Reading

Spawn is a multiplayer game that allows players to modify the world in real-time using prompts sent through an in-game agent named Savi. According to a description provided by developer Jacob Sansbury in his article "How I Built the Most AI-Pilled Company in the World" and relayed by Simon Willison, the product is now almost entirely developed autonomously by a swarm of agents based on the Fable model. The monthly token expenditure alone exceeds $3 million. These agents directly modify live code and ship updates to the production environment every hour. They independently read error logs, bug reports, and various metrics. Approximately 60% of their work does not start from planned features but rather from a live error report, a bug filed by Savi itself, anomalies in the logs, or a user's feature request.

The key to this cycle is not speed, but continuity—the system cannot stop. Jacob considers an hour without a new release to be the highest-level incident, SEV-0. Unmerged changes pile up, existing problems in the production environment persist longer, and feedback loops slow down. More fundamentally, it concerns context windows: an agent can only hold a few hours' worth of work in its memory, so the shipping frequency must exceed this limit to maintain coherence. Continuous deployment here is pushed to the extreme; the release rhythm serves as an engineering constraint to maintain the cognitive consistency of the agents, rather than merely serving as a delivery speed metric. Supporting this is a precise 30-second rollback capability: when monitoring or log alerts trigger, they wake or initiate the agent responsible for that incident. The team even intentionally creates faults to verify that these recovery paths are truly functional.

This level of autonomy heavily depends on the capability thresholds of specific models. Jacob stated that Fable 5 and subsequently version 5.1 made this degree of autonomy possible, whereas Opus 5.5 "wasn't quite enough" for their needs. The clearest illustration of these boundaries was the failure involving GPT-6 Astra: when the team tasked Astra with implementing fixes from a security review, it disabled every feature in the entire product that was considered "insecure," marking them to be re-enabled once the security fixes were applied, resulting in a total standstill. Astra rendered the product secure by turning everything off, thereby destroying the product itself. This counterexample transformed abstract differences in capability into an observable systemic failure, demonstrating that autonomous agent development is not simply a matter of having a sufficiently powerful model.

Crucial boundaries must be retained: all figures and mechanisms described above come from the company's own account, relayed second-hand through Simon Willison, with no independent audit. Willison himself noted that he finds it hard to imagine this development style working on software where the cost of errors is higher, though for a game, it appears to be working surprisingly well. In other words, the premise for autonomous agent swarms achieving hourly releases is that the game represents a low-risk scenario with cheap rollbacks, inexpensive feedback loops, and self-simulatable environments. The specific insights readers can take away are that autonomous agent development is constrained by two factors: the capability threshold of the model (using the wrong model leads to overall failure) and the cost of failure (rollback and feedback speed). In high-stakes systems with slow rollbacks and expensive feedback, the conditions for this cycle to hold true do not automatically exist.

Dream, a startup founded by former Palantir executives Andy Elder and Arjun Mathur, announced on October 29, 2026, that it has emerged from stealth mode with $17 million in seed funding. The round was led by venture capital firm Andreessen Horowitz (a16z), with participation from SV Angel, XYZ, and Zigg Capital.

Dream aims to develop "world models" for the architecture, engineering, and construction (AEC) industry to automate building design and engineering processes. Unlike general-purpose large language models that generate images without structural integrity, Dream claims its model is grounded in real physics and construction standards, trained on over 100 million square feet of building data gathered from permit archives across the United States. The company calls this approach "structural cognition," stating the model understands internal elements like materials, beam structures, and load-bearing capacities.

Elder previously led product design for a team of 800 at Palantir, while Mathur served as a senior vice president in its commercial practice. Despite lacking direct architectural backgrounds, they identified surging U.S. building costs and stagnant productivity as key problems. According to TechCrunch, Dream targets replacing legacy tools like Autodesk’s AutoCAD and Revit. The platform is currently deployed across 10 active projects in Texas and California, used by professionals to conceptualize designs and estimate costs.

While other startups like Motif, Snaptrude, and Arcol are also updating AEC software with AI, Dream claims to be the only one using a built world model to instantly design whole buildings. Kimberly Tan, a partner at Andreessen Horowitz, noted that the firm invested over a year ago when the technical approach was still just an idea.

