The Factory and the Frontier: Reading America's Manufacturing Boom Through Crypto's Infrastructure Narrative

By Mia Brown

Hook: The Signal That Arrived Already Laundered

The morning the U.S. manufacturing numbers hit the terminal, I was doing something embarrassingly unglamorous: reconciling a stale concentrated liquidity position that had quietly bled impermanent loss since October. That's what "digital asset fund manager" actually means in Tallinn most days—half my calendar is spreadsheet forensics and half is defusing someone's panic over a red candle. Then the headline blinked across my secondary screen: U.S. manufacturing hits fastest expansion pace since 2022. Within ninety minutes, the same number had been laundered through at least four crypto news desks, arriving at a conclusion I have watched this industry reach a dozen times since 2017: infrastructure is strengthening, therefore crypto will thrive.

The chain presented was seductive in precisely the way that reminds me of who I was in 2017, when I converted my entire student savings into Ethereum because the community narrative felt so coherent. Manufacturing strengthens. The grid expands. Data centers multiply. AI and crypto infrastructure benefit. Each step sounds logical until you realize that no one in that chain has actually quantified a single watt of new power, a single megawatt-hour of contracted energy, or a single dollar of data center capex tied to the PMI print. The ledger remembers what the market forgets—and the market, in this bull phase, appears to have forgotten that narrative coherence has never once been a substitute for structural analysis.

I am not here to tell you the manufacturing data is irrelevant. That would be intellectually lazy in the opposite direction. I am here to tell you that the story being told about it in crypto circles is dangerously incomplete, that the causal conduit from a purchasing managers' survey to a blockchain infrastructure boom runs through at least six failure points, and that the most likely market outcome of this "good news" is something the narrative's authors did not intend: higher rates for longer, a tighter liquidity envelope, and a crypto market that has to relearn what it means to trade in a world where the Fed is not your co-pilot.

Context: What the Headline Actually Contains

Let me anchor the substance before I deconstruct the metaphor. The underlying economic fact is real: U.S. manufacturing output and new orders have accelerated to their fastest pace since 2022, and the expansion is widely attributed to the Trump administration's industrial policy agenda. Tariffs designed to protect domestic producers. Reshoring incentives that have pushed firms from Southeast Asia back toward the Rust Belt and the Sun Belt. Energy deregulation intended to lower industrial power costs. "American industrial renaissance" is the phrase being used in policy circles, and it has genuinely moved the needle in certain sectors—semiconductor fabs, electric vehicle battery plants, and specialty chemicals have all announced meaningful capacity additions over the past eighteen months.

The source report, distributed through a crypto-native media outlet, drew the obvious connection for its audience: manufacturing growth could "enhance infrastructure" and thereby affect AI and crypto industries. The mechanism, as described, was vague to the point of being unfalsifiable. Enhanced infrastructure could mean more reliable power. More efficient supply chains. Cheaper land for data centers. But it could equally mean something the report did not mention: a scramble for the same electrons that mining rigs and GPU clusters need, with industrial buyers outbidding crypto miners in power purchase agreements.

From my perspective, having spent the 2024 cycle translating blockchain macro-trends into investment theses for over fifty institutional clients, what matters here is not whether the manufacturing expansion is real. It is. What matters is whether the crypto market's reflexive optimism about the data is priced on verifiable mechanics or on a vibes-based extrapolation. The first phase's technical analysis flagged this correctly: the article contains zero protocol details, zero code, zero auditable claims. It is a macro fast-news item dressed in the costume of a sector catalyst. The report cannot be technically verified because there is nothing technical in it to verify.

That is not a criticism of the journalist who wrote it. It is a warning about what happens when an industry-starved-for-good-news starts treating every macro print as confirmation of its own thesis.

Core: The Transmission Mechanism, Examined Link by Link

First Link: What a PMI Number Actually Measures

Before we can assess whether "manufacturing expansion → crypto infrastructure boom" is a valid inference, we need to be honest about what the underlying data measures. The ISM Manufacturing PMI is a diffusion index constructed from a survey of purchasing managers. It measures sentiment about the present and near future—new orders, production, employment, supplier deliveries, inventories. It is not a capital expenditure commitment. It is not a ground-breaking permit. It is not an interconnection queue approval. It is a mood ring worn by supply chain professionals, and while mood rings have been shown to have some predictive correlation with actual industrial activity, the correlation between PMI and the specific assets crypto cares about—grid capacity, data center availability, power prices at industrial load centers—is far weaker than the narrative requires.

