Beyond the AI Hype – Learning From History
- gvalyou
- 36 minutes ago
- 6 min read
Recent headlines are highlighting the challenges and frustrations companies face as they strive to integrate AI into their businesses. Posts and headlines like "MIT report: 95% of generative AI projects are failing," "Almost all AI pilots are delivering no measurable financial return," and "Enterprise AI hype meets a harsh reality check." Many investors and executives are raising red flags, and rightfully so, and the naysayers and pessimistic prognosticators are coming out in full force. Before declaring AI a bust, let's pause. This moment isn't the collapse of AI innovation. It's a familiar inflection point. Almost every transformative wave before it, from the dot-com to mobile, cloud, blockchain, and big data, has run through phases of inflated expectations, disillusionment, and, ultimately, market maturity. The recent MIT study, which influenced many of these headlines, is a wake-up call, not a verdict.
The Gartner Hype Cycle in Action
The Gartner Hype Cycle has proven itself to be an excellent guide for describing how new technologies progress from the "Peak of Inflated Expectations" to the "Trough of Disillusionment," before gradually climbing toward the "Slope of Enlightenment" and eventually reaching the "Plateau of Productivity."
Figure 1. Gartner - Hype Cycle for Artificial Intelligence 2025
There are Five Phases of the Hype Cycle
Innovation Trigger
A new technology or concept emerges — often from research, startups, or a breakthrough.
Media buzz and early experimentation start.
There are no proven products yet; investment is speculative.
Peak of Inflated Expectations
Hype builds as success stories (sometimes exaggerated) circulate.
Early adopters jump in, and vendors race to position themselves.
Expectations are often unrealistic, leading to "hype over substance."
Trough of Disillusionment
Reality sets in. Failures, missed deadlines, and overpromises lead to disappointment.
Many startups fail or pivot; funding dries up for weaker players.
Survivors focus on refining and proving real use cases.
Slope of Enlightenment
Practical applications begin to emerge.
Second- and third-generation products improve reliability.
Enterprises start adopting based on proven ROI and use cases.
Plateau of Productivity
The technology matures and becomes mainstream.
Adoption accelerates, standards emerge, and benefits are widely realized.
Market growth stabilizes, and the tech is integrated into business as usual.
In some areas, according to Gartner, AI has already reached its peak. Now, as unrealistic expectations collide with execution challenges, many areas of AI are entering the disillusionment phase. This is often when critics dismiss the technology, investors pull back, and weak solutions collapse. History shows that this is also when the strongest, most value-driven solutions and companies emerge and scale.
Lessons from the Dot-Com Bubble
A notable example of the Hype Cycle is the dot-com era, which followed a similar arc. Investors poured money into any company with ".com" in its name, many of which had no clear path to profitability.
I recall vividly sitting in client meetings as a partner and director in a very successful dot-com consultancy, working to assign ROI and payback periods, as we strategized around new innovative solutions and often being told by them, "not to worry about it," "build it and they will come," and "they needed to get to market first, let us worry about how to monetize." At the time, I was concerned about their long-term success and wanted to ensure they were creating sustainable, value-driven solutions.
The market crash that followed this Hype Cycle destroyed massive amounts of capital.
I am not predicting that history will repeat itself in quite the same manner. Currently, there are some striking similarities and considerable hype.
The positive result is that out of the dot-com era came Amazon, Google, Salesforce, eBay, eTrade, PayPal, and several clients, whom I wish I could disclose, that created incredible value and significant market-defining solutions supported by a strong base of business fundamentals. The common thread in their success is that they built sustainable businesses based on new and innovative solutions people wanted, with clear customer value and measurable ROI. The dot-com world didn't fail; the hype surrounding it did.
The AI Parallel
Today, billions, perhaps even trillions, of dollars are being invested in AI startups and internal projects worldwide. Some will fail quickly. Many should never have progressed past Seed or Series A funding, or had their project budgets funded in established companies.
The assertion that a majority of companies flame out should come as no surprise to those with fundamental business knowledge, as most new businesses fail. It doesn't matter the industry or era. A 2024 report from the Bureau of Labor Statistics highlights that only 34.7 percent of business establishments born in 2013 were still operating in 2023.
Figure 2. BLS - 10 Year Business Survival Rates
A closer look at some of the individual business successes and failures, including Uber, Instagram, and Venmo during this period, reveals that long-term survivors succeed by addressing value-based needs while remaining grounded in disciplined business fundamentals. History suggests the same will hold true in the AI era: the winners will be those that survive, thrive, and—through a balance of innovation and execution—deliver value that justifies sustained investment. The era may change, but the principle remains the same.
Four-Point Test
When assessing an AI or any other business solution, I use a quick, top-level framework, a Four-Point Test, that helps gauge whether it has the potential to survive the early hype cycle and achieve long-term viability.
