI, and your secure and unstoppable apps or services are created or updated. Cloud engines are designed to directly support this "self-writing cloud" future where we can work hands-free. — Tech sovereignty is becoming a huge issue worldwide, with governments and corporations seeking to create sovereign tech stacks owing to geopolitical tensions. Increasingly, people are realizing that tech provided by foreign nations can come with hidden backdoors and kills switches, from the base platform, right up through hosted apps and services. ICP technology is open source, and those building on ICP using AI own their own source code. When you have the source code, you can verify that there are no backdoors, and when you own the source code thanks to AI, you can update it at will, freeing you from vendor lock-in. But cloud engines take sovereignty much further... — You create a cloud engine by selecting the nodes that will be combined. You can choose the class of nodes used, and their number, but more importantly, you can choose who operates the nodes, and where they are located. Almost any configuration is possible, because the Internet Computer scales the security privileges afforded to hosted software within the network according to configuration (software hosted on cloud engines can directly interoperate with software on other engines and traditional subnets, but base restrictions are applied according to security rules). A cloud engine can be created within a region such as Europe, to comply with regs such as GDPR, or completely within a sovereign state like Switzerland or Pakistan. But cloud engines go further still... — Sovereignty is also about freedom from vendor lock-in. Cloud engines are essentially ICP (Internet Computer Protocol) network configurations, and this means the underlying compute nodes they combine can be swapped out without interrupting their hosted apps and services. This is a big deal. In addition, cloud engines now support nodes that are instances running on Big Tech's clouds, in addition to nodes that are dedicated specialized hardware, as per the Gen I and Gen II nodes that dominate the Internet Computer today. For example, it is possible to have an engine running across different AWS data centers, say, and then reconfigure the engine to run across a mixture of AWS, Google, Azure and Hetzner for even more resilience, without the users of hosted apps and services noticing a thing. That's true freedom. — Sovereign AI is becoming increasingly important too, and cloud engines allow special "AI nodes" to be added to them, so that hosted software can perform inference on hardware provisioned by the owner from a location the owner has selected. Even though the AI nodes are only accessible within the cloud engine, they can still benefit from the forthcoming Internet Intelligence Gateway (IG), which will make it possible to validate inference performed on key frontier open weights LLMs, even when the inference is performed on completely independent AI clouds. When the results of inference are received, this technology can verify that neither the prompt+context (input) nor the inference result (output) have been modified, and that the results were produced by the precise LLM expected. This ensures that AI clouds don't cheat by running inference on cheaper models than are being paid for, and bad actors aren't modifying the inputs or outputs to surreptitiously insert advertising into results, say, or change facts, or insert malware when code is being generated. What's super cool about this technology is the cost of the verification is scalable. A very valuable additional security can be achieved with only 1-2% of extra cost. — Scaling apps and services when they hit capacity limits is another thorny problem that cloud engines help the world address. Engines make scaling possible without rewriting or reconfiguring software. The query workload capacity of hosted software can be horizontally scaled simply by adding new nodes to an engine, and nodes can also be added in geographical proximity to demand. Meanwhile, update workload capacity can first be scaled-up by swapping an engine's nodes out for the next class up, and then when no larger class of node is available, horizontally scaled-out by "splitting" the engine into two, which doubles available capacity. (Technical tip: horizontally scaling update capacity by splitting engines requires multi-canister architectures). — For those who have been following how Caffeine builds apps that can efficiently store large numbers of files, I should mention that apps built on cloud engines will also support the new ICP Blob Storage cloud network (since cloud engines currently have up to about 3 TB of memory, which apps storing large amounts of files can easily exceed). We are also working on allowing blob storage nodes to be added to cloud engines, to enable sovereign mass blob storage within an engine, similarly to how AI nodes can be added currently. — Lastly, but certainly not least, I should mention that cloud engines are multi-blockchain capable, and ready for