I have audited mining operations where the difference between profitable and unprofitable was a single cent per kilowatt-hour. The margin dynamics in crypto infrastructure do not move on vibes. They move on power purchase agreements, on transformer lead times, on substation upgrade approvals that take thirty-six months in the best-case scenario, and on the willingness of landowners in west Texas to host a hundred-megawatt load facility in exchange for a lease payment. None of those variables are captured in a purchasing managers' survey. When I read the crypto coverage of this manufacturing print, I kept asking one question the coverage did not answer: show me the capital expenditure line, not the sentiment line.

The deeper issue is timing mismatch. The manufacturing expansion, if it persists, would take two to three years to translate into the kind of physical infrastructure that could plausibly lower costs for AI data centers or mining operations. Grid interconnection queues in the United States are already staggeringly long—the Lawrence Berkeley National Laboratory's latest queue study showed median wait times exceeding five years for new generation and storage projects. A manufacturing boom does not automatically shorten those queues. If anything, it lengthens them, because every new factory, every new semiconductor fab, every new battery plant is another entrant competing for the same constrained grid capacity.

The market is treating a PMI print as a forward indicator for crypto infrastructure. The more defensible reading is that it is a concurrent indicator for industrial emotions, with a lagged and ambiguous effect on the specific infrastructure crypto actually depends on.

Second Link: The Liquidity Contradiction at the Heart of the Trade

Here is where the narrative is not merely imprecise but actively self-contradictory. The primary macro variable that has determined crypto's fortunes over the past three cycles is not industrial production. It is liquidity. Stability is a myth; liquidity is the only truth. That is not a poetic flourish—it is a description of the mechanism by which risk assets get repriced when the global dollar funding envelope expands or contracts.

Manufacturing expansion at the fastest pace since 2022 is, from the Federal Reserve's perspective, evidence of economic resilience. Resilient economies do not get aggressive rate cuts. Resilient economies with sticky inflation components—and manufacturing expansions tend to lift input prices, wages, and transportation costs—get "higher for longer." The market's own reaction function, refined over the past eighteen months, is to interpret strong macro data as bearish for crypto because it pushes the terminal rate higher and delays quantitative easing. The report's framing, "manufacturing growth is good for AI and crypto," inverts the actual transmission channel that has governed crypto's liquidity-driven rallies since 2020.

Let me be precise about the two competing channels, because this is the analytical crux of the entire piece:

Channel A (the narrative's channel): Manufacturing expansion → infrastructure investment → data center and energy availability → lower costs for AI compute and crypto mining → sector growth.

Channel B (the market's actual historical channel): Manufacturing expansion → economic resilience → inflation persistence → Fed holds rates higher → global dollar liquidity tightens → risk assets, including crypto, face multiple compression.

Channel A is a real-economy story with a multi-year time horizon and a highly uncertain payoff. Channel B is a financial-market story with a quarterly time horizon and a well-documented historical track record. The crypto market, in its current euphoric phase, is choosing to lead with Channel A while ignoring that Channel B is the channel that has actually dictated drawdowns and recoveries in every cycle since I started trading in 2017. My own history is a cautionary tale here: in early 2018, I lost 90% of my student savings because I was trading the narrative channel (adoption, infrastructure, world computer) while the liquidity channel was rapidly closing. I did not understand that a macro tightening cycle could invalidate even the most beautiful technical story. I returned to computer science not because I wanted to abandon crypto, but because I needed to understand the difference between a protocol's promise and its market's liquidity constraints.

Third Link: The Energy Scramble — Manufacturing as Competitor, Not Benefactor

The most under-explored dimension of the "infrastructure boost" narrative is the competition dynamic. Crypto mining, AI data centers, and manufacturing all consume the same grid. They all want baseload power, they all want short interconnection timelines, and they all want low wholesale electricity prices. When manufacturing expands, it does not just create more energy infrastructure through the slow accumulation of tax revenues and utility capital plans. It also, immediately and directly, competes for the existing energy infrastructure.