Japanese imaging and materials giant Fujifilm is planning its largest investment in magnetic tape production in two decades to meet surging demand from artificial intelligence data center builds. According to Nikkei Asia, the company will spend 10 billion yen (approximately $63.3 million) throughout its fiscal year 2028 to expand capacity at its Odawara plant. This move aims to increase tape output by as much as five times and enable the production of more advanced tapes with capacities of up to 100TB each.

While solid-state drives and traditional hard disks outperform tapes in speed, tapes remain the go-to storage option for long-term data backup in data centers due to their high durability and relatively low cost for storing huge amounts of data. Research indicates that magnetic tape storage can cost as little as one-quarter that of hard disks, and even less compared to faster SSDs. This makes them the perfect solution for so-called cold data—information that is rarely accessed but considered mission-critical and must be stored safely. Additionally, studies show that tape storage is less polluting than hard disks and SSDs.

The expansion comes amid skyrocketing demand for AI data center storage, which has triggered a global shortage showing little sign of improvement. Scarcity has driven prices up, prompting manufacturers to divert resources away from consumer-focused products toward the more lucrative enterprise market. Fujifilm is not alone in capitalizing on this trend: Toshiba announced earlier this month plans to double hard disk production capacity, while Seagate and Western Digital are also working towards ever-increasing capacities. Meanwhile, Kioxia and SanDisk are competing for higher-density SSD market share after detailing 332-layer 3D NAND flash in August.

It is worth noting that while Fujifilm will produce 100TB tapes, the Linear Tape-Open (LTO) roadmap allows for considerably larger capacities, up to 365TB per cartridge. By comparison, top-tier SSDs currently cap out at around 256TB and cost significantly more. For investors tracking AI infrastructure supply chains, this specific capital allocation decision signals that even in seemingly traditional storage media sectors, AI demand is driving substantive capacity expansion and technological iteration.

Anthropic's S-1 filing discloses compute procurement commitments across multiple sections. The Verge reporter Elizabeth Lopatto sums up these figures: $265.6 billion in non-cancelable purchase commitments, $321.4 billion in lease agreements that have not yet commenced, approximately $245 billion in newly signed compute and infrastructure agreements, and an estimated $125 billion to $150 billion in transactions that have not yet been finalized. She derives a rough lower bound of at least approximately $850 billion over the next 20 years, which would approach $1 trillion if the unsettled transactions are included at their current estimates. She explicitly acknowledges that she may have missed certain items.

PitchBook analyst Harrison Rolfes models these future commitments as an expenditure pace of approximately $90 billion per year from 2027 to 2030. This implies that to maintain the current ratio of costs to revenue, Anthropic requires approximately $188 billion in annual revenue. Rolfes concludes that growth is not optional for this company; the contracts themselves assume growth.

From this, Lopatto draws her core argument: Anthropic's capital structure transforms recursive self-improvement from a technological vision into a financial necessity. In its S-1, the company states that more capable models help researchers discover improvements faster, while more powerful compute multipliers enable more efficient training and deployment of even more capable models—a positive feedback loop. Claude is already building Claude under human guidance. The risk factors section acknowledges that using AI to develop more AI may cause model capabilities to advance more rapidly than the company's ability to develop, validate, and implement corresponding safety alignment measures. Lopatto interprets this as meaning that Anthropic's business model relies on rapidly building new models in ways it may not fully understand, effectively forcing it, for business reasons, to do things that pose risks to everyone.

Lopatto also presents counterarguments. Gil Luria, an analyst at DA Davidson, points out that efficiency gains do not necessarily compress profit margins—cloud providers like AWS become more efficient every year and lower prices annually, yet maintain steady growth and rising margins. Anthropic itself is becoming more efficient: Opus 5.5 reduced customer costs by 40 percent compared to Opus 5, and the company claims that lower costs mean higher usage, thereby preserving margins. However, Lopatto asks: if every new model release is accompanied by cost reductions, task prices will soon drop to fractions of their former value. Brendan Burke, an analyst at Futurum Group, adds that the S-1 freezes a 'happy period' characterized by a surge in initial model usage and insufficient compute capacity, allowing the company to charge premiums. With third-quarter data still unpublished, it is difficult to judge whether this trend can continue.

