Waiting for the AI crash to happen

It seems that AI is heading toward a crash like the dot-com crash before

We’ve seen this pattern before. The late 90s were full of companies in the early days of the internet rushing to attract huge investments with little value or business impact. Eventually, the bubble burst. Many collapsed, while a few giants became stronger.

AI today shows a lot of similar signs. But these are not just speculations, but are based on patterns. Here are the main pillars that show how the bubble is forming:

🔷Lack of a proper broad organizational AI strategy

Most organizations rush into AI tools without a clear strategy. They experiment with tools or pilots but fail to align AI with business goals, processes, culture, and measurable outcomes. Without strategy, AI is just a technology experimentation.

🔷Very high failure rate of AI projects

MIT reports that only 5% of AI projects reach production successfully. Even then, success depends on whether they create real ROI, not just technical deployment. That small fraction shows what separates hype from value: a strategy, strong adoption, measurable impact, and integration into processes.

🔷Some companies have already gone bankrupt, gone out of business, or couldn't endure costs

We’ve already seen companies failing or unable to prove value: **Humane’s AI Pin, Builder.ai, Stability AI, Retrato, Inflection...etc. Like the dot-com crash, many AI ventures burn cash faster than they generate returns.

🔷Lack of quality data for AI model training

AI systems are only as good as the data behind them. Models now rely on synthetic training data, and quality degrades over time. Models begin learning from their own noise and hallucinations, which leads to inaccurate results.

🔷Massive investment with low returns

More than $500 billion has been poured into AI companies and startups in the last 3 years, but profits remain unseen. NVIDIA is the exception, supplying the infrastructure, yet even its shares are very volatile for AI news or rumors. For most AI firms, the cost of running large models far exceeds the revenue they generate. Even the big players like OpenAI are losing millions in daily operations.

🔷Inconsistent and unreliable outputs

Hallucinations and mistakes are still common, even in the most advanced systems. Businesses cannot rely on tools that may produce brilliant insights one moment and confident errors the next unless built in the right way and based on the right data checks and monitors.

🔷Many AI companies are riding the trend/hype without practical impact

Many companies are simply rebranding or building wrappers around existing models. They ride the AI wave without offering real differentiation, disruption, or value.

🔷The hype of one-size-fits-all automation

True intelligence is specialized, yet many vendors push generic “AI agents for everything” platforms. Broad, unfocused automation rarely solves specific problems deeply enough to deliver business value.

The AI crash might actually be needed now. The lesson from the dot-com crash is not that technology disappears, it’s that it takes away the noise, leaving only the businesses with strategy, vision, resilience, and impact.

AI will follow the same path. The bubble may burst, but organizations that align AI with real strategy will become stronger.

---

If your organization wants to adopt and utilize AI in the right way and create AI strategies that deliver measurable value, let’s talk.