There's a sentence you will never hear on an earnings call: "Our valuation is supported entirely by investor faith."
It came this week from an unnamed Chief Information Officer, relayed through Crypto Briefing, and it carries the quiet violence of an open secret finally spoken by someone with procurement power. The CIO warned that the AI rally relies on investor faith — not realized returns, not validated productivity, not the long-promised wave of autonomous agents that everyone insists is six months away. Faith.
Let that land.
The timing is brutal. We are two years into one of the most aggressive capital-expenditure cycles in the history of industrial technology. Across 2024 and 2025, the hyperscalers — Microsoft, Amazon, Alphabet, Meta — committed hundreds of billions of dollars to AI infrastructure. Nvidia's data center business reshaped the stock market's center of gravity. And yet, at the enterprise level, the deployment reality remains murky. Survey after survey through 2025 — from Gartner, from IDC, from a dozen consulting white papers — keeps showing that a large share of generative AI projects are stuck in pilot-to-production purgatory. They are demos. They are sandboxes. They are "we're exploring the technology" lines on earnings calls.
I know this narrative. I have lived its crypto twin. Buenos Aires, 2017: I am 23 years old, running three Telegram communities for ICO-era Ethereum projects, watching intelligent adults treat whitepaper roadmaps as scripture. The airport had just opened a new terminal; the city was plastered with crypto billboards, and every coworking space smelled of mate, nervous sweat, and unearned conviction. The charts were beautiful. The data underneath was sand. And the name of the fuel that kept that entire market aloft was exactly the word this CIO used: faith.
Let's be honest about the report's information content, because the gap between the facts and the signal is where the story lives. The original piece offers one unnamed CIO, one warning, and roughly zero hard data. No company name. No industry vertical. No budget figure, no revenue multiple, no transaction flow. If I treated this as a dataset, I would flag it as a single respondent with a prior — statistically meaningless.
But single respondents matter when they sit at a specific point in the capital flow. A CIO is not a blogger. A CIO is not a sell-side analyst with a target price. A CIO is the person who sits across from the CFO and argues for budget allocation. They are the institutional demand function made flesh. When an anonymous CIO starts murmuring that the emperor has no clothes, it doesn't matter whether he comes from a Fortune 500 retailer or a mid-market bank. What matters is that the thought now lives inside someone whose job is converting AI promises into purchase orders.
And the messenger has context. Crypto Briefing is not Barron's. Its readership is crypto investors — a community with its own tortured relationship to faith-based valuation. The same week a CIO questions AI's returns, that media ecosystem will quietly suggest that capital is about to rotate back into digital assets. I will dismantle that fantasy in a moment. For now, understand the stage: the AI bull case relies on enterprise budget growth, and the person signaling doubt is exactly the person whose signature unlocks that budget.
Add the macro texture. Public market data from the 2024-2025 window showed cloud and AI capital expenditures growing at double-digit rates, quarter after quarter. At the same time, an uncomfortable share of enterprise AI adoption remained experimental. The gap between the capital going in and the measurable output returning is the void where faith thrives. And this week, an anonymous CIO pointed directly at that void.
There is also a measurement-bias problem the article itself embodies. The report gives us one side of the ledger — a warning — and zero counter-evidence. No rebuttal from an AI vendor. No case study of a deployment that actually worked. No data on the productivity gains that are real. That is not journalism; it is signal extraction. The word "faith" is doing heavy emotional labor, framing a technology market as a religious cult rather than a messy, early-stage investment cycle. I say this as someone who spent 2022 auditing the corpses of decentralized finance protocols: I am allergic to both hype and its mirror image, the smugly one-sided warning.
The Architecture of a Faith-Based Market
You don't need fraud for a faith market. You only need a feedback loop where price action validates narrative, and narrative validates price action.
Here is where my data science training kicks in. In statistical terms, a faith market is a prediction system with a severe overfitting problem: investors are fitting a complex curve to a single, noisy variable — storytelling — while ignoring the validation set. For years, the "validation set" in AI was the capital-expenditure line. Believers pointed to spending as proof of conviction, and spending responded to belief by increasing. That is circular. A system that treats its own input as its output will eventually confuse momentum with truth.
Add the crowding effect. When a narrative dominates for long enough, index funds and momentum strategies mechanically absorb the same names. The marginal buyer stops being an analyst who did homework and becomes a dashboard that rebalances quarterly. That is how you get a market structure where a single doubt can cascade: there is no one left on the other side doing independent thinking. The CIO's warning matters not because it is informative but because it is the kind of thought that, repeated enough times, becomes the consensus — and then becomes the trade.
I wrote about this in 2022, in my Ethics of Code series, after auditing the smart contracts of failed protocols. The pattern was everywhere: centralized decision-making hiding inside decentralized narratives. Multisig keys concentrated in the same founding team. Governance token distributions that made monarchies look egalitarian. What I learned is that beautiful charts are not truth, and neither are large numbers. The 2017 token-distribution analysis that formed my worldview showed roughly 80% of value flowing to early insiders — invisible in the aggregate growth charts everyone was sharing with messianic glee.