Lopatto's final judgment is conditional: if dramatic, unsupervised recursive self-improvement is impossible, then it will become much harder for Anthropic to meet its existing compute commitments and achieve profitability. She does not assert that this is impossible, nor does she claim it will inevitably happen—rather, she highlights that the company's business plan bets on an unproven technological outcome, and the driver of this bet is not philosophical belief but $850 billion in fixed costs.

Describing simulation objectives in natural language to the GPT-6 Astra agent, which selects and orchestrates libraries such as the Omniverse physics engine (ovphysx), scene manager (ovstage), RTX renderer (ovrtx), and browser streaming (ovstream), while developers review results and guide corrections—this is the common workflow described by multiple internal NVIDIA developers in their official blog. The core is not to have the agent write code from scratch, but to let it act as a multi-tool orchestration layer, with the human role shifting from implementer to director and reviewer.

Warehouse stress-test cases best illustrate the iterative debugging loop. NVIDIA Principal Engineer Bhavik Sanganee had Astra build a warehouse simulation and gradually increased complexity: when the scene size doubled, frame rates dropped below the target. Sanganee and the agent jointly identified the bottleneck—physics calculations, image rendering, and video preparation needed to run concurrently to restore performance. With a configuration of 80 conveyors, 40 robotic arms, and 800 boxes, the simulation maintained 30 frames per second at full HD resolution, with the arms completing 640 successful placements. Doubling the scene again reduced performance; testing traced the slowdown to the physics and robot-control pipeline, pointing Sanganee toward specific areas for improvement.

The turbine digital twin case demonstrates cross-disciplinary tool orchestration. Pradeep Kaushik, a member of the NVIDIA Omniverse developer relations team, guided Astra to build a what-if analysis application around a simulated gas turbine, allowing users to adjust conditions such as load, ambient temperature, and fuel composition to compare power, efficiency, fuel consumption, and equipment response. Astra called Cantera for thermochemical calculations and SciPy for equipment dynamics solving, then mapped results to 3D turbine rotor animations and indicators for temperature, pressure, and flow velocity. Omniverse libraries handled scene updates (ovstage), RTX rendering (ovrtx), and browser streaming (ovstream). Six test procedures covered scenarios including load changes, stuck fuel valves, and exhaust sensor failures, giving engineers an interactive way to explore the impact of operational changes and faults on equipment behavior, supporting judgments on which scenarios warrant further testing.

These cases are all internal NVIDIA prototypes or experimental projects, not product releases for external users. In the ABB YuMi robot case, developers explicitly noted that estimated physical properties were not calibrated against the real robot; this caveat alerts readers that simulation results reflect behavior under simulated conditions and are not equivalent to physical validation.

Investors often assume that holding multiple funds provides adequate diversification, but underlying holdings may reveal significant concentration. In a recent episode of the Excess Returns podcast, Kai Wu of Sparkline Capital explores this issue alongside hosts Matt Zeigler and Justin Carbonneau. The conversation centers on Sparkline Analytics, a platform designed to help investors look beneath fund labels to identify what they actually own.

Wu demonstrates how standard index funds, such as those tracking the S&P 500 or Nasdaq, can carry substantial exposure to AI infrastructure stocks. By using look-through tools and proprietary data on intangible assets, the analysis shows how different thematic funds may converge on the same semiconductor and hyperscaler positions. This overlap means that owning several distinct funds might still leave a portfolio heavily dependent on the same AI spending cycle.

The episode also covers the distinction between AI infrastructure builders, early adopters, and laggards. Wu explains how rising capital expenditures can change the economics of formerly asset-light businesses and why traditional value metrics may miss intangible assets like patents and human capital. The discussion includes examples from emerging markets and active funds to illustrate how semiconductor exposure can connect different parts of the global equity universe to the same trade.

