Steve Jobs: The Lesson of the Gulf War is that the Best Software Will Win the War
Steve Jobs, in a speech at MIT's Sloan School of Management in 1992, relayed the judgment of Pentagon CIO Paul Strassmann: the lesson of the Gulf War is that the best software will win the war.
Strassmann found, after researching the IT spending of various U.S. companies in his book "The Business Value of Computers," that successful companies and poorly performing companies had similar IT spending as a percentage of revenue, both around 2%; the difference was not in budget proportion but in IT allocation.
Jobs recounted that poorly performing companies allocated most of their IT budgets to "management productivity"—such as documentation, reports, administrative, and office computer tools; whereas more successful companies directed most of their budgets towards "operational productivity," which involves systems that directly run core business operations, transactions, supply, manufacturing, or service delivery.
Jobs acknowledged that during the first ten years at Apple, his focus on personal computers and Macintosh essentially served management productivity and did not directly engage with core business operations. At that time, he had left Apple and was leading NeXT; later, NeXTSTEP became the technological foundation for Mac OS X after Apple acquired NeXT.
The verifiable key points from the original text are "software determines victory or defeat" and "budget allocation determines productivity," but the statements regarding "hardware reshuffles every 18 months, leading by a factor of two can only last for 6 months, and Microsoft took 9 years to catch up with Macintosh" do not appear in the available transcript of the MIT speech and cannot be considered confirmed quotes from this speech.
From a market mechanism perspective, the true buyers of IT in companies are not the CIO's budget sheets but the business departments responsible for revenue, inventory, pricing, risk control, manufacturing, logistics, and customer delivery. If funds flow towards generic office tools, the benefits typically manifest as localized efficiency; if directed towards operating systems embedded in core processes, software can affect unit costs, turnover speed, error rates, and customer switching costs. The beneficiaries are software vendors that master business data, workflow, and system integration capabilities; vendors that rely solely on hardware specifications or generic functionalities are more easily pressured by standardized competition to compress profits.
Source: Public information
ABAB AI Insight
Strassmann served as the CIO of the U.S. Department of Defense and was cited by Jobs as an authoritative case for software strategy after the Gulf War; at this time, Jobs was in the NeXT phase after leaving Apple. NeXT did not achieve mass market success through hardware sales, but the object-oriented software environment, development tools, and Unix foundation of NeXTSTEP eventually entered Apple in 1996 after the acquisition of NeXT, becoming the technological precursor to Mac OS X. Jobs's low point was not about abandoning software but rather a repositioning from personal computer hardware sales to software platforms and development efficiency.
The watershed in capital pathways is whether software enters the "real-time nodes of revenue generation and cost occurrence." Office software sells employee efficiency; ERP, payment clearing, supply chain, transactions, advertising bidding, pricing engines, and industrial control systems participate in every order, every production scheduling, and every risk decision of a business. The former typically charges by seat and is easily substituted; the latter, once it accumulates business rules, historical data, and organizational processes, incurs replacement costs that translate into multi-year contracts, migration risks, and ongoing service revenue. Strassmann's observation of 2% is therefore not a call for companies to buy more IT but rather a demand for capital to shift from showcase digitalization to digitalization that controls business operations.
Historically, SAP entered manufacturing, procurement, finance, and inventory master data through ERP; Oracle entered transaction records through databases and enterprise systems; Bloomberg terminals entered trading desks through real-time financial information and workflows; the reason they established barriers is not just due to complex code but because systems and customer business processes co-evolve. Later, Apple transformed hardware sales into a platform driven by software rules, developer distribution, and user data through iOS, the App Store, proprietary chips, and services. Jobs's judgment in 1992 foreshadowed Apple's path of emphasizing hardware-software integration after returning from being a PC manufacturer.
Essentially, it is about the transfer of pricing power: hardware performance spreads quickly, and buyers can easily compare prices; software that delves into operations can change business costs, delivery, and revenue, allowing suppliers to anchor prices based on business value rather than material costs. Generative AI is also repeating this filtering: generic chat tools will be pressured by functional homogenization, while products that truly gain pricing power will be systems that connect to orders, payments, approvals, medical records, code releases, or supply chain execution, and take verifiable responsibility for outcomes. The barriers of software have never been about how much code has been written but rather how much business would be interrupted by its replacement.
ABAB News · Laws of Cognition
The same budget directed towards core processes will create barriers.
Hardware sells performance, software sells irreplaceable processes.
Management tools improve efficiency, operational systems rewrite profits.