The AI market is not identical to that carnage. There is real revenue. Nvidia's earnings growth has been extraordinary, and enterprise AI adoption is genuinely real in pockets. But the broad structure — enormous multiple expansion across a long tail of companies with minimal verifiable differentiation in AI profit — carries the fingerprint of belief. When a market runs on faith, the first candle to go out isn't the worst company. It's the fastest narrative.
The Enterprise Validator
Let me use a blockchain lens, because it clarifies the stakes.
In proof-of-stake networks, validators lock up capital to secure the chain. If they lose confidence, they exit, the pool shrinks, and the network's security budget erodes. The system does not fail instantly. It degrades.
CIOs are the validators of the corporate AI economy. They lock up budget — GPUs, API credits, implementation partners, internal data teams — and in exchange, the broader market treats their commitments as confirmation of AI's value. When a CIO tells Gartner that AI is their top priority, that is a block confirmed. When a CIO starts telling anonymous journalists that the returns may never materialize, that is a validator going offline.
And here is the ugly part of the analogy: in crypto, when validators exit, the protocol at least knows the numbers. The chain's security becomes objectively measurable. In corporate AI, there is no public record. The CIO's doubt is a private belief, whispered to a journalist, with no dashboard and no on-chain metric to corroborate it. That is precisely the opacity that makes faith necessary. And faith is precisely what the CIO is now questioning.
I spent DeFi Summer in 2020 organizing weekly deep-dive sessions on impermanent loss and yield mechanics for more than 5,000 participants. We translated complex math into relatable analogies because understanding was our defense against blind conviction. The AI market lacks that shared language. When an investor tries to determine whether an AI company will grow into its multiple, they cannot audit the training compute, the unit economics, or the retention data. They are asked to believe.
The tracking signals are not secret; they are just ignored during bull markets. Watch the hyperscaler earnings calls for the phrase "AI ROI" — when executives stop talking about capability and start defending payback periods, the faith discount has begun. Watch the down rounds in AI startups; a single visible valuation cut can do more damage than a hundred anonymous CIO warnings. And watch the Gartner and IDC enterprise IT spending surveys, the closest thing this industry has to an on-chain validator dashboard. When the word "pilot" starts appearing less often than the word "production," we will know the chain confirmed a block.
What AI Could Learn From DeFi
There is a painful irony here. The industry built to replace faith with proof — blockchain — has spent years being mocked as the faith-based market par excellence. And now the AI industry, the technology world's darling, is demonstrating exactly why verification matters.
DeFi is far from perfect. My 2022 audits showed me the game: protocol teams copying code they did not understand, governance systems engineered to consolidate power, yield farms that were pyramid structures with extra steps. But despite all of it, DeFi offered something AI investors will never get from an earnings call: the ability to look. I can check Uniswap's liquidity pools. I can trace a suspicious transaction. I can evaluate a smart contract's access controls. Transparency is not a solution by itself. But it is the precondition for every solution.
The AI industry's measurement infrastructure barely exists. Not for compute efficiency. Not for model quality. Not for enterprise return on investment. Model cards remain inconsistent. Benchmarks get benchmarketed into irrelevance. And "AI ROI" is whatever the vendor's marketing deck says it is. When that is true, the market's only check on narrative is a human being like the anonymous CIO — and his warning is a crack in the system's only defense.
What would a public ledger for AI look like? That is the question I have spent the past year chasing with Verifiable Minds, the project I founded to explore zero-knowledge proofs for AI agent identity. The premise is simple: in a world flooded with synthetic content and opaque model behavior, we need cryptographically verifiable proof of who — or what — is acting. Verifiable inference. Zero-knowledge model verification. On-chain attribution of AI revenue. These are not PowerPoint slides. They are the minimum infrastructure for a market that does not require faith.
I prototype three zero-knowledge proof systems in six months, and I can tell you the technology is not the bottleneck. The bottleneck is incentive. Nobody in the AI supply chain wants to publish their unit economics. Vendors do not want you to verify their benchmarks. Enterprises do not want to admit their pilots are failing. The result is a market where the CIO's faith is the only oracle. And oracles, in both crypto and finance, are famous for being manipulable.
The Mirror: Crypto's Own Faith Problem
Here is where I turn the lens on myself, because crypto circles are already circulating this story like it validates every Bitcoin maximalist sermon ever preached. It does not.
Crypto's faith problem is not historical. It is present tense. I founded LatinWeb3 Arts in 2021 with 150 emerging Latin American artists and watched "community" become the most abused noun in our industry. We ran a DAO-governed grant fund that was genuinely beautiful in intention and administrativel-ocliatic in execution. I watched traders enter yield positions without understanding where the returns came from — and I watched the pile collapse in 2022. We were not the adults in the room. We were the warning label on a different bottle.
