The “Button Problem”: When Innovation Becomes Incremental
The promise of artificial intelligence meets the reality of workplace integration
History has a way of humbling technological forecasts. Predictions that the internet was going to collapse in 1995 and dire warnings about cameras and television have all fallen flat because experts have continually misjudged transformative technologies. Today, the artificial intelligence revolution is facing its own peculiar limitation: it’s being reduced to a series of simple point-and-click features in existing software platforms. Critics have dubbed it the “Button Problem” — a phenomenon that threatens to turn the most hyped technology of our time into little more than a collection of convenient shortcuts.
Grand promises, grander numbers
The projections for AI’s impact have been dazzling. Investment analysts paint a picture of unprecedented economic transformation: while some forecast a $7 trillion boost to the global economy in the next decade, consultants at McKinsey envision annual gains rivaling the entire British economy. This could potentially reverse a two-decade slump in productivity growth that has troubled economists. According to productivity forecasters, knowledge workers should be able to complete the same tasks in three-quarters of the time while providing 40% better results. More dramatic predictions of knowledge worker productivity doubling over the next six years can be found among the optimists at ARK Investment Management.
When reality bites
Forecasts vary wildly. While optimists foresee annual labor productivity growth potentially increasing by 0.5–1.5 percentage points over the next decade, conservative economists suggest the technology’s contribution might be barely noticeable — even to the point of improving only by a fraction of a percentage point. Early studies do indicate early promising productivity gains in a variety of sectors, but converting these gains into actual cost savings proves difficult. History suggests this pattern is familiar — new technologies often take decades to boost productivity, creating a “J-curve” effect where performance actually dips before surging.
The integration conundrum
This disconnect stems largely from how organisations are implementing AI. Rather than fundamentally reimagining their operations, most companies are simply grafting AI capabilities onto their existing tools. Virtual assistants appear in customer service software, predictive text emerges in email clients, and smart suggestions pop up in productivity tools. These additions boost efficiency but do not deliver the promised workplace transformation. Maybe that is not too surprising — after all, AI is best viewed as the continuation of the IT revolution that started in the 1970s, not as a distinct revolution but rather as another general-purpose technology along with electricity, computers, and the internet.
Zero-based redesign
Some companies have found a better way. Those willing to overhaul their processes entirely while introducing AI have seen great success, with some companies reporting efficiency gains of 25% or more. But these are exceptions. Real productivity leaps require more than just augmentation — an AI chatbot here or a summarizing copilot there — but rather the complete elimination and automation of processes: what consultants call “zero-based redesign”. Success with AI requires more than technological adoption — it demands organizational reinvention. Companies must fundamentally restructure their operations to leverage the technology’s capabilities. This challenge is particularly acute in manufacturing, where AI’s limited exposure to physical production processes has hampered its effectiveness.
We’ve been here before
Since the 1770s, five major technological transformations have reshaped our economy, from the Industrial Revolution through the steam age, electricity era, mass production period, and the current IT revolution. Each needed new institutions — from trade unions to regulatory agencies and welfare states — to manage their far-reaching changes. AI integration requires not only the adoption of new technologies but also institutional reinvention.
Innovating innovation
As organizations make their way through this technological shift, one reality becomes clear: success will require more than the sprinkling of AI features across existing systems. History often shows that predictions about the transformative nature of technologies are often wrong — whether wildly optimistic or deeply pessimistic. The real breakthrough may come from using AI to revolutionize the very process of innovation itself while building the necessary institutional frameworks to manage these changes. This challenge requires unprecedented collaboration between academics, businesses, and policymakers to turn the “Button Problem” into an opportunity for systematic reinvention.
References
“The Button Problem”, https://every.to/napkin-math/the-button-problem-of-ai?utm_source=tldrai
“Generative AI could raise global GDP by 7%, https://www.goldmansachs.com/insights/articles/generative-ai-could-raise-global-gdp-by-7-percent
"The economic potential of generative AI: The next productivity frontier”, https://www.mckinsey.com/capabilities/mckinsey-digital/our-insights/the-economic-potential-of-generative-ai-the-next-productivity-frontier
“How AI Can Change the Way Your Company Gets Work Done”https://hbr.org/2024/07/how-ai-can-change-the-way-your-company-gets-work-done
“Artificial intelligence and the economy: implications for central banks”, https://www.bis.org/publ/arpdf/ar2024e3.htm
“How to fine-tune AI for prosperity”, https://www.technologyreview.com/2024/08/20/1096733/how-to-fine-tune-ai-for-prosperity/
“Artificial intelligence, the economy and central banking”, https://www.bis.org/review/r240924c.htm
“Download: Insights for success in AI-driven organizations”, https://mitsloan.mit.edu/ideas-made-to-matter/download-insights-success-ai-driven-organizations
"Why it is too soon to call the hype on AI’s productivity promise”, https://www.ft.com/content/04260e91-37e6-4e7b-95ec-2ddfe48f24d7
"Zero-Based Redesign: The Key to Realizing Gen AI’s Cost Savings Potential”, https://www.bain.com/insights/zero-based-redesign-the-key-to-realizing-gen-ai-cost-savings-potential/