Consider who wins in a power auction between a semiconductor fab that employs five thousand people and represents $20 billion of capital investment and a bitcoin mining operation that can move its containers to another jurisdiction within weeks. The factory wins. The factory always wins. Manufacturing jobs carry political weight; mining operations are often viewed as parasitic loads that create few jobs and arbitrary energy demand. When the manufacturing boom accelerates, utilities and regulators become less accommodating to crypto mining loads, not more. The political economy of energy allocation moves against us precisely when the macro narrative claims it is moving in our favor.

This is not speculation about the future. We have already seen the pattern in Texas, where winter storm demands triggered curtailments of mining loads, and in multiple jurisdictions where data center announcements prompted moratoria on new high-load connections. Manufacturing expansion intensifies this dynamic. Every new factory announcement is another reason for a utility to reserve its remaining transmission capacity for "productive" industrial uses. The narrative assumes crypto infrastructure will ride the manufacturing wave. The more likely outcome is that crypto infrastructure gets caught in its undertow.

Fourth Link: Mining Centralization — The Infrastructure Boom's Quiet Concentration Effect

Now let me add my specific technical contention, one I have been developing since the fourth Bitcoin halving compressed miner revenue. The manufacturing infrastructure story, even if it fully materialized, would not democratize mining. It would accelerate its concentration.

Here is why. After the fourth halving, block subsidies dropped from 6.25 to 3.125 BTC, and miners who had built operations on the assumption of higher revenues found themselves in a brutal margin squeeze. Hash price—the expected revenue per terahash per day—fell to levels that made marginal operators unprofitable. What I observed, auditing mining economics through that consolidation, was a market bifurcating into two tiers. Tier one: large, well-capitalized operators with contracted power at sub-$0.05/kWh, often with direct relationships with utilities and even grid participation agreements. Tier two: everyone else, scraping by on spot power, unable to secure long-term PPAs, and increasingly dependent on pooled hash services or exit.

A manufacturing-driven infrastructure boom—new substations, expanded grid capacity, lower electricity costs from deregulation—would be captured by tier-one operators who have the balance sheet and the legal teams to execute power purchase agreements at scale. The small miner, the home operator, the community micro-miner: none of them get access to newly available industrial power because they cannot commit to the load profiles utility companies demand. The infrastructure story is a story of concentration, not democratization. By my estimates, and I have run this analysis repeatedly across the past twelve months, the top three mining pools now account for a substantial majority of total network hash. The decentralization consensus that Bitcoin was supposed to embody is hollowing out precisely as its physical infrastructure improves. We are building better roads, and the roads all lead to the same three toll booths.

This is uncomfortable to say in a bull market where any caution reads as bearish. But I have run this fund through enough cycles to distrust comfort. The fourth halving did not produce a decentralized mining ecosystem. It produced a professionalization wave that consolidated power, and the manufacturing expansion, to the extent it touches energy infrastructure at all, will pour fuel on that consolidation fire. If you are bullish on this narrative because you hope it helps small miners survive, you are reading a fairy tale that my spreadsheets cannot reproduce.

Fifth Link: DePIN and AI Compute — Where the Narrative Overpromises

Let me turn to the other beneficiary the macro story claims to serve: decentralized physical infrastructure networks. As someone who led a pilot program in 2025 connecting AI researchers with GPU providers through a decentralized compute marketplace, I have a grounded view of where DePIN actually stands. The network worked, in the narrow technical sense. We verified compute integrity using consensus-based attestation, we built escrow logic for fair payment, and we demonstrated that blockchain-based coordination could reduce settlement friction. But the pilots also revealed a chasm between the DePIN promise and the enterprise reality.

Enterprise AI labs do not primarily care about whether compute is decentralized. They care about latency, about data residency, about verifiable trust, about uptime guarantees, and about whether a contract can be enforced in a jurisdiction they recognize. A manufacturing boom in the United States does nothing to solve these frictions. It does not make an enterprise client trust a decentralized GPU market more than a hyperscaler's cloud contract. It does not resolve the legal ambiguity of enforcing a smart contract when a GPU provider in South America fails to deliver. And critically, it does not address the demand generation problem: enterprise AI workloads are overwhelmingly routed to centralized clouds because the procurement path is simpler, and a macro infrastructure narrative does not change procurement behavior.

The deeper problem is the one I keep circling in my own analysis: most of these infrastructure projects are solving a supply problem when the binding constraint is demand. The industry loves to talk about compute scarcity and GPU availability. The actual limiting factors are trust verification, integration overhead, and corporate procurement's risk aversion. A cheaper kilowatt-hour or a new data center building does not produce an enterprise client willing to trust code over contract. Code is law, but trust is the currency—and trust cannot be manufactured by PMI data. It has to be earned through reliability, through compliance, through the unglamorous work of making decentralized infrastructure boring enough for legal review.

