Trick or Treat? Decoding AI

Artificial intelligence has quickly evolved from a niche technology to a driving force behind market enthusiasm, corporate strategy, and everyday conversation. From the boardroom to the classroom, discussions about AI’s potential to reshape productivity and its implications for the workforce are becoming increasingly common. Yet, despite the surge in investment and the extraordinary market performance of AI-related companies, a key question remains: Is AI truly delivering on its promise of productivity, or are we witnessing another speculative cycle fueled by hype?

To examine this more closely, we decided to turn the question back on AI itself, specifically ChatGPT. We posed a series of prompts examining the relationship between AI and recent stock market gains, the paradox of rising investment without corresponding productivity growth, and the complex financing structures driving the current AI economy. Over the course of the conversation, ChatGPT generated pages of responses, providing insightful, data-driven, and at times, cautionary content.

As any user quickly learns, the quality of an AI’s output depends mainly on the precision of the prompts it receives. Well-crafted prompts can surface meaningful analysis, while vague ones tend to yield superficial answers. This exercise wasn’t just about seeing what AI “thinks,” but about understanding how it processes complex economic ideas when guided by thoughtful, targeted inquiry.

What follows is a distilled, plain-English summary of that dialogue designed for readers who are curious about AI’s impact on the economy, skeptical of the surrounding hype, and mindful of what this technological shift may mean for both investors and the next generation entering the workforce.

Here are the prompts I gave ChatGPT.

• How much of the stock market gains over the past two 2 years are attributable to the AI boom? Please list sources.
• AI-related stocks have gone up rapidly, and an epic amount of capital is flowing into the space. MIT and others report that though companies are investing in AI, few are realizing productivity gains. Is AI the ultimate “build it, and they will come”? What if productivity gains do not come through as anticipated?
• Please comment on the complex financing and circular nature of the AI economy and deal-making. What are the risks inherent in how many of these co-dependent deals are being structured?

Over the course of this conversation, ChatGPT generated over 600 words over 21 pages.

Therefore, my final prompt was: “There is a lot of information here. Boil this down for a reader who has heard cocktail party talk about AI, has dabbled in Language Learning Models) LLM, and has talked to other parents about AI giving bad answers, or has concerns about AI reducing the number of employment opportunities for their soon-to-graduate children. Write at a college degree level and in a conversational tone. Limit it to no more than 1000 words.

Chat GPT’s Response

Here’s a concise, plain-English summary you can share with educated readers who are curious—but skeptical—about the AI boom.

———————————————————————————————————————–

AI’s Wild Ride: Promise, Hype, and the Productivity Puzzle

Over the past two years, AI has gone from a tech-industry curiosity to the topic of dinner parties and school-pickup conversations. ChatGPT and other “large language models” gave everyone a taste of what machines that understand and generate language can do. The result? A tidal wave of investment, new companies, and big promises that AI will transform the economy.

But there’s a catch: despite trillions in market value created, the promised productivity boom hasn’t shown up yet. MIT, McKinsey, and others have found that while most large firms are investing in AI, only a small fraction are seeing measurable efficiency gains.

So what’s going on—and what happens if the big payoff never arrives?

———————————————————————————————————————–

1. Why the Money Is Pouring In

AI feels like electricity in 1900 or the internet in the 1990s: a “general-purpose technology” that could touch every industry. The belief is that whoever controls the data, chips, and models will dominate the future.

That belief has created a massive, almost circular investment ecosystem:

• Chip makers like Nvidia fund AI startups that, in turn, spend billions on Nvidia chips.
• Cloud providers (Microsoft, Amazon, Google) invest in model developers like OpenAI and Anthropic—who buy their cloud capacity back.
• Venture funds pour into data-center builders who rely on demand from those same AI firms.

It’s a self-reinforcing loop: everyone finances everyone else, partly to guarantee future demand and partly to keep capital within the AI “family.” It makes sense while enthusiasm is high—but it also means if demand softens, the whole network feels it at once.

———————————————————————————————————————–

2. The Productivity Paradox

Economists call this the productivity paradox: you can see AI everywhere except in the statistics.

Just like the early computer era, big gains take time because companies must rebuild how they work—not just add a new tool. AI often helps most when paired with new processes, data cleanup, and worker training, which can take years.

Early users report improvements in specific tasks—summarizing documents, drafting code, analyzing data—but the savings are often offset by new costs: cloud bills, model tuning, compliance checks, and human oversight to catch mistakes. Think of it as installing a jet engine on an old prop plane: impressive power, but you still have to rebuild the wings.

———————————————————————————————————————–

3. The “Build It and They Will Come” Problem

Much of the current AI spending assumes that productivity gains will come, eventually. Data centers are being built faster than electricity grids can keep up; companies are pre-ordering chips years out. The risk is that we’re in a “Field of Dreams” phase—building infrastructure in hope that profitable use cases follow.

