A Lightbulb is Not a Brighter Candle: 3 Mental Models to Understand the Impact of AI
We explore three mental models illuminating AI's potential for transformative, not just incremental, impact.
In 1881, Thomas Edison established electricity-generating stations at Pearl Street in Manhattan and Holborn in London. Yet by 1900, less than 5% of mechanical drive power in American factories came from electric motors. The age of steam lingered on. Why? Because initially, electrification merely resulted in the replacement of steam-powered machines with those powered by electricity. While more cost-efficient, electric machines didn't help unlock the real value of this new power source.
That insight came in 1908 when Henry Ford realized that electricity could enable an entirely new factory layout and production model - the assembly line. It was a breakthrough that supercharged manufacturing productivity and efficiencies in a way simply upgrading existing machines could not.
Today, we find ourselves in a similar transition period with artificial intelligence. Current AI applications are akin to those early electric machines - impactful upgrades on existing processes but not yet realizing AI's full disruptive potential. Just as a lightbulb is not merely a brighter candle, AI represents a technological shift that could reshape industries, business models, and how work itself is conducted.
In this piece, we'll revisit three mental models explored in-depth over the past year that can help us wrap our minds around AI's potential. By viewing the impact of AI through these lenses, we can better understand the changes this new technology could unlock.
The Economics of Abundance: When Personalization Becomes Virtually Free
In the not-so-distant past, getting braces was a rite of passage filled with discomfort and awkwardness. Orthodontists would bond brackets to your teeth, connect them with wires, and then tighten the contraption over countless visits to shift your smile into place. It was a one-size-fits-all solution requiring professionals to meticulously adjust the wires every few weeks over 2-3 years of treatment.
Then in 1997, a Stanford student named Zia Chishti looked at the antiquated procedure and thought, "This is nuts. Couldn't this be done with clear, removable aligners that are 3D-printed for each patient's mouth?" And just like that, the multi-billion dollar idea for Invisalign was born.
Invisalign's breakthrough was using 3D printing and digital imaging to create customized, clear plastic aligners for each user. Patients simply swap out a new set of printed aligners every week or two, with the molds gently shifting their teeth into the ideal position. No wires, no brackets, no monthly orthodontist visits - just a steady progression toward a straighter smile through mass personalization.
Like 3D printing enabled this personalized orthodontics revolution, AI has the potential to drastically reduce the cost of customization across many industries to levels impractical today. From generating personalized summaries and illustrations on demand to creating custom software user interfaces, we're just beginning to scratch the surface of AI's mass personalization capabilities. The implications could be just as disruptive as Invisalign has been for orthodontics, as we explored in "The Economics of Abundance: What Do You Charge When Software is Free to Produce?"
You Can't Sell Horseshoes to People Who Don't Own Horses
When a disruptive new technology emerges, it's easy to get fixated on the immediate impacts to your own workflows and business model. But the secondary ripple effects through your customers' industries can be even more consequential - and hazardous to ignore.
Take the mini-lab companies that enabled 1-hour photo development before the digital camera revolution. Firms like Noritsu were printing money, providing the machinery that allowed grocery stores, pharmacies, and mall kiosks to quickly process film rolls and print pictures on-site. Their largest mini-labs could crank out 3,000 developed rolls per hour to meet booming demand.
Kodak's wildly popular disposable FunSaver cameras were fueling this need for fast turnaround photo printing. The photography giant sold tens of millions of the single-use cameras each year in the 1990s. For companies like Noritsu supplying the mini-lab equipment, the future looked bright.
But almost overnight, digital cameras began disrupting the entire film-based value chain that Noritsu and others were serving. Suddenly, there was no need to develop and print photos. No matter how advanced Noritsu's machines became, their customers' business models were being rendered obsolete by digital cameras.
Despite its dominance in the heyday of 1-hour photo processing, Noritsu was ultimately decimated by the digital transition. As we explored in more depth in "Reinvent or Relinquish: You Can't Sell Horseshoes to People Who Don't Own Horses", when new technologies upend your customers' industries, your own business will inevitably feel those shockwaves.
AI's Bespoke Fit for the Digital Assembly Line
When we picture a factory, Henry Ford's revolutionary assembly line model immediately comes to mind. By breaking down production into a sequential series of small tasks performed by different workers, Ford ushered in a new era of streamlined manufacturing for the masses.
Today's companies may not produce physical goods, but their operations still rely on finely-tuned digital assembly lines humming with the flow of information and tasks. Sales, marketing, product, and other departments all have their own processes akin to a factory floor - with work packets progressing through each stage like products being assembled.
Just as a factory utilizes Andon cords to identify issues and gemba walks for managers to observe operations, these digital assembly lines employ Slack channels for quickly surfacing blockers and review meetings to optimize handoffs between stages. Software ecosystems like Salesforce, Marketo, Jira, and countless other tools drive these knowledge factories.
Keeping this intricate machinery running smoothly has spawned entire new job categories. Roles like Marketing Operations Manager, Sales Operations Analyst, and Revenue Operations have skyrocketed over 100% in demand over the past five years according to job sites. These "biz ops" professionals are the new production line supervisors - monitoring, fine-tuning, and ensuring the entire workflow system operates at peak efficiency.
In my previous article, "The Future of Knowledge Work: Software as the New Assembly Line", we explored how AI has the potential to radically streamline and optimize these digital production lines. By intelligently automating tasks, seamlessly routing work, and identifying bottlenecks, AI could unlock new operational efficiencies for knowledge factories.
As we stand on the precipice of the artificial intelligence era, it's tempting to view this new technology through the narrow lens of how it can incrementally improve our current workflows and business operations. But if the history of transformative technologies like electricity taught us anything, it's that the true disruptive impact of AI likely won't be realized until we shed those constrained perspectives. The lightbulb wasn't just a brighter candle, and AI won't simply be a smarter software program.