When I hear the crypto market claim "U.S. manufacturing expansion will boost AI and crypto infrastructure," I translate that into: "A macro data point with a multi-year, low-probability transmission chain to our sector is being used as a narrative justification for buying infrastructure tokens." And I note that the similar infrastructure narratives of 2021—the Web3 infrastructure glut, the "pick-and-shovel" thesis—produced a remarkable number of networks with beautiful architectures and negligible usage. We built the cathedral before the saints arrived, and then discovered the saints had decided to worship in a centralized cloud instead.

Sixth Link: What My Institutional Clients Actually Asked

After the 2024 ETF approvals, I spent the better part of a year in rooms with traditional finance clients—wealth managers, fund-of-fund analysts, sovereign wealth advisors—translating blockchain's macro dynamics into language their risk committees could digest. I wrote a whitepaper on liquidity flows in the post-ETF era that tracked correlations between ETF inflows and on-chain activity, and I learned something that reshaped my own framework: institutional clients do not care about PMI data in the way crypto natives think they do.

They care about correlation regimes. Is Bitcoin a risk asset correlated to Nasdaq, or a hedge asset correlated to gold? They care about custody infrastructure, about insurance, about redemption liquidity. They care about the Treasury General Account balance, about the pace of quantitative tightening, about overnight reverse repo usage. The macro data that actually moves their allocation decisions is liquidity data, not manufacturing data. When a manufacturing print hits, the question my clients ask is not "Does this help AI compute?" It is "Does this change the Fed's reaction function?" And the answer to that question, for nearly every strong-data surprise since 2023, has been: it pushes rate cuts further out, and that is bearish for risk assets that benefited from easy liquidity.

This is the analytical chasm between the crypto-native interpretation of this story and the institutional interpretation. The crypto-native reading treats manufacturing strength as a bullish infrastructure signal. The institutional reading treats manufacturing strength as a bearish liquidity signal. The institutional reading has the better historical track record. I have the scars from 2018, the spreadsheets from 2022, and the client calls from 2024 to confirm it.

Contrarian: The Decoupling Thesis and the Narrative Saturation Signal

The Argument for Decoupling

Let me steelman the contrarian case, because it is not the one crypto media is running, and it deserves a hearing. There is a plausible scenario in which crypto decouples from the traditional manufacturing cycle entirely—not because infrastructure improves, but because the monetary conditions that drive crypto become divorced from industrial outcomes.

The U.S. fiscal trajectory is the elephant in every macro conversation. The manufacturing expansion is being purchased with tariffs, with subsidies, with industrial policy that expands the federal deficit at precisely the moment interest costs are compounding. If the manufacturing boom accelerates while fiscal deficits remain structurally elevated, the long-term dollar purchasing power trajectory worsens even as the short-term economy looks resilient. A manufacturing renaissance financed by deficit spending is a currency debasement story wearing a hard-hat costume. For Bitcoin specifically, which has increasingly traded on monetary debasement and geopolitical reserve dynamics rather than on industrial production, the manufacturing boom could be miscalibrated: the more the government spends to fund the renaissance, the more attractive the hard-asset hedge becomes.

This is the decoupling thesis. It argues that Bitcoin's price is a function of monetary trust, not industrial output. Under this thesis, a PMI print is almost noise. What matters is the size of the deficit, the direction of central bank policy, and the accumulation behavior of sovereign and institutional entities. The manufacturing expansion is bullish for AI and crypto only if it increases the fiscal burden in ways that undermine confidence in the dollar's future purchasing power. If I were being maximally contrarian, I would argue that the manufacturing boom is actually a powerful accelerant for the debasement narrative: trillions of dollars of industrial subsidies, chained to reshoring projects that upended global supply chains, with tariff revenues entirely insufficient to pay for the reconstruction they imply. The fiscal hole at the center of the "industrial renaissance" is the most quietly Bitcoin-bullish thing the administration has done.