Real Problem: Does it address a genuine problem or need that people or businesses will care about?
Scalability: Can it scale effectively beyond early adopters?
Defensible Growth: Is its growth trajectory and the barriers to entry sufficient to sustain an advantage or competitive pressures?
Return on Investment: Can it demonstrate a payback period or rate of return that aligns with investor expectations and available resources?
Think of it as a version of Shark Tank for AI. On that show, investors like Mark Cuban decide whether a business idea is worth backing. My Four-Point Test applies the same kind of lens. Does an AI or any solution really deserve the investment of time, money, and attention?
Solutions that fail this simple checklist usually stall in the trough—they burn through momentum, money, and interest before reaching maturity. Those that pass cleanly are already on a trajectory toward sustainable growth. And for solutions that land somewhere in between, the message is clear: they must find stronger answers to the questions where they can’t yet say a confident “yes.”
How to Avoid Bad AI Projects and Investments
Leaders can tilt the odds in their favor and ensure their AI solution does not get stuck in the trough by following a disciplined playbook:
Anchor to Business Value – Start with problems worth solving, not experiments in search of justification.
Define ROI Early – Establish payback expectations upfront. If you cannot demonstrate value within a reasonable period, it's a red flag.
Pressure Test Use Cases – Separate features from solutions. Many AI plays are enhancements, nifty magic tricks, not standalone solutions or companies.
Avoid the never-ending "Pilot Graveyard" – Success is not a proof-of-concept; it's integrated solutions that can be applied to the real world to deliver measurable impact at scale. The point being, you have to get it to market.
Invest in Execution Capacity – The biggest failures come not from technology gaps, but from talent, governance, and adoption gaps.
Dream, but Be Realistic – Encourage bold ideas and experimentation, but anchor them in sound market, financial, and operational discipline. Don’t discard creative concepts too quickly if they are not a go today—hold onto them until execution becomes viable. After all, many of today’s everyday technologies once seemed unrealistic, imagined only in the minds of Hollywood screenwriters, yet they now shape our daily lives.
Where Do We Go From Here?
AI isn't dying, it's evolving and maturing. The Gartner Hype Cycle reminds us that disillusionment is not the end, it's the middle of the journey. Just as the dot-com era emerged from its bubble stronger than ever, AI will most likely follow the same path. The path to success is disciplined investment, focused on real value. Those who stay the course, avoid the noise, and invest wisely won't just survive the cycle. They will become the market leaders of the future.
Recognition and Thanks Special thanks to my friends at ProceedAI for reintroducing me to the Gartner Hype Cycle during a recent meeting. If you're not familiar with the Gartner Hype Cycle, I encourage you to take a moment to learn more about it. References and Citations
Tom's Hardware. (2025, June 2). 95% of generative AI implementations in enterprises have no measurable impact on P&L,' says MIT. Retrieved August 2025, from https://www.tomshardware.com/tech-industry/artificial-intelligence/95-percent-of-generative-ai-implementations-in-enterprise-have-no-measurable-impact-on-p-and-l-says-mit-flawed-integration-key-reason-why-ai-projects-underperform
Investors' Business Daily. (2025, June 2). Why MIT's Study On the Enterprise Market Is Pressuring AI Stocks. Retrieved August 2025, from https://www.investors.com/news/technology/artificial-intelligence-stocks-ai-stocks-mit-study/
Gartner. (2024). Understanding Gartner's Hype Cycle. Retrieved August 2025, from https://www.gartner.com/en/research/methodologies/gartner-hype-cycle
AI's current transition from inflated expectations into the trough of disillusionment. https://www.gartner.com/en/newsroom/press-releases/2025-08-05-gartner-hype-cycle-identifies-top-ai-innovations-in-2025
U.S. Bureau of Labor Statistics. (2024, January 12). 34.7 percent of business establishments born in 2013 were still operating in 2023. The Economics Daily. U.S. Department of Labor. Retrieved August 26, 2025, from https://www.bls.gov/opub/ted/2024/34-7-percent-of-business-establishments-born-in-2013-were-still-operating-in-2023.htm
Yahoo Finance Website. (2019, December 6). The 16 most innovative new companies of the 2010s. Retrieved from https://finance.yahoo.com/news/the-16-most-innovative-new-companies-of-2010-s-190021422.html
Images and Media
Cover image, Wix Stock Image Photos, August 24, 2025
AI Tools
Greg Valyou—me, a real person, wrote this article. In addition to drawing on my experiences, knowledge, and research, I utilized AI tools to augment the process.
Grammarly. https://www.grammarly.com. Accessed August 22-25, 2025
Spelling, grammar, sentence structure, and plagiarism checks
ChatGPT (4o). https://chatgpt.com. Paid Account, Apple Store Application for Mac. Accessed August 22-25, 2025
During draft creation, reviewed suggested options for how to adjust some paragraphs for clarity