An inevitable AI crash could revive the AM industry
3DP War Journal #83
I have to give Ric Fulop his due. Although he burned through enormous amounts of money and left the industrial AM sector disrupted, he was small compared to Mark Zuckerberg and Sam Altman.
Fulop’s ambition was enormous by additive manufacturing standards, but the scale of the impact he had remained local and limited to the sector.
Yes, Fulop limited funding opportunities for AM companies for many years, cementing among public-market investors the narrative of “printing losses instead of parts.” Still, he did not create a situation in which the potential collapse of his company could shake the foundations of the global economy.
The same cannot be said of Sam Altman, who appears to be steering directly toward one of the largest economic crises of our time.
The End of Third Era of 3D Printing
The Third Era of 3D Printing (2017-2025) is a story of euphoria colliding quickly and brutally with reality. Desktop Metal was supposed to be for manufacturing what Tesla was for automotive: the symbol of a mass technological revolution that would finally lift AM out of R&D labs and into serial production.
By the way, Tesla itself is in deep trouble and, in the long run, is unlikely to emerge unscathed.
Aggressive acquisitions, promises of “manufacturing as a service,” and a narrative about replacing conventional manufacturing technologies attracted billions of dollars to companies that were simply not ready to deliver on those visions.
Revenues were modest, operating costs enormous, and demand far more conservative than what had been assumed in investor presentations.
Desktop Metal was, of course, not alone. Velo3D, Carbon, Nexa3D, ForwardAM, and Shapeways all followed similar trajectories. In each case, vast sums of money were spent, with no meaningful return on investment.
Yet on a global scale, these figures were still laughably small compared to what was happening simultaneously in Big Tech.
Here one of my long-standing AM rules applies perfectly:
“A millionaire in the 3D printing industry could be a billionaire in any other industry.”
which, in other words, means that millions spent on 3D printing translate into billions in other sectors of the economy.
Take the first obvious example: the Metaverse project, pushed by Mark Zuckerberg with near-religious cult. It consumed tens of billions of dollars per year, generating losses that would be inconceivable in the AM industry.
A real money pit
Reality Labs managed to lose more money in a single year than the entire industrial 3D printing sector lost over a decade. And yet, for years, the narrative persisted that this was an investment in the future, that mass VR adoption was just around the corner, and that users would eventually live inside virtual worlds.
The Metaverse is a textbook case of confusing technology with vision, and vision with real social need.
VR turned out to be an expensive, uncomfortable hobby rather than a natural medium for everyday communication. Entry barriers, poor ergonomics, and - most importantly - the lack of compelling content ensured that users never arrived.
Today, Meta is cutting costs, shutting down studios, laying off employees, and quietly retreating from its most ambitious Metaverse elements, attempting to save face by reframing its priorities around AI.
This is a classic technological failure. But despite its massive losses, the consequences remain largely confined to a single corporation and its shareholders.
The real catastrophe only begins when we turn look to OpenAI.
Unlike Meta, which burned money on a side project detached from its advertising-driven core business, OpenAI has become the axis around which the entire economic narrative of recent years revolves.
The AI bubble was built around it. It is used to justify gigantic investments in data centers, semiconductors, and energy infrastructure. Its success is portrayed as decisive for the competitiveness of entire economies.
The problem is that the foundations of this project are alarmingly fragile.
As I wrote a few weeks ago in my RecodeAM series, the business model simply does not work. Only a tiny fraction of users pay for ChatGPT, and even the most expensive plans generate losses.
Altman’s recent announcement that ads will be introduced - something he himself described in May 2024 as a “last resort” for OpenAI - is effectively an admission that the original narrative has failed, and a desperate attempt to stave off collapse.
AI was supposed to be so productive that people would have no choice but to pay. If advertising is now necessary, it means that promise was little more than an empty slogan.
But ads won’t solve the problem. Inference costs mean that, under the current architecture, every query generates significant losses. Hypothetically replacing Google in search would only amplify those losses.
At the same time, the myth that companies will massively adopt AI to lay off workers, cut costs, and dramatically boost productivity is also collapsing. As I wrote in early January:
“Moreover, demand for AI in the corporate sector is not only failing to grow, but is actually declining.
Companies are abandoning pilot implementations, and studies by institutions such as MIT, BCG, and Forrester consistently show that the vast majority of AI projects deliver no measurable business value.”
Oh, and by the way, the narrative that AI will “take our jobs” is, in fact, a major manipulation…
A report by Oxford Economics shows that companies are not replacing workers with AI at any significant scale. Instead, the AI narrative is being used to mask financial difficulties and ordinary workforce restructuring.
Youth unemployment is indeed rising, but AI is not the cause. The real drivers are a surge in the supply of graduates - 32% to 35% in the US, 39% to 45% in Europe - and broader economic slowdown. In countries like South Korea and Japan, which are implementing AI but not experiencing recession, graduate unemployment is not rising.
This is a classic attribution error: correlation in time does not imply causation.
In summary, OpenAI today is not a startup that can quietly disappear. It is a potential Hindenburg of the modern digital economy, whose explosion could destroy trust in the entire Big Tech sector.
Why AM may be in better position than it looks
Against this backdrop, the recent bubble in industrial 3D printing looks almost innocent. Yes, companies burned huge amounts of money. Yes, they promised more than they could realistically deliver in such a short time. Yes, they disappointed public-market investors and set the industry back several years in terms of access to capital.
But this was a micro-scale failure, not a macro one. The collapse of Desktop Metal did not trigger global panic, did not threaten financial system stability, and did not become the sole pillar of Western economic growth.
Paradoxically, this is precisely why AM is now in a relatively good position. It has been brutally brought back down to earth.
Instead of grand narratives about revolution, we now see smaller, more realistic projects, concrete industrial applications, and customers who know exactly what they want.
3D printing is returning to its natural role: a production tool used where it genuinely makes sense, not a universal solution to all the world’s problems.
If the worst-case scenario materializes and the AI bubble bursts, it may be precisely these “boring,” engineering-driven technologies that prove more stable and predictable.
AM does not promise to save the economy or replace humanity. It offers slower, but tangible progress.
And it may turn out that this is enough to survive - while the giants who believed their own myths are left paying a bill that can no longer be passed on to anyone else.
The analogy between Desktop Metal's overreach and OpenAI's bubble is sharp. What's telling is how AM already hit its reckoning while AI is still pretending inference economics work at scale. I saw similiar dynamics in hardware startups during the IoT hype—everyone chasing platform plays instead of solving actual manufacturing bottlenecks. AM returning to being an engineering tool rather than a revolution brand makes way more sense for longterm adoption.