I cannot fully dismiss this thesis. In the 2025 AI-crypto convergence work I did, I watched the market treat compute narratives as geopolitics: control over AI infrastructure is becoming a national security question, and national security spending is, historically, the most potent currency debasement engine in existence. A manufacturing expansion linked to national security and AI dominance is, from a hard-asset perspective, a dollar-weakness machine. The crypto market's infrastructure narrative might be aiming at the wrong target. The real benefit is not cheaper energy or new data centers. It is the fiscal deterioration that flies in under the cover of economic patriotism.

The Signal That We Are Narrative-Searching

Here is my less metaphysical concern. When a macro data point is picked up by crypto media and stretched to fit the "AI + crypto" story, it tells me less about the data and more about the market's emotional state. We are in a bull market that has already consumed multiple narratives: ETF inflows, Bitcoin as a strategic reserve, the AI-crypto convergence, DePIN, meme cycles, and now manufacturing-led infrastructure. The consumption rate of narratives is accelerating. Each new story is treated as the next leg of the bull run, and each story is exhausted faster than the last.

In my experience—and I have ridden this emotional rollercoaster enough times, from the ICO frenzy of 2017 through the DeFi Summer of 2020 and the post-ETF euphoria of 2024—the willingness of a market to import macro data into sector narratives is a late-cycle signal. Early-cycle markets trade on fundamentals. Mid-cycle markets trade on emerging narratives. Late-cycle markets import anything into the narrative machine, including data that should arguably be bearish. The manufacturing story is being converted into a crypto bull case by inversion: a data point that nominally argues for tighter liquidity and higher rates is being celebrated as a physical infrastructure boon. That inversion is a sign that the narrative engine is operating on fumes.

I do not mean to time the top. That is not my job as a fund manager, and I have learned that macro timing is a fool's game. But I have also lived through enough cycles to recognize the structural pattern: when every piece of news, regardless of its direction, is interpreted as bullish, the market is telling us it is not discriminating between information and narrative. And the information-to-narrative ratio is one of the most reliable (if under-discussed) sentiment indicators I have ever found. The ledger remembers what the market forgets—and the market is currently forgetting the difference between a purchasing manager's survey and a capital expenditure plan.

The Enemy of the Good Narrative

There is also a micro-level contrarian point that gets lost in macro discussions. Even if the manufacturing expansion does, against my expectations, translate into improved energy and data center infrastructure for crypto within a two-year horizon, the market's current pricing has front-run that outcome by at least eighteen months. Infrastructure tokens, DePIN networks, AI compute projects—many of them are trading at valuations that assume the physical buildout is already complete. I audited the token economics of several DePIN projects in late 2025, and I found the same pattern I found in DeFi during the 2021 bull market: the token's market capitalization implied astronomical future usage, while the underlying network's actual utilization rates remained a rounding error.

I have a standing belief, based on my experience with liquidity mining programs during DeFi Summer, that most incentive-driven networks are subsidizing their own metrics. Stop the incentives and the real users vanish. The infrastructure narrative creates a version of the same problem: it subsidizes expectation. It tells investors that a physical buildout is coming, and that the buildout will validate their token's valuation. But physical buildouts do not care about token valuations. They follow permitting timelines, construction schedules, and utility commission approvals. The narrative will be exhausted long before the concrete is poured.

The Risk Matrix I Cannot Ignore

Let me lay out, in the spirit of the risk discipline I built during the 2022 bear market, the specific risks this macro story carries. When my fund faced a 60% drawdown in 2022, I organized daily resilience circles with my team and investors, and the conversations were not about price recoveries. They were about identifying which narratives were real and which were self-soothing. The same discipline applies here.

The first risk is narrative over-simplification: treating a PMI print as a direct causal link to crypto infrastructure when the actual transmission chain is long, weak, and subject to multiple failure points. The probability of this narrative being wrong in its strong form ("manufacturing expansion substantially boosts crypto infrastructure within a relevant horizon") is high. The impact of this being wrong is moderate: investors allocate to infrastructure narratives, the buildout does not materialize, and the tokens underperform.

The second risk is the liquidity counter-signal: if manufacturing strength persists, the Federal Reserve's path to rate cuts narrows, and crypto's liquidity tailwind fades. This risk is the most under-discussed because it inverts the story crypto media is telling. A strong economy is, for risk assets, a double-edged sword flagged only as a single edge. The probability is moderate; the impact is potentially significant, because a multiple compression across the crypto complex would punish every project that priced in easy liquidity.

