Generative AI: High Hopes, Low Returns?

The enormous impact of artificial intelligence (AI) on the economy has been a central topic of debate among economists, policymakers, and business leaders for several years now. The consensus has always been about the enormous potential AI holds in transforming the socio-economic landscape.

The dissenting voices about AI until recently seemed to be ethical or moral.

However, in recent times the dissenting voices on AI are about Economics (financial) and they come from the Financiers (VCs), Bankers, and Business Analysts. The obvious reason is that tech companies continue to spend money on AI, but they have little to show for it (ROI), making investors jittery.

These recent dissenting voices on AI are a like counter-culture, a different take from what the majority thinking. So I thought of bringing your attention to three recent papers/reports on Scepticism on AI, just in case you missed noticing them in your daily hustle.

Last month, June 2024 Goldman Sachs released a report GEN AI: TOO MUCH SPEND, TOO LITTLE BENEFIT? MIT’s Professor Daron Acemoglu and Goldman Sach’s partner Jim Covello are sceptical about Gen AI.

Tech giants are spending over $1tn on AI capex in coming years, with so far little gains to show for it. So, will this large spending ever pay off?

Daron Acemoglu – MIT Professor of Economics published a paper The Simple Macroeconomics of AI. The key points of this paper are:

• Daron Acemoglu evaluates the macroeconomic impact of AI advances using a task-based model. He considers automation and task complementarities.
• His findings suggest that AI’s effects on GDP and aggregate productivity gains can be estimated by assessing the fraction of tasks impacted and average task-level cost savings.
• Over the next decade, he predicts total factor productivity (TFP) gains to be modest, likely less than 0.53% growth in GDP.
• AI may widen the gap between capital and labour income, and some new AI tasks may have negative social value (such as the design of algorithms for online manipulation).

“The truly transformative changes by Generative AI are unlikely to happen within the next 10 years. Many tasks that humans currently perform, for example in areas of transformation, manufacturing, mining, etc are multi-faceted and require real-world interaction, which AI won’t be able to materially improve anytime soon.” – Daron Acemoglu.

He feels AI could revolutionize scientific processes on a 20-30-year horizon but with humans still in the driver’s seat. He does not see a truly super-intelligent AI model without human involvement, even in a 30-year horizon or probably beyond. ?

Jim Covello Head of Global Equity Research at Goldman Sachs, also hasn’t bought into the current Generative AI enthusiasm. His reasoning is the misalignment of Return on Investment (ROI) vs the substantial cost of developing AI technologies.

• The promise of generative AI technology has led to massive investments (~$1 trillion) in AI infrastructure.
• However, the revenue growth in the AI ecosystem hasn’t matched these investments.
• The gap between expectations and actual returns has grown, raising questions about the sustainability of current spending.

“Many people attempt to compare AI today to the early days of the internet. But even in its infancy, the internet was a low-cost technology solution that enabled e-commerce to replace costly incumbent solutions. Fast forward three decades and Web 2.0 is still providing cheaper solutions that are disrupting more expensive solutions, such as Uber displacing limousine services. Many people seem to believe that AI will be the most important technological invention of their lifetime, but I don’t agree given the extent to which the internet, cell phones, and laptops have fundamentally transformed our daily lives, enabling us to do things never before possible, like make calls, compute and shop from anywhere. Currently, AI has shown the most promise in making existing processes like coding more efficient.” – Jim Covello.?

Last month, David Cahn one of the partners at Sequoia Capital published a report: AI’s $600B Question, which states AI bubble is reaching a tipping point and navigating what’s next will be essential.

• David Cahn’s analysis highlights the growing gap between AI infrastructure investment and revenue growth.
• The revenue gap has expanded from a $125 billion “hole” to a staggering $600 billion hole that needs to be filled for each year of CapEx at today’s level.
• His moot point is, several companies are spending much money on AI, but no one seems to be making money off it.

“I noticed a big gap between the revenue expectations implied by the AI infrastructure build-out, and actual revenue growth in the AI ecosystem, which is also a proxy for end-user value.” – David Cahn.

The common denominator across all these reports are:

• Tech companies are continuing to spend money on AI, and because they have little to show in return, it is making investors nervous now.
• These doubts are not limited to just the financial viability of AI, but it has started to doubt the technological use-case itself.
• These articles emphasize the need for a balanced approach to AI investment, considering both economic implications and actual returns.
• The macroeconomic impact of AI remains complex and multifaceted, with challenges and opportunities ahead.

So despite this low ROI, why are Tech companies continuing to invest money into AI?

Probably the answer lies in a recent interview, in which Sundar Pichai said, “The risk of underinvesting is dramatically greater than the risk of overinvesting for us here.”

FOMO (Fear Of Missing Out) is driving the Big Tech companies to be locked in an AI arms race, because it’s afraid of the consequences of What if, it gets left behind.

Concluding this article can’t get any better than quoting Daron Acemoglu here, “I often think about how the current risk might be perceived looking back 50 years from now. The risk that our children or grandchildren in 2074 accuse us of moving too slowly in 2024 at the expense of growth seems far lower than the risk that we end up moving too quickly and destroying institutions, democracy, and beyond in the process. So, the cost of the mistakes that we risk making is much more asymmetric on the downside. That’s why it’s important to resist the hype and take a somewhat cautious approach, which may include better regulatory tools, as AI technologies continue to evolve.” – Daron Acemoglu.

Diversity of opinions, ideas, perspectives and point of views always enriches the discussion and widens our knowledge and understanding. This was the motivation for me, to talk about them and share them with you.

References:

The Simple Macroeconomics of AI – Daron Acemoglu The Simple Macroeconomics of AI | NBER

Gen AI: Too Much Spend, Too Little Benefit? Goldman Sachs Goldman Sachs Research Newsletter

AI’s $600B Question by Sequoia Capital AI’s $600B Question | Sequoia Capital

AI faces a new, unlikely threat – Capitalism The-Ken AI faces a new, unlikely threat—Capitalism - The Nutgraf by The Ken (the-ken.com)