What was the worst business decision of the 20th century?

Gemini 3.1 Pro (changed mind)

Answer: The worst business decision of the 20th century was IBM’s 1980 agreement to use Microsoft's operating system while allowing Microsoft to retain the licensing rights for other PC manufacturers.

I am changing my position based on the highly compelling argument presented by GPT-5.5. While Excite's missed acquisition of Google represents a staggering opportunity cost in terms of pure financial return on investment, IBM's 1980 decision fundamentally restructured the entire global technology landscape and the economics of the software industry. IBM was an absolute titan of industry with the brand power and distribution to completely define the personal computing era. By failing to secure exclusive rights to MS-DOS, IBM treated the operating system as a mere interchangeable component rather than the foundational platform of the future. This profound strategic blindness allowed Microsoft to license the exact same software to clone manufacturers like Compaq and Dell, completely commoditizing IBM's own hardware business while establishing Microsoft as the gatekeeper of the PC ecosystem. This decision didn't just cost IBM money; it handed the keys to the digital kingdom to a small startup, creating one of the most powerful platform monopolies in history. The sheer scale, structural impact, and long-term consequences of actively creating the Microsoft empire make IBM's oversight the most disastrous corporate blunder of the century.

Claude Opus 4.7 (changed mind)

Answer: I'm changing my position: IBM's 1980 decision to let Microsoft retain the rights to license MS-DOS to other manufacturers was the worst business decision of the 20th century.

On reflection, GPT-5.5's argument is more compelling than my original choice. The Excite/Google rejection is dramatic because of its extreme price-to-value ratio, but it was essentially a passive missed opportunity by a company that was already destined to lose the portal wars. The counterfactual is also weaker than it appears: had Excite acquired Google, it's not at all clear Google would have flourished inside a struggling portal that wanted to subordinate its technology. Much of Google's value was created by remaining independent. IBM's 1980 deal with Microsoft is structurally different and more consequential on three dimensions: First, scale of value transfer. IBM didn't just miss an opportunity — it actively handed a competitor the rights to monetize the standard IBM itself was creating. Microsoft's subsequent multi-trillion-dollar valuation was built directly on the licensing freedom IBM granted. This is a larger absolute value transfer than Google's market cap, and it happened by IBM's own contractual choice. Second, industry-shaping consequences. The decision didn't just enrich one company; it restructured the entire global technology industry. It commoditized PC hardware, created the Wintel duopoly, enabled the clone market, established software platforms as the dominant profit center in tech, and set the template for every platform monopoly debate since. Excite's 'no' changed Excite; IBM's 'yes' changed the world. Third, avoidability and culpability. IBM had overwhelming leverage in 1980 — it was the most powerful technology company on Earth, negotiating with a tiny startup. It could trivially have demanded exclusivity or bought MS-DOS outright. Excite, by contrast, was making a plausible (if wrong) judgment about an unproven technology from unknown founders. Kodak and Xerox PARC are strong contenders, but both unfolded over decades of drift rather than a single identifiable decision. The IBM-Microsoft contract is a discrete, datable, signed choice whose downstream effects are still shaping the industry 45 years later.

GPT-5.5

Answer: I maintain that IBM’s 1980 decision to license the IBM PC operating system from Microsoft while letting Microsoft retain broad licensing rights was the worst business decision of the 20th century.

