The World of Atoms, Not Bits⚛️…

1. Key Themes

Deep Tech & Physical World as the Next Wealth Creation Frontier

Capital is rotating from software into physical infrastructure. The article reports that "the next trillion-dollar companies will be built in the world of atoms, not bits," with robotics, energy, aerospace, advanced materials, and compute identified as the key sectors. Quantitatively, "the U.S. and Europe attracted $156.6B in deep tech funding, while new unicorns increasingly require longer timelines and larger capital commitments." France is a leading indicator of this shift: deep tech "reached a record $4.7B and tripling from 2019 levels," with "growth since 2020 [having] exceeded every major European market."

The VC Liquidity Crunch Is Reshaping Capital Markets

Traditional exit pathways have stalled, forcing the industry to adapt structurally. "Private company exits remain scarce, leaving a large backlog of venture-backed businesses waiting for liquidity events. As distributions slow, investors are increasingly using discounted secondary transactions to unlock capital." Private equity is compounding this: "longer holding periods and widening valuation gaps are making portfolio exits harder to execute." The structural shift is visible in public markets too — "PE-backed companies now significantly outnumber UK-listed firms as sponsor-to-sponsor transactions continue to rise."

AI Is Evolving From Prompts to Autonomous Orchestration

The nature of AI tooling is shifting from single-interaction prompting to self-managing workflow systems. Claude Code's new dynamic workflows show "how software can generate and manage its own execution logic, reducing the need for manually linked prompt chains." The building blocks — "parallel evaluation, verification loops, and ranking systems" — are described as "reusable" for complex tasks. Separately, a solo founder playbook frames this operationally: "founders manage context, systems, and decisions while agents handle much of the underlying work."

AI Infrastructure Commanding Extraordinary Valuations

The deal flow confirms that AI infrastructure is attracting capital at previously unseen valuations. Prometheus raised "$1.2B Series B at a $41B valuation to scale its AI infrastructure platform and accelerate deployment of artificial general engineering systems" — a striking valuation for a Series B. AlphaSense raised "$350M at a $7.5B valuation," and PhoenixAI raised "$80M Series B to scale its AI-native database infrastructure for autonomous agents." The pattern signals investor conviction that AI picks-and-shovels companies will capture outsized value.

Private Markets Replacing Public Markets as the Default Exit Path

The IPO market is no longer the default endgame for venture-backed companies. "PE-backed companies now significantly outnumber UK-listed firms as sponsor-to-sponsor transactions continue to rise." While "lower interest rates, listing reforms, and private share trading venues are helping revive deal activity," the structural preference has shifted toward staying private longer. This has downstream consequences for LP liquidity and fund-raising cycles.

2. Contrarian Perspectives

The AI investment opportunity is NOT narrowing to just infrastructure giants. The consensus view holds that only dominant foundation model companies and cloud infrastructure players are worth betting on at this stage. Sarah Guo directly challenges this: she "challenges the belief that only dominant infrastructure companies remain investable in the current market cycle. The argument centers on capabilities and advantages that cannot be easily replicated through scale or training data alone." This implies investable moats exist in application-layer or specialized AI businesses that incumbents cannot easily absorb.

The "world of atoms" represents a more durable opportunity than continued software investment. The prevailing VC playbook for the past two decades has been software-first. Private equity is now signaling a reversal: "firms are reallocating capital away from software toward sectors viewed as more resilient to technological disruption." Combined with the $156.6B flowing into deep tech, this suggests a multi-decade reallocation away from pure-software bets — a meaningful departure from consensus venture doctrine.

Finishing, not starting, is where value is actually created. Most startup culture lionizes launching and iterating quickly. Alfred Lin inverts this: "he argues that difficult final steps are where quality, differentiation, and outcomes become visible to others. Projects often fail not from poor starts but from never reaching the level of completion that compounds results." The implication for founders and investors is that execution depth — not speed of initiation — is the true differentiator.

3. Companies Identified

SpaceX Description: Aerospace and space technology company Why mentioned: Cited as a market milestone — "SpaceX surpasses TSMC to become the world's 6th most valuable company," with a 3% probability on Polymarket of becoming the largest company by year-end. Illustrates the "atoms not bits" thesis in action. Quote: "3% chance SpaceX is the largest company at year-end."

Prometheus Description: AI infrastructure platform Why mentioned: Largest deal in the issue — raised "$1.2B Series B at a $41B valuation to scale its AI infrastructure platform and accelerate deployment of artificial general engineering systems." Emblematic of the extreme valuations AI infrastructure companies are commanding.

