Tuesday, June 30, 2026 Trend Press · Cloudflare Pages

The Trend Tribune

"All the trends that are fit to read" Morning Edition Free of Charge
TODAY'S LEAD STORY

Chamath-Led AI Coding Startup Raises $135M—The Era of 'Don't Write Code, Design It' Is Coming

Prominent investor Chamath Palihapitiya raised $135M in a Series A round and took on the role of CEO himself. The very nature of what it means to be a developer in the AI era is being fundamentally reimagined.

By tech_tanaka · June 30, 2026
Chamath-Led AI Coding Startup Raises $135M—The Era of 'Don't Write Code, Design It' Is Coming
(Photo: illustrative)

The Day an "Investor" Became an "Operator"

One of Silicon Valley's most outspoken investors has made his move. The AI coding startup founded by Chamath Palihapitiya announced a $135 million Series A raise on June 29, 2026 — and in the same breath, Chamath revealed that he would be taking the helm as CEO.

It's not unheard of for a VC to step into the CEO role at one of their portfolio companies. But given Chamath's reputation — first as the founder of Social Capital, then as the architect of the SPAC boom — the question rippling through Silicon Valley circles is: "Why build a product yourself, and why now?"

Is the AI Coding Market a "Bubble" or "the Real Deal"?

The AI coding market is clearly running hot right now. From GitHub Copilot to Cursor to Amazon Q Developer, products are launching in rapid succession from incumbents and upstarts alike, and the daily reality of software development has changed dramatically over the past two years. VCs have been racing to write checks, eager to ride the wave.

This latest raise fits squarely into that narrative. At $135 million, the round is an exceptionally large Series A — most typically fall in the range of a few million to tens of millions of dollars. The sheer size symbolizes a market-level bet, and signals that the VC community's appetite for AI coding tools hasn't waned in the slightest. If anything, it's accelerating.

From "Writing Code" to "Directing the Design"

The core of Chamath's vision, as he tells it, is "the end of the era where developers write code line by line." His argument is that AI will handle implementation, while humans focus on architectural design and defining specifications — and that this division of labor will become the standard in the near future.

This isn't just a catchy tagline. In practice, it's becoming increasingly common at Silicon Valley startups to see a single engineer running multiple AI agents — autonomous AI programs that execute tasks on their own — and producing output that rivals what a team of five to ten people once delivered. We used to talk about the "10x engineer," but now you're even starting to hear people say "100x engineer."

Anthropic Teams Up with "The Government"

Slightly outside the AI coding storyline, but equally hard to ignore, is another piece of news that dropped the same day. Anthropic has struck an agreement with California Governor Gavin Newsom, allowing state government agencies to access Claude at half the standard price.

Bringing AI into government is a fundamentally different challenge than selling to private companies. Security requirements, legal constraints on data management, the complexity of procurement processes — clearing all of those hurdles while still landing at half-price pricing represents a significant strategic commitment on Anthropic's part.

The backdrop here is worth paying attention to, though. The federal government has been signaling a cooler stance toward Anthropic, and the company finds itself in an increasingly delicate political position as OpenAI's chief rival. The deepened relationship with California can be read as an attempt to offset that federal-level headwind at the state level — and for Governor Newsom, it's a prime opportunity to reinforce California's brand as a leader in AI adoption.

Takeaway: The Reshaping of "Who Builds" and "Who Uses" Is Accelerating

What today's news makes clear is that the AI industry is moving into its next phase. Capital continues to pour into startups building the tools, while the organizations on the other side — governments included — are getting serious about AI adoption.

What will Chamath build as CEO? How far will Anthropic's influence extend through its government partnerships? Both will be important things to watch from late 2026 into 2027. It's well worth keeping an eye on these developments over your morning coffee.

References: techcrunch.com / techcrunch.com
AIコーディングスタートアップ資金調達Chamathシリコンバレー

'Throw and Deliver': New Frontiers in Drone Logistics — Using Reinforcement Learning to Precisely Release Cable-Suspended Payloads

By thesis_yamamoto

A recent paper published on arXiv presents a reinforcement learning-based approach to achieving precise dropping of cable-suspended payloads by drones, with promising applications in search and rescue and medical delivery.