digital assets, thanks to the clever math at their core. For example, an e-commerce service built on a cloud engine can securely accept and custody stablecoin payments, or a multi-chain DEX could be hosted. Further, engines can support software autonomy (software orchestrated and controlled by other autonomous software, in a decentralized way) and can themselves be orchestrated by SNS technology, and thus run autonomously too. Today, though, the focus is on *mainstream* cloud. This year, the cloud industry will generate approximately one trillion dollars in revenue. That number is already huge, but is expected to grow to two trillion dollars by 2030. After years of continuous development, which have seen more than $500m spent on R&D, the Internet Computer network is now tacking directly toward this mainstream cloud market with cloud engine technology. In their first version, cloud engines are not meant to be a cloud panacea. For example, currently they are not ideal for working with big data. You should use something like DataBricks for that. Cloud engines are carefully targeted at enabling AI to produce traditional online applications and services, including SaaS, in a safer and more productive way, which represents a new market segment with tremendous potential. Of course, DFINITY will continue to work relentlessly to push forward ICP's capabilities, so expect further developments. It's worth mentioning that this cloud segment isn't just about creating new apps and services using AI, it's also about replacing legacy systems and apps built on super expensive SaaS services. Caffeine Labs is working to produce technology (Caffeine Snorkel) that can study an enterprise's legacy systems and app built on SaaS, create replacement systems and apps, and migrate the data, while supporting key stakeholders through the process over email and chat, with full automation. Thus the legacy systems and SaaS markets shall also be addressed by cloud engines. Zooming out, and reasoning in a more metaphysical way, we believe, as we always have, that there is room for a new kind of cloud created by mathematical networks, that provides seminal advances in the fields of security and resilience, as well as true sovereignty and freedom from lock-in. That this same technology, with the help of additional technologies like orthogonal persistence and Motoko, enables AI to build for us without the need for so much oversight, and to create more backend sophistication while consuming fewer AI API tokens, enables ICP to bring game-changing advances to the world. Cloud engines will work synergistically with the Intelligence Gateway, which will enable apps and services running on engines to seamlessly leverage AI, wherever that AI is running, while providing verifiability at extremely low cost for open weights frontier models. We believe that cloud engines represent an inflection point in the storied history of the Internet Computer project, and I'm very proud to be sharing the details with you on the network's fifth birthday 💪 I'll be back with more news soon!!

We are open-sourcing blcli: an Agentic Infra Stack, battle-tested at 30M+ user scale. It allows coding agents like Codex or Claude Code to help manage your whole cloud infrastructure through code, PRs, dry-runs, and deterministic apply workflows. A solid & serious infra that can support to millions of users. This is a collaboration across multiple teams, the same stack that powers @AlvaApp, @Galxe, @GravityChain, and @ReahPlatform. Check it out here: Docs: blcli: Production stack template: Personal account starter: A common take today is: AI agents are useful for toy apps and prototypes, but not for serious infrastructure. The conclusion is wrong, because the issue is not that agents cannot work on real systems. The issue is that real infrastructure requires a large amount of expert context to get it correct in the first place, and even more context to guide agents through the next 18 months of iteration. Production infrastructure is not just a few Terraform files or Kubernetes YAMLs. It includes: cloud projects IAM boundaries networking VPC / subnet / firewall design Terraform state and backend management Kubernetes clusters cluster add-ons secrets management Git-based deployment workflows observability and telemetry (logs, metrics, traces. All integrated together and ready for your Agents to debug live on your prod env) databases, often self-hosted for cost efficiency and control environment separation: stg / beta / prd operational runbooks rollback paths production failure patterns Most of this knowledge usually lives in senior engineers’ heads, internal docs, shell scripts, Slack threads, old runbooks, and lessons learned from real incidents. If an agent does not have that context, of course it will build toy infrastructure. So the real question is: How do we package production infrastructure expertise into a form that AI agents can read, reason about, modify, and operate safely? That is what blcli does. At its core, blcli is a CLI tool plus a whole package of best practices of Infrastructure as Code. The key design principle is simple: Agents are already very good at reading and modifying code. So we make infrastructure code-first. The generated repo is intentionally self-explanatory. An agent can open the repo and understand what happened, and what's next. Who blcli is for? We built blcli for two types of users. 