Helion Description: Commercial fusion energy company Why mentioned: Raised "$465M Series G to advance commercial fusion energy and support deployment of its first power plant." A flagship example of the "world of atoms" investment thesis and sustained deep-tech capital commitments.

AlphaSense Description: Market intelligence and enterprise AI platform Why mentioned: Raised "$350M at a $7.5B valuation to expand its market intelligence and enterprise AI platform." Demonstrates the scale of AI application-layer valuations.

PostHog Description: Product analytics and developer tooling company Why mentioned: "Published a practical marketing manual for new founders" — cited as a practitioner resource for operators building repeatable growth before hiring dedicated marketing staff. Positioned as a trusted operational voice in the founder community.

GPS Renewables Description: Bioenergy and compressed biogas infrastructure company in India Why mentioned: Raised "$66.3M Series C to expand bioenergy infrastructure and compressed biogas production across India." Notable as a non-Western deep tech deal, signaling the global reach of the atoms-not-bits investment wave.

Theker Description: AI-native legal technology platform Why mentioned: Raised "$85M Series A to expand its AI-native legal technology platform and accelerate international growth." Represents the vertical AI application opportunity Sarah Guo alludes to.

PhoenixAI Description: AI-native database infrastructure for autonomous agents Why mentioned: Raised "$80M Series B to scale its AI-native database infrastructure for autonomous agents and enterprise applications." Signals investor demand for AI agent infrastructure as a distinct category.

Vanta Description: Security compliance automation platform Why mentioned: Newsletter sponsor — framed around the high-stakes risk of compliance failure: "The fastest way to lose an enterprise deal is to fail the security questionnaire."

4. People Identified

Alfred Lin Description: Partner at Sequoia Capital; early Zappos executive Why mentioned: Shared the insight that "finishing the last 10% creates most of the value," arguing that "projects often fail not from poor starts but from never reaching the level of completion that compounds results." Quote: "Difficult final steps are where quality, differentiation, and outcomes become visible to others."

Sarah Guo Description: Founder and General Partner at Conviction VC; former Greylock partner Why mentioned: Offered a contrarian view on AI investing — "she challenges the belief that only dominant infrastructure companies remain investable in the current market cycle," pointing to "capabilities and advantages that cannot be easily replicated through scale or training data alone."

Ruben Dominguez Description: Author of The VC Corner newsletter Why mentioned: Curator and author of the issue; also referenced for prior content including "The Claude Code system that replaces a 5-person team" and various founder resource guides.

5. Operating Insights

Treat AI agents as a functional team, not a tool — and architect accordingly. The solo founder playbook featured in the issue reframes how operators should think about AI: "founders manage context, systems, and decisions while agents handle much of the underlying work." The implication is that the strategic bottleneck shifts from labor capacity to workflow design — founders who invest early in agent orchestration infrastructure can scale operations without proportional headcount growth.

Build marketing systems before you need them — not after product-market fit. PostHog's marketing guide is explicitly "aimed at operators who have early customer demand but limited experience with distribution. It provides a starting framework for building repeatable growth efforts before hiring dedicated marketing talent." The operating insight: distribution infrastructure built early compounds; distribution assembled reactively always lags demand.

SOC 2 certification is a revenue issue, not just a compliance checkbox. Vanta's sponsorship message is framed as a commercial warning: "The fastest way to lose an enterprise deal is to fail the security questionnaire." For B2B founders targeting enterprise customers, SOC 2 certification is a sales prerequisite — procurement friction from missing compliance can kill deals that would otherwise close.

6. Overlooked Insights

France is emerging as a potential European deep tech compounder worth tracking early. Buried in the reports section, France's deep tech trajectory is striking: funding "reached a record $4.7B and tripling from 2019 levels," with "growth since 2020 [having] exceeded every major European market, placing France on a trajectory comparable to the U.S. expansion cycle." For investors building European exposure, France — not just the UK or Germany — may be the highest-beta deep tech market right now.

The secondary market is quietly becoming a primary liquidity mechanism — not a last resort. The article notes that "as distributions slow, investors are increasingly using discounted secondary transactions to unlock capital." This is typically framed as a distress mechanism, but the scale of the backlog suggests secondaries are becoming a routine portfolio management tool. For LPs and GPs, pricing discipline in secondary transactions — and the emergence of dedicated secondary funds — may be as strategically important as primary deployment decisions.