The third risk is policy change: the Trump administration's industrial policies are high-conviction but low-continuity. A midterm election or a subsequent administration can reverse tariffs and incentives faster than the infrastructure they funded gets built. Policy-driven narratives are inherently fragile because policy is inherently reversible. The probability is high; the impact is moderate.

The fourth risk is narrative exhaustion: the market's willingness to import macro data into bullish crypto frameworks is itself a signal of narrative saturation. If the manufacturing story becomes the main bull case and the data then disappoints—a PMI miss, a supply chain reversal, a tariff negotiation collapse—the narrative deflates quickly precisely because it was already over-extended. The probability is moderate; the impact is moderate.

What I Would Actually Watch

If I were advising an investor who wanted to position for the macro-infrastructure story without getting caught in the narrative trap, I would redirect attention to a different set of indicators entirely. Not the PMI. Not the manufacturing headlines. The variables that actually transmit macro reality into crypto infrastructure outcomes.

First, industrial electricity prices at major load centers. Electricity is the single largest variable cost for both mining and AI data centers. Track the wholesale price trends in ERCOT, PJM, and CAISO. If the manufacturing boom is genuinely creating infrastructure benefits, we should see power prices stabilize or decline relative to grid average. If the competition dynamic I described is winning, we should see industrial power costs rise, squeezing crypto infrastructure margins. The data will tell us, long before any narrative, whether manufacturing is a competitor or a benefactor.

Second, grid interconnection queue data. The Lawrence Berkeley National Laboratory publishes detailed interconnection queue statistics. The key metric is not the total MW waiting in line; it is the median queue duration, and whether data centers and high-load facilities are being prioritized or deprioritized relative to manufacturing loads. If the manufacturing expansion lengthens queues for high-load facilities, that is a definitive negative for the infrastructure narrative. If grid operators run separate fast-track channels for data centers and AI loads, the narrative gains support.

Third, hash price and mining pool concentration. Hash price tells us whether the marginal miner is making money; pool concentration tells us whether the decentralization consensus is holding. I have been watching the post-halving consolidation for over a year, and my read is that the infrastructure boom accelerates concentration rather than reversing it. Set alarms not for bitcoin's price but for the Herfindahl index of mining pool concentration. A rising concentration index tells you more about the real infrastructure trajectory than any PMI print ever will.

Fourth, enterprise demand for decentralized compute. The pilot I ran in 2025 involved three major AI labs. The question I get most often from that work is whether enterprise trust in decentralized compute is growing. The honest answer: slowly, and only where the alternative (centralized cloud) is unavailable due to regulatory or geopolitical constraints. Watch the procurement behavior of AI firms, not their conference keynotes. Watch whether DePIN contracts are being signed on the strength of verifiable compute, or on the strength of a token incentive. The former is structural; the latter is the yield-farming subsidy pattern I know too well.

Takeaway: Positioning for the Cycle, Not the Narrative

I have had to make peace with the reality that my job, as I see it, is not to be maximally optimistic or maximally cautious. It is to be useful. And the most useful thing I can say about the manufacturing-infrastructure narrative is this: position for the liquidity cycle, and let the infrastructure buildout be a long-duration option rather than the core thesis.

We have lived through enough market phases to know that bull markets always feel permanent while they are running. The phrase "surviving the winter makes the spring inevitable" got my team through 2022, when our resilience circles taught me that the most important infrastructure in crypto is not the fabs or the grid or the fiber. It is the community of people who keep showing up, keep building, and keep honest when the narrative engine runs hot. Community is the ultimate infrastructure layer, and in bull markets the community is best served not by cheerleading but by reminding itself that data is not destiny.

The manufacturing expansion is real. Its transmission to crypto infrastructure is, at best, unproven and multi-year, and at worst, counterproductive. The market's decision to frame it as bullish tells me more about where we are in the emotional cycle than about where the economy is going. From the frontier to the foundation, we are building something that will take a decade to calibrate. The question that matters is not whether a PMI print helps a DePIN token. The question is whether we are building the foundation at all, or merely selling each other pictures of the cathedral.

I still believe the spring is inevitable. I just need to be confident we can survive the season of overpromises first. And I have learned, from the ledger and from the markets, that the surest way through that season is to trust the data that compounds slowly: the kilowatt-hours actually delivered, the compute actually verified, the users who stay when the incentives dry up. Everything else is a narrative waiting to be corrected.

— Mia Brown, Tallinn