1. Product teams that need to scale beyond prototypes The first group is teams building real products that need infrastructure capable of growing beyond the prototype stage. These teams want the speed of AI-assisted development, but they cannot afford toy infrastructure. 2. Frontier labs and agent teams building self-improving systems The second group is frontier labs, data companies, and agent teams that need infrastructure not just to run applications, but to train, evaluate, and improve agents. If you are building coding agents, infra agents, or long-horizon autonomous systems, blcli stack is a good agent harness/env. Authors: @SiriJhui @p0pUBhv35I8308 @alvinFu1 @ryan4yin @algoxstonk

1. Intro Vitalik recently wrote about where the EF should go; Aya added a note to explain how we got here, and why. I’ll write about the execution. We now have enough clarity to stop treating “what is the EF for?” as an open-ended question. Our mandate is clear: The EF exists to ensure Ethereum is, becomes, and remains real permissionless infrastructure for self-sovereignty: censorship (and capture) resistant, free and open source, private, and secure; and capable of supporting sovereignty-preserving coordination at scales where trusted institutions hitherto have been unavoidable. The following are my thoughts on some of the points that follow from the mandate and how we are translating it to action. But first, a short reminder about 2. What the EF is not for We are not here to optimize for EF importance, corpo/pol appeal, or ecosystem popularity. We are also not here to please short-term speculators, prop up TBTF neo-SIFIs, market every app on Ethereum, help anyone look good to their crypto or investor friends, or provide on-demand entertainment for dinner parties and private retreats. 3. What the EF is for: Eliminating weaknesses We are here to defensively strengthen places where Ethereum is, or can still become, extractive, totalizing, or vulnerable to cartel or state capture, or authoritarian tools of surveillance or coercion. We will base our actions on a full examination of what Ethereum is and can be at the protocol layer (what is actually running as “Ethereum”), the access layer (what users use to interact with the protocol), the user layer (the end-users who need and will need Ethereum), and the institutional layer (the intermediated paths that scale self-sovereign usage). The EF exists to harden every surface of Ethereum, including those where Ethereum can remain formally permissionless while becoming practically captured. Some obvious surfaces are the transaction pipeline, staking and network security, access layer standards and interfaces, self-sovereignty norms, privacy expectations, institutional adoption patterns, and social layer governance processes. The primary concerns are similar across most of them: does the status quo and its future trajectory minimize trusted dependencies, minimize points of leverage and capture vectors, make user privacy the default, preserve exit, and make trust assumptions legible? The work starts with the EF itself. We are moving compensation and major financial relationships toward ETH and mandate-compliant Ethereum-native stables, with exceptions where positive law or unavoidable operational constraints require exceptions. Rather than a purity ritual or instruction for people to take unmanaged personal risk, it is robustness, alignment, and product pressure. If the EF’s work is to make Ethereum usable as infrastructure for self-sovereignty, everyone at the EF will increasingly live inside the constraints of the system the EF exists to improve: wallet UX, volatility, accounting, privacy gaps, payment friction, stablecoin trust assumptions, recovery, dependency risk, etc. If we can’t use these tools ourselves, it is unrealistic to expect others to. Ethereum is already mature; those who do not depend on the user-facing stack have no business trying to shape its future, at any layer. The transaction pipeline is next. Preventing toxic MEV capture is core EF work, not a peripheral market-structure concern. Transaction supply, ordering, inclusion, block construction, propagation, and settlement are part of Ethereum’s neutrality boundary. Some MEV may persist as an adversarial phenomenon the protocol contains, but it must be absolutely minimized and, for that to be possible, we must guard against the acquisition of unwarranted influence by its beneficiaries. If credibly neutral execution is subverted by privileged orderflow, cartelized builders, trusted relays, opaque routing, or validators outsourcing into a narrow supply chain, Ethereum will look permissionless while users experience it as intermediated at the moment value moves. EF protocol work will therefore