By Matt Fitzpatrick ’25 Principal
The Latest Chapter of AI
ChatGPT-4, the latest update of the artificial intelligence (AI) tool created by OpenAI, has recently gone live and it’s only gotten more impressive. The AI’s capabilities have now expanded to interact with images and creative writing. Before it could write papers, implement code, and complete homework for students. Now, it can analyze paintings and writing styles, write musicals and plays, and even compose symphonies. The possibilities of this technology are endless, and AI will play a huge role in the future of global education and the global economy. As to be expected, the hype around AI has never been greater. This has led to “AI” becoming a common buzzword, a word or phrase that is fashionable or desirable at a particular time, which has carried over and taken control of the startup industry.
Looking at Past Experiences
This level of buzzword takeover in the startup industry has happened before in the late 1990s and early 2000s during the “.com bubble”. With the emergence and rapid growth of the internet, an absurd number of startups started adapting the “.com” term to their company operations. This led to internet-based companies or startups with “.com” in their name or marketing strategies, seeing their valuations skyrocket even with little to no revenue. Some companies had billion-dollar valuations or even went to market with IPOs (Initial Public Offerings) despite never turning a profit, having no finished product, or having any proprietary technology. This phenomenon showed the power of buzzwords. Returning to the present, we see a similar occurrence. This time, the buzzword is “AI”. Startups are using the term “AI” to capitalize on the hype around the technology to attract investors. Such investors will offer lucrative deals based on the potential perception of the technological sophistication of the product. Additionally, founders use the term to give their startups a competitive edge and differentiate themselves in their designated industries.
Misleading Potential Investors
In comparison to the “.com bubble” fiasco, the buzzword itself is not the problem, it is the over-usage and misrepresentation of the term that is causing issues. Numerous startups are using the term despite their company having no functioning AI. In many cases, these companies’ products and services are powered by simple algorithms, rule-based systems, or what-if models designed in Microsoft Access or Excel. In reality, these startups’ tech stacks lack machine learning, a key subset of AI that allows a system or program to learn from experience, which ultimately distinguishes AI from simple algorithms. A study by MMC Ventures analyzed over 2,800 AI startups across Europe and found that only 60% of them currently have a functioning AI. Here in the United States, a report by the Harvard Business Review found that 85% of AI startups in the U.S. have “little to no AI involvement”. The number of founders misleading or overhyping their AI will lead to problems down the line.
Implications and Risks of Overhype
Whether founders are purposely misleading investors or simply they are misinformed about their own AI, these fallacies are leading to unrealistic expectations from investors and customers. When startups overhype their tech stack or their technologies’ capabilities, investors and customers may expect too much too soon, leading to disappointment, a loss of trust, and poor relationships among the parties involved. This is further exacerbated when companies become overvalued, or their valuation increases exponentially, due to the hype around AI. For example, in 2017 the start-up, Outcome Health, advertised a product that integrated AI to improve patient outcomes. The company was valued at $5.5 billion until it was discovered that the company misled investors about the effectiveness of its so-called “AI”. After the truth was revealed that their “AI” was ineffective along with other infractions, the company’s founders were charged with fraud and its valuation plummeted. Outcome Health is one of the many examples of severely overvaluing a company based on misleading hype of technology instead of genuine innovation, leading to unrealistic expectations for performances. In conclusion, investors must dive deep into the tech stack and pay attention to what is under the hood while asking themselves, is this really AI?
Matthew is a junior from Hudson, NH pursuing a B.S. in Information Systems & Business Analytics, and a minor in Data Science. This is his second year with the fund, and he is excited to continue to learn private equity investing and the startup industry through first-hand experience. Matthew is part of the Paul Scholar class of 2025, which recognizes a cohort of some of the most academically promising students enrolled at the Paul College of Business and Economics. Outside of campus, he is a Beta Gamma Sigma International Business Honors Society member. In the past year, Matthew completed an internship at Reveneer, as part of their Business Technology team and is continuing to look forward to his career in Data Analytics.