AI: More Hype Than Hypergrowth (For Now)
Fresh & Hot curated AI happenings in one snack. Never miss a byte 🍔
This snack byte will take approx 4 minutes to consume.
The AI era, hailed as the harbinger of economic transformation, has yet to deliver the revolutionary impact many anticipated. It’s been nearly two years since OpenAI released GPT-3.5 to widespread acclaim, with tech luminaries like Bill Gates comparing its significance to groundbreaking advancements like the graphical user interface of the 1980s.
Predictions of AI reshaping global economies and rendering millions jobless sparked both awe and anxiety. Yet, the reality today is more muted.
Data from America’s Census Bureau reveals that only 6% of businesses actively employ AI for production, while economic indicators such as labor productivity and output growth lag far behind the computer age surge of the 1990s.
To understand AI's slow uptake, it helps to revisit the computer age’s trajectory. Back in 1965, computing pioneer Herbert Simon predicted machines would soon perform any task a human could. Yet, productivity gains didn’t materialize as expected.
Decades later, in 1987, economist Robert Solow quipped that the computer age was visible "everywhere but in the productivity statistics." Not until the late 1990s did businesses fully harness computers' potential, transforming industries and prompting Solow to revise his stance.
The late '90s productivity boom was driven by three factors: significant investment in IT infrastructure, plummeting hardware and software prices, and innovative integration of technology into operations. For example, Walmart revolutionized retail by embedding real-time data systems like Retail Link into its supply chain.
These changes were underpinned by businesses ramping up IT investment, which soared by 20% annually between 1995 and 2000. By comparison, today’s AI investments, though promising, are a shadow of that fervor. Business spending on information-processing equipment and software has grown at a modest 4% annually over the last two years.
One explanation for AI’s slower impact lies in its intangibility. Unlike physical computers or networks, AI investments often focus on algorithms, data, and custom solutions, which are harder to quantify in traditional economic metrics.
Payments to AI startups for tailored tools might appear as operational expenses rather than capital investment. Yet, even software spending—critical for AI implementation—remains lackluster. Recent growth in this area is three times lower than the late 1990s, signaling a need for stronger momentum.
Meanwhile, AI's cost dynamics are another hurdle. While the underlying technologies, such as cloud computing and GPUs, are becoming cheaper, middlemen repackaging these tools into enterprise solutions often add hefty margins. This contrasts sharply with the '90s, when the quality-adjusted price of IT equipment dropped by a third, fueling widespread adoption.
Perhaps the biggest challenge, though, lies in the need for businesses to fundamentally rethink their operations to unlock AI’s full potential. Current AI use cases are often narrow—fraud detection in finance or chatbots in customer service—failing to revolutionize workflows at scale.
Data infrastructure remains a significant barrier; most firms lack the custom datasets necessary to train sophisticated AI models.
Economist Rudi Dornbusch’s observation that economic changes happen “slower than you thought they would and then faster than you thought they could” might be apt here. The AI revolution, much like the computer age, seems to be following a prolonged incubation period.
Comparing today’s AI landscape to the 1970s, when innovations like microprocessors and memory chips emerged but failed to immediately impact productivity, offers a sobering perspective. Back then, despite breakthroughs like email and the early internet, labor productivity in America grew at a tepid 1.7% annually between 1975 and 1994.
AI may indeed hold transformative potential, but history suggests we might need to wait a little longer—and invest a lot smarter—before its full economic impact becomes evident.
So, while the AI hype train is gathering steam, its station stop at "Revolution Boulevard" may still be a few years down the track.