prioritize lower barriers to block building and validation, stronger inclusion guarantees, reduced extraction opacity, competitive transaction pipelines, user-facing legibility of trust assumptions, and more aggressively exploring the open orderflow solution space. None of this is simple. A good solution in one place can aggravate problems elsewhere. FOCIL is good for censorship resistance, but it may introduce more cross-block MEV. While ePBS solves the relayer trust problem, we must make sure that its implementation does not inadvertently obstruct long-term solutions to even larger problems. It would be unacceptable, for example, if ePBS enshrining the builder economy ends up making it harder to reduce reliance on the private orderflow that has emptied out the public mempool. Encrypted mempools may not only reduce pre-execution transparency and pending orderflow visibility, but also shift competitive advantage to new privileged actors, including specialized hardware operators in some designs, while adding protocol complexity. In order to avoid wasting time playing whack-a-mole, we must commit to solving the extraction problem at a whole system scale. Doing so will require creativity, courage, and the understanding that failure to solve this problem is unacceptable. If we fail, we will have left in place an unnecessary barrier to institutional adoption, but, more importantly, we will also have surrendered a core part of the promise of Ethereum - the replacement of extractive middlemen with permissionless, credibly neutral infrastructure and competitive markets. That must not happen. MEV is likely to be the next major front in the cypherpunk war. We must set ourselves up to win here. Privacy is just as fundamental. A public ledger without serious privacy defaults is a surveillance substrate with settlement guarantees. That is not an acceptable end state for the world computer. Unconditional privacy will be readily available across Ethereum, with programmability on top for selective disclosure, proofs, auditability, compliance logic, reputation, governance, identity, and other constraints chosen by users and their communities. The temporal order matters: unconditional privacy must exist first, opt-in constraints come second. It is also important to avoid forcing users to assemble a fragile stack of special wallets, RPCs, bridges, apps, compliance providers, and operational habits to attain privacy. Deep privacy must be more secure than this. Privacy is a condition for Ethereum’s viability as freedom-respecting coordination infrastructure and as such must be robust. Staking must be treated as protocol infrastructure risk. Staking is not merely a yield product, and liquid staking is not merely an app-layer market. If stake, liquidity, validator access, DeFi collateral, and governance influence concentrate around a small set of issuers or operators, Ethereum’s security layer becomes vulnerable to capture through capture of the economic layer around it. EF will support research, specifications, and designs that keep staking permissionless, private where possible, plural in operation, and resistant to intermediaries becoming permanent control points. The access interfaces are where users access either the protocol directly or through intermediated defaults. The primary problem to solve here is not getting Ethereum into more rooms directly, but making its users, both end users and institutions, more self-sovereign and less susceptible to coercion, and avoiding normalization of soft coercion in exchange for reach. EF will not help Ethereum become more acceptable by sanding off the properties that make it uniquely valuable. Ethereum does not need to become another permissioned settlement backend with better branding. It needs to show, in production, that self-sovereign coordination at scale is possible. Across Ethereum, the EF’s defensive work seeks to ensure that Ethereum is infrastructure people can still use when counterparties fail, platforms censor, governments overreach, intermediaries extract, and coordination problems become infeasible for trusted systems to handle. A core part of that is to make that infrastructure secure and robust against capture at every layer wherever capture opportunities can hide. 4. What the EF is also for: Seizing opportunities Shoring up the fundamentals is not enough. Ethereum’s potential is still largely unrealized, but that does not mean that the path ahead is going to be straight. Opportunities must be seized when the time is right. At this moment in time, a number are visible, including: * Ethereum becoming the first quantum-resistant global infrastructure. Ethereum researchers will lead the post-quantum cryptographic migration before the threat becomes urgent, not after it becomes a governance emergency. That means hardening Ethereum’s cryptographic foundations while there is still time to design carefully. The same applies to other long-horizon risks, where waiting for market demand means waiting until the window for principled design has already closed. * Verifiably