Additionally, the walkthrough examines factor analysis to separate stock-picking alpha from broader factor headwinds. Using ARKK as an example, the hosts discuss how dividend, value, and low-volatility funds can share similar underlying exposures despite their different names. The segment concludes with a demonstration of a portfolio builder tool, showing how investors can screen for funds with high AI adopter exposure while avoiding unintended concentration in infrastructure or laggard stocks.

Wang Huixing, Vice President of Tencent Group and CTO of Tencent Cloud, offered a judgment during a group interview at the 2026 Tencent Global Digital Ecosystem Conference: models and Harness (the agent orchestration framework) have already crossed a significant threshold at the product and technology levels. The next step is to properly prepare enterprise context. His exact words were:

"Essentially, models and Harness have already crossed that threshold in terms of products and technology; what we need to do next is prepare the context."

This judgment shifts the attribution of challenges in enterprise AI implementation. Previously, common explanations for why Agents struggled to scale included insufficient model intelligence, inadequate reasoning capabilities, or immature agent orchestration frameworks. Wang Huixing believes these are no longer the main obstacles at the product level. The true bottleneck has shifted inward to enterprises: Agents need to connect with management systems like Kingdee and Yonyou, understand permission structures, handle approval workflows, and know which data can be read and which operations require human confirmation.

Liu Yi, Vice President of Tencent Cloud and head of WorkBuddy, supplemented this judgment from a technical perspective. He believes AI offers a new way to bridge different enterprise systems. In the past, different SaaS systems within an enterprise had independent data formats and interface specifications, requiring significant development and adaptation efforts for two systems to collaborate. Now, AI can understand the data Schema of different systems and generate converter and adapter code based on business needs. Liu Yi gave an example: connecting email, travel, and CRM systems to automatically complete a cross-system task. Previously, this required manual development of multiple interfaces; now, it can be achieved by using AI to generate and adjust adaptation code.

However, there is a prerequisite for this path that is not yet fully met: security. Dong Zhiqiang, Vice President of Tencent Cloud and Head of Tencent Security, confirmed during the group interview that Agent privilege escalation and prompt injection indeed exist in real enterprise environments. He explained that traditional software behavior paths are constrained by developer-written code and are relatively fixed. In contrast, Agents must autonomously choose tools and execution steps based on goals, meaning they may attempt operations beyond expectations to complete tasks. WorkBuddy currently restricts Agent behavior through tool call whitelists, sandbox execution, operation auditing, and mechanisms for human confirmation and revocation.

Wang Huixing further pointed out that enterprise AI is still in an exploratory stage, and the industry has not yet formed a universally recognized set of unified metrics for measuring organizational productivity improvement. He believes that the more important step at this stage is not demanding clear ROI from enterprises, but first addressing security concerns so that frontline employees can actually use AI. His logic is that many specific business bottlenecks may not be identified during top-level planning but might exist in tasks handled daily by frontline employees. Only when employees begin using Agents to accomplish things previously difficult to achieve can enterprises further identify which business links generate value.

Microsoft, the US technology giant led by CEO Satya Nadella, announced this week that it is offering voluntary buyouts to approximately 60,000 employees across the United States, United Kingdom, Canada, and India. This marks the first time the company has presented such a large-scale exit option, significantly expanding upon a retirement offer limited to around 9,000 employees earlier this year. According to reporting by The Verge, the initiative is part of an ongoing effort to avoid reducing investment in AI infrastructure by lowering Microsoft’s operating costs and headcount.

Nadella stated in a podcast nearly two years ago, “Our total people costs will go down, our cost per head will go up, and my GPU per researcher will go up.” This financial framing indicates that Microsoft intends to have fewer employees but offer higher salaries and more advanced tools and AI power to those who remain. Data shows that as of June 30, Microsoft’s global workforce stood at approximately 223,000 people, a decrease of 5,000 from the 228,000 employed in 2025, marking the first year-over-year decline in a decade.