Drones Enter the "Throwing" Era — The Next Frontier of Logistics Robotics

On June 30, 2026, a series of papers with the potential to overturn conventional wisdom in drone logistics were published on arXiv in the robotics field (cs.RO). Among them, drawing particular attention is research on "dynamic payload release of cable-suspended loads" using quadrotor drones.

From "Carrying" to "Throwing" — What Makes It So Hard?

The paper "Learning to Throw: Agile and Accurate Cable-Suspended Payload Delivery with a Quadrotor" (arXiv:2606.27603) proposes a technique for drones to accurately deliver a cable-suspended payload to a target location.

Conventional drone logistics research has focused primarily on landing payloads "safely and slowly." However, in time-critical scenarios such as search and rescue or emergency medical delivery, the requirement is to deliver quickly, accurately, and precisely to the target location. The problem is that a payload hanging from a cable swings like a pendulum, and its dynamics are extraordinarily complex. Wind effects, cable tension fluctuations, and the drone's own inertia all intertwine in ways that make hitting a target with simple control laws extremely difficult.

The research team confronted this "high-dynamics release" problem head-on by using reinforcement learning (RL). The agent repeatedly undergoes trial and error within a simulation, receiving target coordinates, cable length, and flight speed as inputs, then autonomously learns "when and along what trajectory to fly while releasing the payload."

From Simulation to Real Hardware — Another Paper Breaks Through the "Sim-to-Real" Barrier

Published at the same time, "Support-Constrained RL Enables Real-World Policy Improvement without Real-World Experience" (arXiv:2606.27475) tackles the "Sim-to-Real" problem that is common across robotics research broadly.

When a policy learned in simulation is transferred directly to real hardware, degraded accuracy, behavioral fragility, and unexpected actions occur frequently — a challenge that has long plagued robotics researchers. This paper attempts to solve it through a method called "support-constrained RL." The idea is to constrain, during training, actions that fall outside the distribution (support) of real-world data, thereby greatly increasing the probability that improvements made in simulation will also hold on actual hardware.

If this approach becomes widespread, it could allow performance improvements in high-dynamics tasks like drone payload throwing while minimizing costly real-world testing.

Autonomous Docking in Space — The Application Domain of Robotics Expands

Furthermore, "Spacecraft Fiducial Marker for Autonomous Rendezvous, Proximity Operations, and Docking" (arXiv:2606.27566) proposes a new design for fiducial markers (visual position reference indicators) intended for autonomous docking in outer space. Whereas existing markers assume a single scale intended for ground use, this design's hallmark is a multi-scale approach suited to the extreme variations in distance and lighting conditions found in the space environment.

It is worth noting that across arXiv at the end of June 2026, approaches using RL and newly designed markers to address the shared challenge of "precise localization and dynamic control" are appearing in a concentrated burst — spanning everything from ground-based drone logistics to autonomous docking in space.

Conclusion — The Triangle of "Throw," "Constrain," and "See" Opens a New Future

What this collection of papers demonstrates is the reality that the frontier of robotics research is shifting from "executing movements" to "accurately completing tasks in high-dynamics situations." Reinforcement learning-based payload throwing control, a support-constrained approach to the Sim-to-Real problem, and space-ready marker design — viewed individually, these may seem like niche research topics, but combined they bring the dream of "a robot that delivers accurately in any environment" closer to reality. Practical deployment in application areas with direct implications for human life — search and rescue, medical delivery, space infrastructure maintenance — is eagerly awaited.

References: arxiv.org / arxiv.org / arxiv.org
ドローン強化学習ロボティクス物流arXiv

France Maneuvers to Exclude UK from EU's €5 Billion Scale-Up Fund——The 'Money Dividing Line' in European VC Geopolitics

By fin_suzuki

France has been revealed to be working to block UK participation in the EU's €5 billion scale-up fund. The rift within Europe over the allocation of policy capital has reached a critical juncture that will determine the future direction of tech investment.