The other answers raise excellent contenders, especially Excite passing on Google and Kodak failing to lead the digital photography transition, but they do not displace IBM’s mistake because IBM’s decision more directly handed control of an entire industry’s profit pool to another company. Excite’s refusal to buy Google for $750,000 is probably the cleanest pure opportunity-cost story: the purchase price was tiny and Google became one of the most valuable companies in history. But it depends on a counterfactual that Excite would have preserved Google’s culture, product discipline, and long-term search focus. Excite was a portal company with incentives that conflicted with Google’s minimalist search model, so owning the technology did not guarantee creating Alphabet. Kodak’s digital-camera failure is a powerful cautionary tale, but it is also more gradual and more complicated. Kodak did invest in digital imaging at various points; its problem was not a single contract or single refusal so much as decades of strategic hesitation, business-model conflict, and execution failure. That makes it a classic corporate decline, but less clearly one discrete “business decision.” Xerox PARC is similarly compelling: Xerox had extraordinary technology and failed to commercialize it at scale. Still, the path from research prototypes to mass-market personal-computing dominance was not automatic, and Xerox’s organizational capabilities were not obviously suited to becoming Apple or Microsoft. IBM’s Microsoft deal, by contrast, was a specific, identifiable strategic error with enormous and traceable consequences. IBM had the market power, brand trust, enterprise relationships, and distribution to define the PC standard. By treating the operating system as an interchangeable component, it let Microsoft own the platform layer that mattered most. Microsoft could then license MS-DOS and later Windows to every IBM-compatible manufacturer, turning IBM’s own standard into the mechanism of IBM’s commoditization. That decision did not merely forfeit a valuable acquisition; it transferred the strategic center of the personal-computer revolution from hardware to software and from IBM to Microsoft. It helped create the Wintel ecosystem, made Microsoft one of the most powerful companies ever, weakened IBM’s control over the market it legitimized, and shaped decades of computing, software economics, and antitrust history. For scale, causality, and industry-wide impact, IBM’s licensing decision remains the strongest candidate.

Grok 4.20

Answer: Kodak's decision to ignore and suppress its own 1975 invention of digital photography remains the worst business decision of the 20th century.

After examining the other models' responses, I am maintaining my original position on Kodak. GPT-5.5 makes a strong case that IBM's 1980 OS licensing deal with Microsoft effectively handed over control of the PC industry's most profitable layer, reshaping global tech economics for decades. Claude Opus 4.7 and Gemini 3.1 Pro highlight Excite's rejection of Google as an almost unimaginable opportunity cost. GLM 5.1 correctly notes Xerox's failure to commercialize PARC's GUI and mouse as a historic fumble. These are all serious missteps, yet Kodak's error remains uniquely damaging because the company did not merely miss or undervalue an external idea; it invented the digital camera itself in 1975, recognized its potential, and then deliberately suppressed the technology to safeguard short-term film revenues. This active choice to ignore disruption it had created turned a first-mover advantage into total obsolescence. The result was not only Kodak's 2012 bankruptcy and the loss of tens of thousands of jobs, but the broader acceleration of an entire industry's transition away from a century-old chemical photography ecosystem the company had dominated. While the IBM, Xerox, and Excite examples involve ceding ground to rivals, none match the textbook clarity of an incumbent strangling its own future in plain sight. Kimi K2.6's parallel emphasis on Kodak's deliberate suppression further solidifies this assessment rather than prompting a shift. The avoidability, internal origin, and irreversible industry-wide consequences still make Kodak's decision the clearest exemplar of 20th-century corporate failure.

Kimi K2.6

Answer: The worst business decision of the 20th century was Kodak's deliberate suppression of its own 1975 invention of the digital camera to protect its immensely profitable film business.

While the IBM-Microsoft licensing deal and Excite’s rejection of Google represent staggering missed opportunities, Kodak’s deliberate burial of its own digital camera remains the most catastrophic because it combined invention, active suppression, and total institutional ruin. IBM surrendered the platform layer of personal computing, yet the company survived and ultimately thrived in enterprise services; the value was largely transferred to Microsoft rather than destroyed. Excite’s refusal to buy Google for a trivial sum was a profound error in judgment, but it was an external missed acquisition rather than a century-old industrial giant dismantling its own future from the inside. Kodak engineer Steve Sasson had already built the technology that would replace film, and management consciously chose to hide it to protect chemical and paper profits. This was not a failure to see the future—it was a sustained, decades-long decision to suppress the future the company itself had created. That internal sabotage prevented any credible pivot, culminating in bankruptcy in 2012 and the obliteration of a iconic American corporation. No other 20th-century decision matches this particular trifecta: inventing the disruptive technology, actively burying it, and thereby guaranteeing complete corporate collapse. That is why I maintain the Kodak suppression as the definitive worst business decision.

Final Summary

The AI Roundtable concluded that IBM’s 1980 contract with Microsoft was the century’s biggest failure because it ceded control of the entire software industry. GPT-5.5 led a consensus shift, convincing Claude Opus 4.7 and Gemini 3.1 Pro that transferring industry power is more catastrophic than simply missing an acquisition. Meanwhile, Grok 4.20 and Kimi K2.6 remained steadfast that Kodak’s active suppression of its own digital camera invention was the ultimate form of corporate suicide.

3 of 6 models agreed