If those gains don’t appear soon, some projects may have to scale back, merge, or write down investments. It wouldn’t end AI—just shrink it to a more realistic pace, similar to how the dot-com bust cleaned up the excesses of the early internet.

———————————————————————————————————————–

4. The Human Side: Jobs, Skills, and Anxiety

Parents naturally wonder: will AI wipe out jobs before their kids even start careers?
So far, the evidence says no—but it will reshape work.

AI excels at pattern recognition and language processing, but it still struggles with judgment, context, and ambiguity—the very areas where people add value. Think of it less as replacing humans and more as reshuffling tasks. Jobs that combine human oversight, creativity, and technical literacy will grow fastest.

Students who learn how to work with AI tools—prompting, checking, and integrating them into workflows—will be in demand. The big risk isn’t obsolescence, it’s complacency: assuming a degree alone guarantees employability in a world where learning agility matters more.

———————————————————————————————————————–

5. What Could Go Wrong

If productivity gains stall, the consequences could ripple through markets and policy:

• Financial correction: Some AI-heavy stocks are priced for perfection. If usage plateaus or margins shrink, valuations could correct sharply.
• Capital hangover: Circular financing means the same dollar might appear multiple times in different balance sheets. When credit tightens, that illusion unwinds fast.
• Consolidation: Smaller startups or data-center builders tied to over-optimistic demand could be absorbed by larger players.
• Public pushback: If workers see automation without broad benefits, political pressure for regulation or “AI taxes” will rise.
• Energy strain: AI data centers already consume enormous power. Bottlenecks in electricity generation could limit growth or make it costlier.

None of these spell doom—just a reminder that revolutions rarely move in straight lines.

———————————————————————————————————————–

6. What a Real Productivity Boom Would Look Like

If the technology fulfills its potential, here’s what we’d see over the next five years:

• AI tools quietly embedded into everyday software—email, spreadsheets, design apps—without users realizing they’re using “AI.”
• Fewer middle steps: insurance claims settled in minutes, contracts drafted in hours, software built in days.
• Lower barriers for small businesses that can finally afford capabilities once reserved for large firms.
• Measurable macro effects: GDP per worker rising faster than wages, suggesting real efficiency gains.

That’s the optimistic—but plausible—scenario. It just requires patience and smarter implementation than we’ve seen so far.

———————————————————————————————————————–

7. How to Think About It as an Investor or Parent

• Separate the technology from the trade. AI will change the world—but that doesn’t mean every AI stock will make money. The internet changed everything, yet most dot-com investors lost money.
• Look for proof of use, not proof of hype. Companies showing real workflow improvement or cost savings—not just big model launches—are the ones to watch.
• For your kids: Encourage them to learn AI fluency, not just coding. Understanding data, critical thinking, and ethics will matter as much as algorithms.
• For society: The best safeguard against job loss is faster diffusion of useful AI—tools that augment people rather than displace them.

8. The Takeaway

AI is neither salvation nor scam. It’s a powerful new toolkit that, like electricity or the internet, will take years to find its full footing. We’re in the expensive middle stage—high excitement, uncertain payoff, and complex financing that keeps the flywheel spinning.

If history rhymes, the frenzy will cool, weak hands will fold, and durable applications will emerge quietly. The winners will be firms—and individuals—who treat AI not as magic, but as leverage: a force multiplier for human judgment, creativity, and discipline.

In short: build wisely, question the hype, and don’t panic about the robots. The future still needs thoughtful people to make sense of the machines.

———————————————————————————————————————–

I personally found Chat GPT’s output very insightful and useful.  It also made me a lot more productive in reaching my goal, which was to show, through a real-world example, what ChatGPT can do by using AI to explain AI as an investment theme, including its opportunities, limitations, risks, and what it could mean for work and the job market.

Ultimately, AI is best viewed as a robust set of tools that drive productivity and innovation, rather than a magic solution. The excitement and massive capital flowing into AI have created arenas for circular investing and interdependent deals, amplifying both opportunity and risk.

The potential for these technologies to enhance productivity, streamline workflows, and unlock new opportunities is real. But realizing that potential will take time, careful implementation, and continued human judgment. AI can make mistakes. It can misinterpret data or amplify biases, and its effectiveness is always tied to the quality of the inputs and guidance we provide. Far from replacing people, these tools are most valuable when they augment human insight and decision-making.

So, is AI a trick or a treat? It may be both. It offers an exciting opportunity tempered by caution and the recognition that the human role remains indispensable in shaping meaningful, lasting results. The answer depends on how thoughtfully we use it and how patiently we wait for its promise to unfold fully.