The Demand Elasticity Paradox: More than Meets the AI
Wondering how AI might affect your job? Ask this: if your work gets easier to do, will demand for it increase?
In the 1930s, as industrial machinery was transforming traditional labor practices, the renowned economist John Maynard Keynes envisioned a future where such technological advances would bring about a 15-hour workweek. He prophesied that machines would liberate us from the rigors of laborious tasks, freeing up leisure time. Fast forward to today, and Keynes's forecast seems misplaced. Full-time work is still the norm, and the technological utopia of vastly reduced work hours has yet to materialize.
This gap between Keynes's prediction and our reality has led to various interpretations. A prevalent narrative points towards the "Luddite fallacy," an economic theory named after the 19th-century English textile workers later known as Luddites. The Luddites, fearing job loss, protested against the introduction of labor-saving automated looms. This fallacy suggests that technological advancements inevitably shrink the economy's overall 'lump of labor', leading to job losses.
However, experience and observation indicate a different reality. Rather than uniformly eliminating jobs, advancements in efficiency can, paradoxically, increase demand for certain types of work and even create new job opportunities. Understanding this complex interplay between efficiency, demand, and job creation becomes crucial as we navigate the era of AI and automation. It challenges the prevailing notion of technology as a mere job destroyer, suggesting a nuanced and potentially more positive outlook on the future of work.
To fully grasp the often counterintuitive dynamics between technological advancements and job creation, we need to understand demand elasticity and "Jevons Paradox".
Demand Elasticity
Demand elasticity refers to how the demand for a product or service changes in response to a change in price. If the demand is elastic, it means that even a small change in price can lead to a change in the quantity demanded. Conversely, if the demand is inelastic, changes in price have little impact on the quantity demanded.
A similar principle applies to jobs and technology. If the demand for a certain type of job is elastic, making that job easier or cheaper to perform – for instance, through automation – could lead to an increase in demand for that job, not a decrease.
Busting Myths with Jevons: When Efficiency Fuels Demand, Not Dwindles It
This leads us to Jevons' Paradox. In the mid-19th century, English economist William Stanley Jevons made a startling observation about coal usage. He noticed that as new steam engine technologies made the use of coal more efficient, the overall demand for coal did not decrease. Instead, it actually increased. Jevons concluded that improvements in efficiency could paradoxically lead to higher demand, offsetting the anticipated reduction in consumption from the efficiency gain.
Jevons detailed this seemingly contradictory phenomenon in his 1865 book, "The Coal Question." His insights continue to resonate today as we grapple with the implications of automation and artificial intelligence.
While Jevons' observations centered around coal usage, a more relatable modern example might be found in urban highway planning. Planners often fall into a "lump of traffic" fallacy, projecting current traffic levels and adding lanes under the assumption that today's traffic encapsulates the total freeway demand. Yet, as anyone who has watched a newly widened freeway slowly fill up with cars can attest, this is not the case.
Just as Jevons observed with coal, the increased capacity doesn't ease congestion for long. Instead, it induces demand until equilibrium is restored, the congestion returning much like a campfire flaring up with the addition of a new log. It's an expensive phenomenon, with newly laid lanes ending up just as jammed as before.
This concept isn't foreign to the tech industry - we recognize it in the form of network effects. Consider David Sacks' analysis of Uber's total addressable market, for instance. He argues it isn't confined to the traditional taxi and limousine industry size. Sacks illustrates how Uber, by reducing wait times, unlocks latent demand that taxis couldn't access. As wait times drop, demand rises, creating a virtuous loop of increased usage. This is the Jevons paradox in action - an efficiency improvement leading to increased demand, a pattern that echoes across coal mines, freeways, and ride-sharing apps alike.
Reshaping Roles, Not Replacing: The AI Revolution in Translation
We encounter a similar dynamic if we shift our gaze from highways and ride-sharing apps to the translation industry.
In this arena, there was a widespread expectation that AI and machine learning advancements would monopolize the field, rendering human translators obsolete. But the reality, just like the traffic patterns on a freshly expanded freeway, turned out to be quite different.
Two main factors disrupted this expectation. The first hinges on specialization. Areas like legal or medical translations continue to require human proficiency. Despite AI making impressive strides, it still struggles to match human expertise when the stakes are high, and inaccuracies can be costly.
The second factor lies in the rise of hybrid translation services (Machine Translation Post-Editing or MTPE)—services where AI creates a first draft, and humans refine and correct it. This blending didn't axe human roles; instead, it reshaped them. More significantly, this fusion sliced costs by around 40% against traditional human translation. Just as freeway expansions or reduced wait times for rides can unlock demand, so too did this cost reduction, leading to an uptick in translation service demand among a broader clientele.
In numbers, the Bureau of Labor Statistics marked a 19% rise in translators and interpreters—58,400 in 2010 to 69,400 in 2021. But growth wasn't without drawbacks. Lowered service costs resulted in an inflation-adjusted 8% dip in median income for this cohort.
This case study underscores the multi-faceted impact of AI on jobs: understanding demand elasticity helps predict tech-induced shifts. However, it's crucial to remember that financial returns may not follow suit even as demand grows.
Harnessing AI Potential Within the Enterprise
To apply these insights to building an AI company we have to pinpoint roles within an organization where the expense of human resources curbs investment.
Take key account management in sales organizations, for instance. Here, strategic customers get extra attention—through account management, analytics, and more—to ensure account retention and growth. While this increased focus may chip away at the account's profitability, it can boost revenue and extend the account's lifespan.
Now, imagine if the cost of providing these additional services could plummet, akin to what we saw with translation services. The result? The ability to service more key customers profitably by offering personalized services.
Jevons' Paradox and demand elasticity aren't mere economic theories; they're guiding principles for our tech-driven future. From Uber unlocking latent ride-hailing demand to AI reshaping the translation industry, the real story is clear: AI doesn't just replace—it reinvents. As we venture forward, the true litmus test for AI startups lies in uncovering those use cases that don't just adapt to the AI revolution, but thrive under it, buoyed by the transformative power of demand elasticity.