self-sovereign stack, from soup to nuts, whether local or remote, with no censorship or extraction openings: browsers, wallets, intents, broadcasts, orderflow, inclusion, block construction, proposal, proving, exit, and recovery. Minimal MEV, and zero toxic MEV entrenchment, either in or around the protocol. No execution layer that is formally permissionless but practically gatekept by privileged supply chains. If there’s a funnel towards an extractive private lane, there’s other options that keep the game live. The goal is not only to prevent extraction or capture, but to make credibly neutral execution competitive enough that serious users prefer it. * Making ETH normal digital cash: a private, dignity-respecting, debasement-resistant and surveillance-resistant medium of exchange and store of value, as well as the native asset of private computation and private coordination for both humans and their agents. If Ethereum can make private economic life and private institutional life possible without routing users back through the friction and potential abuse of custodians, surveillance vendors, or permissioned ledgers with softer branding, as well as provide a venue for secure and competitive machine economics, the value unlocks will be immense. * Personal wallets with personal AI agents that users can actually own and run on their own personal computers. Not your keys, not your coins; not your model, not your mind. As agents become interfaces for more economic and social action, the question of who owns the wallet, the model, the memory, the policy, and the signing authority becomes an existential question about sovereignty instead of UX details - we are all users above any other roles, and no one at EF will forget this. * Institutional and enterprise use cases where Ethereum wins by not disappearing into an invisible backend, gatekept by intermediaries or terrible UX, and by not compromising into a compliant fintech rail with web3 branding. Rather, we will win through proving that credibly neutral infrastructure can handle disintermediated coordination so competitively that trusted intermediaries have to meet Ethereum users on Ethereum’s terms. * Security-preserving scaling. L2s and related infrastructure will be able to meet institutional-level needs without accepting dependencies on closed operators, opaque sequencing, custodial UX, or upgrade committees that users cannot realistically exit. Scale is not throughput alone. Scale is the guaranteed availability of self-sovereignty under real load. We are ensuring Ethereum remains the hardest bedrock for settlement, local and worldwide; and beyond that, a civilizational ledger and execution substrate to stand the test of time. When future civilizations speak of the infrastructure they inherited from the Antiquity of the Information Age, their first example should be Ethereum. Ethereum will outlast all of us. More than enough people watching understand this. Many wondered why it needed saying at all, but it did. If you don't believe us or don't get it, we don't have time to try to convince you, sorry. 5. Addressing departures There has been a lot of online speculation about departures from EF, both before and after the mandate. Some people resigned, others were terminated. Some departures were about strategy, some about role fit, some about normal institutional change, and some simply about people deciding that their best work for Ethereum should happen somewhere else. We will not litigate individual personnel matters on Twitter. That is the default because it is better for EF, better for the people involved, and better for Ethereum. People who contributed through EF deserve dignity on the way out. They do not deserve to have their employment history turned into factional content. Where possible, we have let people describe their departures in their own words as a matter of courtesy, and not concession. If public claims materially mislead people about EF’s direction, decision-making, or mandate, we may correct the record at the level of policy, process, and institutional facts. We still will not turn personal files into public spectacle. Ethereum is permissionless. People may disagree, criticize, compete, fork, and build elsewhere. We intend to keep exits dignified and expect others to do the same. It will suffice to say that we are thankful for what all contributors have built; we will continue to do work Ethereum needs. 