I read two articles this week about why AI has not replaced software engineers, and why it is unlikely to. Both resonated with me deeply. I am grateful to the authors, Arvind Narayanan and Sayash Kapoor of Princeton, in one, Michael Lawrence in the other. Reading them side by side opened a wider frame. Sources at the foot.

Narayanan and Kapoor look at the corporate layoffs and find that most of those announced as AI driven are nothing of the kind. The real drivers are financial pressure, pandemic over hiring, activist investors and bloated management layers. AI is the cover story that plays better with stakeholders than the truth.

Lawrence argues that AI is not replacing developers. It is deforming them. Judgement and muscle memory erode quietly, while engineers are told they are more productive than ever.

I wrestle with versions of these arguments almost daily. Companies are cutting people for reasons that have little to do with AI. At the same time, the genuine effect of AI on the people they retain, on the formation of judgement and the maintenance of expertise, is going almost entirely unnoticed. The media focuses on the wrong problem.

I see the same hypocrisy, or to use a better word, the same froth on every frontier. AI. Digital currencies. Quantum.

The froth around AI

Narayanan and Kapoor reach back to Fred Brooks, who in 1986 wrote No Silver Bullet. Brooks separated the complexity of software into two kinds. Accidental complexity is the friction of tools and languages. Essential complexity is the work of specifying what software should actually do, understanding the consequences of it, and being accountable for what it does in the world. Tools have reduced accidental complexity, repeatedly, for forty years. None has ever touched the essential kind. Brooks said it would not happen. It has not.

The froth around AI runs in two voices. The cover story, that AI is driving the layoffs, plays better with stakeholders than the truth. The replacement headline, that AI will replace developers, radiologists, knowledge workers, sells better than the data. Both leave the essential complexity exactly where Brooks left it. Human.

The work the industry should be doing is developing human capital that sits above AI, not beneath it. Training juniors to think, not just to prompt. Putting senior engineers on the questions AI cannot answer: what should be built, what the consequences will be, who is accountable when the system behaves in the world. Building the verification, oversight, and review muscle that compounds with AI rather than competing with it. And being honest, in the meantime, that the workforce question is one of redirection and elevation, not of replacement.

The froth around Digital Currencies

In digital currencies, the froth has names. Web3 as the next internet. Tokenise everything. NFTs as a new asset class. Stablecoins as a parallel monetary system. A decade of breakthroughs announced and re-announced, most of which amounted to a reduction in accidental complexity. The blockchain, in many of its forms, is a different way of writing to a ledger. Useful, occasionally novel, rarely revolutionary on its own.

The essential complexity sits quietly underneath, mostly untouched. Who actually holds the legal title to a tokenised asset. What happens when the chain forks. How a court will recognise a transfer. What custody means when the key is the asset. How settlement finality is reached without a central counterparty. How identity is established without surrendering it to a platform.

The work the industry should have been doing all along is on those questions. Legal frameworks for tokenised ownership. Insolvency rules for custodians. Settlement finality in distributed systems. Identity primitives that work across jurisdictions. The serious people are doing this work. It is slow, unglamorous, and gets a fraction of the attention given to the next announcement.

The froth around Quantum Readiness

In quantum readiness, the froth is Q-day. The countdown clock. The vendor briefings that open with a picture of a quantum computer and a year on a slide. Panic about what a sufficiently capable machine will do to public key cryptography, sold as urgency rather than understood as risk. The capability question is real. The marketing wrapper around it is froth.

The accidental complexity, which algorithms, which standards, which key sizes, which lattice constructions, will resolve. NIST is doing that work. The essential complexity will not be resolved by an algorithm. It is knowing where, in your organisation, public key cryptography is actually being used. What depends on what. Which secrets are long lived. Which vendors hold which keys. Whether your hardware refresh cycle allows you to swap cryptographic primitives at all.

The work the industry should be doing is the inventory, not the countdown. A cryptographic bill of materials. A migration plan tied to real refresh cycles, not to a date on someone’s slide. Honest conversations with vendors about which protocols are pinned to which algorithms and how that gets unpinned. None of this lends itself to a keynote. It is, however, where the essential work begins. And the work cannot wait for a cryptographically relevant quantum computer to arrive. Regulators, insurers and clients have already set their deadlines.

The froth itself has evolved

Throughout a long career I have watched the dial move. There was always a little gloss. A press release that exaggerated a product, a capability claimed that did not quite exist yet. The dial has moved from gloss to something else. The volume of unfounded claims, on AI, on digital assets, on quantum, on cybersecurity, has reached a level where the only honest description is froth. It is bad for customers. It is bad for the people doing the actual work. It is bad for boards, who must make capital decisions on the basis of it.

The job of a serious adviser, in my view, is to say plainly which parts are accidental complexity that tooling will eventually take care of, and which parts are the essential complexity that no slide deck can wish away.

That is the only honest position. Everything else is just froth.

Sources

About the Author

Viren Mantri is a cybersecurity advisor and former senior technology leader across Standard Chartered, UBS, McAfee, and KPMG. After three decades at the intersection of technology, risk, and regulation, he now helps organisations cut through complexity and make better security decisions.

CC-BY Viren Mantri, 2026, licensed under a Creative Commons Attribution 4.0 International License.

Disclaimer: All views expressed here are entirely mine.