Who Does "European Money" Belong To? — The Battle Over the Distribution of a €5 Billion Fund

A quiet ripple is spreading through the European tech world. Sifted's exclusive reporting has revealed that France is attempting to block the UK from participating in the "EU Scale-Up Fund," a vehicle totaling €5 billion (approximately ¥800 billion) that the EU is working to establish. France is said to be objecting to a proposed arrangement that would involve UK participation through British Business Bank, a UK government-backed financial institution, and EQT, a major Swedish-origin VC firm.

Post-Brexit UK-EU relations have outwardly performed a kind of "pragmatic cooperation," but when interests directly collide over the distribution of policy capital, the old fault lines resurface. From a chartered accountant's perspective, this is not merely a political dispute — it is a question of "sovereignty within the LP (limited partner) structure of a fund." At stake is who draws up the blueprint for who contributes capital and who enjoys the returns.

The Structure and Purpose of the EU Scale-Up Fund

This fund was designed to solve the problem of "capital export" — the situation in which European unicorn companies have no choice but to rely on US or Middle Eastern funds at the later stages of their growth. In Europe, while seed and Series A funding can be raised domestically, large rounds exceeding ¥10 billion have often depended on US firms like Tiger Global or SoftBank's Vision Fund.

Large funds such as EQT are among the candidates to manage the vehicle, but while EQT is Swedish-owned, it has deep roots in the UK market as well. France is seen as being wary of the fund effectively coming to operate under a management structure skewed toward the UK.

Consolidation Among European VCs and the Question of Capital Efficiency

At the same time, expectations are growing that mergers and acquisitions among European VC firms will accelerate. The founder of Italy's P101 has stated explicitly that "many more VC mergers will happen going forward," against a backdrop of mounting difficulty for smaller funds to survive on capital efficiency grounds.

A fund's fixed costs — personnel, due diligence, legal fees — decrease as a proportion of AUM (assets under management) as that AUM grows. The smaller the AUM, the more the standard 2% management fee fails to cover actual operating costs. In Europe, "mid-sized VCs" in the ¥20–50 billion range are under the greatest pressure, and are being forced to choose between scaling up or specializing.

A similar dynamic is playing out in the United States, where Bloomberg has reported that Valor Equity Partners is targeting the formation of a "Fund VII" at approximately $2.5 billion. The pursuit of economies of scale through fund enlargement has become a common trend on both sides of the Atlantic.

How Flawed "Policy Capital" Design Ripples Outward

The crux of the problem is that the EU Scale-Up Fund has a dual nature: it is simultaneously a "tool of public policy" and a "commercial VC investment vehicle." Policy objectives (the independence of Europe's tech industry) and commercial objectives (maximizing returns for LPs) do not always align.

Behind France's push to exclude the UK, one can also read a concern that "policy benefits will leak to UK companies." If portfolio companies concentrate their headquarters and jobs in London, the EU would lose its ability to account for the fund as an industrial policy instrument.

This is highly instructive as a debate over the design of a European-style "industrial policy finance" vehicle. Unlike subsidies or policy loans, an equity-style policy fund will not produce its intended policy effects unless the ownership stakes, governance arrangements, and exit routes are all designed with precision.

Conclusion — The "Geopolitical Cost" of EU Tech Investment

€5 billion is a large number, but it is by no means sufficient to meet the demands of the European tech ecosystem as a whole. If the UK and France come into conflict over those limited resources, it is ultimately the startups that will pay the price. Delays in fundraising rounds, deteriorating terms, and the emigration of talented founders to the United States — all of these will come back as "the cost of failed policy design."

The next focal point of European tech investment is shifting away from who will manage this fund, and toward the more fundamental question: "For whom was this fund designed?"

References: sifted.eu / sifted.eu / techcrunch.com
EUスケールアップ基金欧州VC政策英仏関係スタートアップ投資公共ファイナンス
Advertisement300 × 250