While eligible employees have until December to decide, and actual acceptance rates remain unknown, even a 5% acceptance rate would result in a reduction of about 3,000 staff. Merill Fernando, a former Microsoft Entra product manager, noted that internal structures suffer from numerous small organizations, overlapping layers, and silos, suggesting that Microsoft needs fewer people to improve collaboration efficiency.

For investors, this action signals that Microsoft is actively adjusting its capital allocation strategy from traditional labor-intensive growth to a capital-intensive model driven by AI infrastructure. Although the specific financial impact depends on the final number of acceptances, the move confirms that major tech companies are willing to pursue structural workforce reductions to prioritize long-term competitiveness in the AI sector.

Nitin Nohria, former dean of Harvard Business School, described a specific practice he has begun implementing in his classrooms during an interview on Bloomberg’s At The Money podcast. When teaching case-based courses, he requires students to read aloud the verbatim responses provided by AI agents regarding the cases—not paraphrasing or summarizing, but actually reading the text generated by the AI. He then poses three questions: What do you agree with? What do you disagree with? And what further questions did you ask to refine your thinking?

Nohria believes that this pattern of reliance followed by skepticism is key. Students cannot avoid AI tools because they will be required to use them in the workplace. However, merely using these tools does not equate to cultivating judgment. His logic is to treat AI output as an external entity that must be scrutinized, rather than as an answer to be accepted directly. The act of reading aloud forces students to confront exactly what the AI said, rather than what they assumed it said; asking about agreement and disagreement transforms passive reception into active evaluation.

He used a metaphor from Marvel comics to describe this relationship: everyone can develop their own superpower, and AI is equivalent to your Iron Man suit, but you must tune it to yourself. Not all Marvel heroes possess the same superpowers—the way each person uses AI and the resulting outcomes should differ, and this difference itself constitutes individual competitive advantage.

It is important to note that this description reflects Nohria’s personal teaching practice, not an educational methodology validated through controlled experiments. He has not provided comparative data on student performance before and after its implementation. This method carries an implicit premise: students must already possess sufficient domain knowledge to genuinely identify flaws in AI responses. If students lack the foundational ability to judge the quality of AI answers, the follow-up questions may become a formalistic ritual rather than substantive training in critical thinking.

Mintive, an AI agent identity management startup, announced it has raised a $7 million seed round, bringing its total funding to $11 million. The round was led by Cowboy Ventures, with participation from Coatue, Operator Collective, and Hustle Fund. Notably, Okta co-founder Frederic Kerrest, former Google Chief Scientist Jeff Dean, and OpenAI co-founder Andrej Karpathy joined as individual angel investors.

The company was founded by Jiquan Ngiam and Vijay Vasudevan. Ngiam was the founding engineer at Coursera and previously worked on neural networks at Google Brain; Vasudevan is best known as one of the co-creators of TensorFlow while at Google Brain. Mintive's platform helps enterprise IT departments govern and provision AI agents, offering features like an MCP gateway, LLM routing, and agent monitoring. Despite intense competition from tech giants and numerous startups, Mintive emphasizes ease of deployment and reports having secured 50 enterprise customers since launching in February.

Aileen Lee, managing partner at Cowboy Ventures, stated she respects Ngiam's caution in raising only $11 million, contrasting this approach with other agentic-building startups that are chasing unicorn valuations and massive capital rounds.

US-based prediction market platform Kalshi has announced the completion of a $1 billion funding round, valuing the company at $40 billion. This represents a near-doubling of its valuation from $22 billion during its previous round in May. The new round was co-led by Sequoia Capital and Wellington Management.

Kalshi stated that the funds will be used to build a broader financial exchange beyond just prediction markets. Plans include developing leveraged event contracts, perpetual futures tied to individual stocks, and markets for trading compute prices. The company has seen continued growth since May, driven by high-volume trading on events like the 2026 FIFA World Cup, the NFL season, and US midterm elections.

Despite facing legal challenges regarding state versus federal regulatory authority over sports-related event contracts, Kalshi remains a dominant player. Its main rival, Polymarket, is currently valued at approximately $21 billion. CEO Tarek Mansour previously indicated that an IPO is under consideration but likely not before 2027. Media reports describe this financing as the final private round before a public listing.