6. Addressing EF spinouts Some work should and will leave the EF in the months to come. We hope and expect this process to result in some excellent work being done in service of scaling self-sovereign adoption, but we also must take care lest it becomes an abdication of responsibility or an excuse for undisciplined spending. Some work is not mandate-compatible and should not be carried forward with EF funds or EF endorsement, either inside or outside the Foundation. The efforts carried out by the spinouts will vary widely. Some efforts will leave EF because another org would be a better home for them; others will leave because markets should decide on their worth. Some will leave because they are not compatible with the direction set out in the mandate; others because they are useful but not EF work. Just as a spinout is not automatically good because it reduces EF headcount, former EF affiliation is not a claim on EF funding. The question we ask when deciding on funding is not “did this come from the EF?” But, rather the questions that should be asked about all external funding: “Is this work mandate-critical? Would the EF do this work internally if it had the organizational and financial capacity? Is there no better natural home? Can the external party execute without increasing capture risk, private extraction, opacity, or dependence? Does supporting it reduce Ethereum’s dependence on the EF over time, without prematurely transferring resources and legitimacy to new organizations and thereby risking operational failure or mission drift?” EF funding for work being done externally can be appropriate when it is a capacity solution for mandate work - work the EF should responsibly want done; work that protects CROPS; work that advances self-sovereignty and scales it; essential work that no actor can or will reliably do without EF funding; and work that can be scoped, reviewed, and held accountable without creating a permanent dependency. Such funding is not appropriate when it is a lazy continuity payment, a friendship payment, a reputational hedge, a way to avoid making a hard decision, or a way to support work that is not compatible with the mandate. EF has finite funds, finite legitimacy, and a specific mandate. We will spend all three as if they matter. When we say “EF is one of many nodes”, we mean that we intend to be one of many nodes working to keep self-sovereignty and its scaling the North Star, and working to keep CROPS the undisplaceable first-class properties of the network. We don’t mean that we will support orgs or projects with different priorities. Diversity that leads to ecosystem resilience, coordination cost right-sizing, and better decision-making is good. Diversity that leads to mission drift is not. We are not neutral on the direction Ethereum takes. CROPS are not just things we “believe in”, they are characteristics we understand must be thoughtfully prioritized at every fork for Ethereum to realize its potential. We are partisans for and builders of something of such incredible neutrality that it will fundamentally reshape the world we live in; we wish to work with everyone committed to this shared purpose.

We're introducing GLM-5.2, our latest flagship model for long-horizon tasks. It marks a substantial leap in long-horizon task capability over its predecessor GLM-5.1 and, for the first time, delivers that capability on a solid 1M-token context. GLM-5.2's new capabilities include: Solid 1M Context: A solid 1M-token context that stably sustains long-horizon work Advanced Coding with Flexible Effort: Stronger coding capabilities with multiple thinking effort levels to balance performance and latency Improved Architecture: We propose IndexShare, which reuses the same indexer across every four sparse attention layers, reducing per-token FLOPs by 2.9× at a 1M context length. We also improve GLM-5.2’s MTP layer for speculative decoding, increasing the acceptance length by up to 20% Pure Open: An MIT open-source license — no regional limits, technical access without borders Supporting long-horizon tasks starts with making long context engineering-usable: the model must maintain quality across long, messy coding-agent trajectories, not just accept more tokens. A 1M context is easy to claim, but much harder to keep reliable under real engineering pressure. To this end, we substantially expanded 1M-context training for coding-agent scenarios, covering large-scale implementation, automated research, performance optimization, and complex debugging. The result is a long-context system that is not only wide in scope, but solid in execution: a practical substrate for sustained engineering work. This capability is reflected in GLM-5.2's performance on three long-horizon coding benchmarks. FrontierSWE measures whether an agent can complete open-ended technical projects at the scale of hours to tens of hours, spanning systems optimization, large-scale code construction, and applied ML research. On this benchmark, GLM-5.2 trails Opus 4.8 by only 1%, while edging out GPT-5.5 by 1% and Opus 4.7 by 11%. On PostTrainBench, where each agent is given an H100 GPU and evaluated by how much it can improve small models through post-training, GLM-5.2 outperforms both Opus 4.7 and GPT-5.5, ranking second only to Opus 4.8. On SWE-Marathon, an ultra-long-horizon software engineering benchmark covering tasks such as building compilers, optimizing kernels, and developing production-grade services, GLM-5.2 still has room to grow, trailing Opus 4.8 by 13% while remaining second only to the Opus series. Across all three benchmarks, GLM-5.2 is the highest-ranked open-source model, showing that its 1M context has translated into practical long-horizon delivery capability.