AI for Small Enterprises: Conclusion

In this final part of our series, we reflect on the state of (Generative) AI from the perspective of small enterprises and share our personal perspective.

What’s in a Name: GenAI

Over the past four parts of this series, we’ve explored what "Generative Artificial Intelligence" (GenAI) is and where its strengths and weaknesses lie, how it can be used for creative writing in everyday business, how it can assist in idea development, and how it can search and aggregate data.

In all these articles and examples, we’ve been careful not to use GenAI for generating factual content or solving logical problems. We consider it a misappropriation of the technology. Its essence lies in the name: Generative Artificial Intelligence. Generating creative content by transforming previously known texts, images, and music is the core competency of GenAI.

Any use beyond that is likely to fall short of the expectations fueled by the current hype. This holds true for personal use, when GenAI is employed to solve arithmetic problems, as well as for businesses trying to mask a shaky data foundation with the use of (Gen)AI.

The Hype and the Bubble

Speaking of hype: Gartner already saw GenAI reaching the peak of the AI Hype Cycle in 2023 https://www.gartner.com/en/articles/what-s-new-in-artificial-intelligence-from-the-2023-gartner-hype-cycle. On the stock market, there are rising concerns that GenAI might be a bubble on the verge of bursting https://www.bloomberg.com/news/articles/2024-07-18/goldman-s-top-stock-analyst-is-waiting-for-ai-bubble-to-burst. Comparisons are even being drawn with the dot-com bubble of the late ’90s https://www.reuters.com/markets/echoes-dotcom-bubble-haunt-ai-driven-us-stock-market-2024-07-02/https://finance.yahoo.com/news/stock-market-crashed-dot-com-091200099.html.

However, from our perspective, the current sentiment around GenAI is different from the dot-com bubble. Instead of misguided investor speculation, we see massive investment by large tech companies that is not yet justified by profitable use in business. Recent reports from Deloitte https://www2.deloitte.com/content/dam/Deloitte/us/Documents/consulting/us-state-of-gen-ai-q3.pdf</justament-reference aria-label="Reference to article 'The state of AI in early 2024: Gen AI adoption spikes and starts to generate value' on https://www.mckinsey.com"> and McKinsey https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai indicate that we’re just now beginning to see returns on this investment.

We (along with others https://www.washingtonpost.com/technology/2024/07/24/ai-bubble-big-tech-stocks-goldman-sachs/) suspect that the pragmatically relevant development of GenAI will follow the Gartner Hype Cycle: the AI-driven stock market may experience a downturn, but the underlying technology will become a steady part of everyday (work) life; as businesses find the best use cases for it.

So, it’s likely not a bubble, but rather an (expensive) discovery phase.

Why GenAI is Nevertheless Disruptive

This view might seem like a stark limitation when compared to the contemporary marketing around AI. Nevertheless, we believe that GenAI is a disruptive technology that can usher in a technological shift. The disruption doesn’t lie in GenAI answering all our questions or making us all redundant. Instead, we hypothesize that GenAI is disruptive in two key aspects:

• GenAI is a new interface between humans and machines. Natural Language Processing (NLP) and voice recognition have long worked on enabling inputs through natural language — now, GenAI has succeeded in facilitating coherent conversations with virtual personalities in natural and diverse language. This could open the door to computer systems that were previously only accessible with expert knowledge.
• GenAI provides technological support for creative processes. By generating new connections and combinations of ideas, GenAI could render the status quo in innovation management obsolete https://www.sciencedirect.com/science/article/pii/S0148296324000468.

Internet Research and SEO: The First Small Disruption by GenAI?

Beyond how big tech companies use GenAI for business innovation, marketing, or customer service, there’s a (small) shift in the area of internet search that will affect us all. As we hinted in our previous post, searching for information via GenAI chatbots like ChatGPT, Perplexity, or Bing/Copilot represents a new paradigm; where information is asked for through conversation rather than sought by keywords. This shift, in itself, has the potential to simplify access to information or speed up the process of information retrieval.

Gartner predicted earlier this year https://www.gartner.com/en/newsroom/press-releases/2024-02-19-gartner-predicts-search-engine-volume-will-drop-25-percent-by-2026-due-to-ai-chatbots-and-other-virtual-agents that traffic on keyword-based search sites will drop by 25% by 2026, due to AI chatbots and other virtual agents. If this scenario unfolds, where GenAI-supported search engines gain a significant market share, it would have implications for Search Engine Optimization (SEO) and user behavior on websites. Asking for information through GenAI is tied to the concept of Zero Click Results, where the user’s query is answered without ever visiting the page that originally provided the information. As a result, long tail keywords will become more relevant: visitors will visit a website for specific information that is no longer covered by the AI-aggregated content https://www.semrush.com/blog/google-sge/.

Moreover, the aggregation of search results by AI brings with it an aspect that we should critically examine: the natural limitations of GenAI also apply in the context of internet search. GenAI introduces factual errors when summarizing search results https://arxiv.org/pdf/2304.11076 and affirms biases https://arxiv.org/pdf/2304.09848. These limitations become more significant because the application of GenAI is partly beyond the user's control. It’s not a conscious decision by the user when Bing or Google display GenAI-aggregated content on the results page. We see this as a risk of users lacking awareness regarding the nature and quality of the information they consume.

Conclusion

With this commentary, we conclude our series "AI for Small Enterprises." We aimed to highlight the use cases for Generative AI in everyday work and how you can test them without making investments. Hopefully, we’ve also managed to familiarize you with the basics of GenAI along the way.

In line with the nature and goal of the posts, we’ve focused on general-purpose chatbots (i.e., those not trained on specific datasets), which are publicly available and don’t require infrastructure on premise. Beyond that, there are, of course, a vast number of specialized (Gen)AI applications. One aggregator currently lists over 14,000 AI tools https://theresanaiforthat.com/. Feel free to explore beyond the major GenAI chatbots like ChatGPT, Claude, Copilot & Co - but always keep in mind our guidelines for interacting with GenAI chatbots on the internet.

We will certainly follow the development of GenAI with interest and are curious to see if it truly turns out to be a disruptive technology for human innovation. Until next time!