Nor is there any guarantee that AI's loss becomes crypto's gain. The rotation narrative is the waterbed fallacy. When institutional investors lose faith in growth assets, they do not rotate into higher-beta, less-regulated speculative assets. They rotate to cash and Treasuries. In 2022, the S&P fell roughly 20 percent while Bitcoin fell roughly 65 percent. Faith assets are correlated. When the faith leaves, every candle flickers. Crypto cheering for an AI correction is a mouse celebrating a house fire.
This is why the anonymous CIO matters to us. His doubt is our doubt, refracted through a different technology. If the enterprise market is finally demanding verifiable returns — not PowerPoints, not pilots, not faith — then that standard applies to us first. We built our entire pitch on replacing trust with proof. We should be the most audited, most verified, most evidence-driven industry on Earth. We are not.
The Case for the Other Side
Now let me argue against my own thesis, because a faith-market analysis only earns credibility if it steelmans the bull.
The anonymous CIO could simply be wrong. Not dishonest. Wrong. "AI is overhyped" has been a job description for short-sellers since 2023, and the trade has been painful. Nvidia's data center revenue did not just meet expectations across this cycle; it repeatedly annihilated them. Microsoft's AI services line grew into a genuine business unit. OpenAI's revenue trajectory became the fastest in software history. In a winner-take-most market, the existence of thousands of failed pilots is entirely compatible with enormous aggregate value creation. Amazon does not need ten thousand successful AI projects. It needs two.
The anonymity of the source should also temper our certainty. A CIO at a mid-market insurance company still running legacy systems from the 1990s is not the same informational authority as a CIO at a technology-forward enterprise. Every great transition — the internet, cloud, mobile — had a phase where traditional IT leaders dismissed the new paradigm as overvalued. Some were early. Some were wrong. Most were both.
And there is a missing variable in the bear case: AI's deflationary economics. Inference costs have been falling with remarkable speed. A project that fails a return-on-investment test today can become obviously worth it eighteen months from now. The enterprise ROI that the CIO says is missing might simply be a lagging indicator. The same argument applied to cloud computing in 2012: it was unprofitable for most enterprises that year, and utterly transformative by 2016.
One final point in the bulls' favor. Faith is not always irrational. In early-stage technology, faith is the input that makes evidence possible. Without belief in outcomes that had not yet arrived, the internet would never have been built. The CIO's warning says the market's faith outpaced its evidence. That can be true. But the question is not whether the rally runs on faith. Every revolution does. The question is whether the faithful are accumulating real value while they wait. Some — not all — most certainly are.
Faith markets, remember, can run for years. The dot-com rally survived multiple skeptical CIOs and skeptical Fed chairs before reality arrived. The useful stance is not shorting the faith; it is respecting the duration and sizing for the correction that eventually corrects every narrative that outgrew its evidence.
Positioning In the Chop
For those of us navigating this sideways, consolidated market, the practical takeaway is positional, not emotional. The CIO's warning is not a sell signal. It is a signal to demand more from everything we hold. In a market where narratives swing violently and nothing moves in a clear direction, the assets that protect you are the ones with verifiable fundamentals: revenue you can trace, metrics you can audit, community activity you can witness. The projects built on pure storytelling are the ones that bleed when the next faith crisis hits.
In a sideways market, chop is for positioning. That means rotating out of narrative-heavy positions and into assets with measurable traction. It means asking the same question the CIO asked of AI, but aimed at our own portfolios: if faith disappeared tomorrow, what would still be standing?
Short term, watch the hyperscaler capex guidance on the next earnings cycle. Medium term, watch the Gartner CIO surveys and the frequency of AI startup down rounds. Long term, watch whether AI actually transforms a single traditional industry — health, manufacturing, finance — in a way that shows up in audited financial statements. None of that is exciting. All of it is real.
Takeaway: Faith Is a Terrible Oracle
Faith is a terrible oracle. The CIO's warning is a reminder that any market running on consensus belief is a market running on borrowed time. For AI, the fix is not more aggressive optimism. It is measurement infrastructure — standardized ROI reporting, auditable efficiency, verifiable outcomes. It is building the transparency layer that every mature market eventually demands.
For crypto, the lesson is harder. We do not get to smirk. We do not get to hope for AI's collapse. We built our entire pitch on replacing faith with verification. The only way to win that argument is to be the most inspected, most rigorous, most quantifiable industry in technology.
We don't need more belief in our corners. We need proof. Freedom isn't found in conviction; it is found in verification — the ability to check, audit, and confirm without trusting anyone. That was the promise of 2017. It is still the promise of today.
And if we keep it, the next cycle will be a future built by our shared vision — and backed by something far stronger than faith.
