The AI Big Crunch Is Starting

The AI emperor has clothes. Just much less than everyone thinks.

Ben Rothke

3 min read

·

8 hours ago

--

The Bang Comes Before the Crunch

Everyone knows about the Big Bang.

Not the television show, but the scientific theory describing the explosive expansion of the universe from an unimaginably dense point nearly 14 billion years ago.

Far fewer people have heard of the *Big Crunch — *the hypothetical end of the universe where expansion eventually stops, gravity wins, and everything begins collapsing back in on itself.

While cosmologists largely believe that won’t happen to our universe, the metaphor fits artificial intelligence in one crucial way: the collapse is coming after the expansion.

We’ve lived through the AI Big Bang. Now the AI Big Crunch is beginning: the phase when hype gives way to limits.

The AI Gold Rush

Today it’s almost impossible to buy a product that isn’t “AI-powered.” Every software vendor claims to have AI. Every board wants an AI strategy. Every executive fears being left behind.

The problem is that many organizations have confused adopting AI with understanding AI, and that confusion is already producing costly mistakes.

One of the best books separating AI fact from fiction AI Snake Oil: What Artificial Intelligence Can Do, What It Can’t, and How to Tell the Difference by Arvind Narayanan and Sayash Kapoor of Princeton University.

Their premise is refreshingly simple: AI is extraordinary at some things, mediocre at others, and completely incapable of many of the tasks marketers claim it can perform.

They define AI snake oil as technology that cannot work as advertised. That’s not a criticism of AI. It’s a criticism of those selling unrealistic expectations.

Reality Arrives

Eventually, reality catches up with hype. Forrester found that more than half of employers now regret AI-related layoffs, showing that many organizations eliminated employees based on what they hoped AI would eventually accomplish — not what it could actually do.

Robert Half reached a similar conclusion. Nearly one-third of hiring managers admitted eliminating positions because of AI, only to hire people back when reality intervened.

Perhaps the clearest example comes from Ford. The company recently brought experienced engineers back into critical quality processes after discovering that AI could not replace decades of engineering judgment.

Charles Poon, Ford’s Vice President of Vehicle Hardware Engineering, put it perfectly: “Artificial intelligence is a fantastic tool, but it’s only as good as the information you use to train it.” He went on to acknowledge that Ford had undervalued the experience of veteran engineers who had seen multiple product generations.

That’s a remarkably honest admission. It’s also one many companies will eventually make. According to Bloomberg, correcting those mistakes is expected to cost Ford billions. That is an expensive reminder that replacing expertise with algorithms isn’t always progress.

Bruce Schneier — Right Again

Bruce Schneier famously observed: “If you think technology can solve your security problems, then you don’t understand the problems and you don’t understand the technology.” He was talking about cybersecurity. He could just as easily have been talking about AI.

Technology doesn’t solve poorly understood problems. It merely automates them. Organizations that rush into AI before clearly defining the problem often end up solving the wrong one — faster.

AI Is a Tool, Not a Strategy

None of this means AI is overhyped. Nor does it mean AI has failed. Quite the opposite. Generative AI is arguably the most significant advance in information technology since the Internet became mainstream. It will transform nearly every industry.

But transformational technologies follow a familiar pattern. They generate enormous excitement. Then inflated expectations. Then disappointment. Finally, they settle into practical, valuable use. We’re entering that third phase: the correction from hype to reality.

The companies that succeed won’t be those that bought AI first. They’ll be the ones that understood where AI creates value — and where it doesn’t. The lesson isn’t to avoid AI. The lesson is to stop believing that AI is the strategy. The thesis is simpler: AI is a tool, not a plan.

It’s just another tool. An incredibly powerful tool. But still just a tool.

Surviving the AI Big Crunch

The organizations that survive the AI Big Crunch will be those that remember a simple truth: Technology has never replaced sound judgment. It has only amplified it.