Warnings About Generative AI

This material is excerpted from the introduction to my new book, Bots in Suits: Using Generative AI to Revolutionize Your Business, a practical demonstration of how to converse with a generative AI chatbot (like ChatGPT) for business use. The ebook is on sale now at various retailers (with print coming soon), but for the time being, subscribers to Customers, Costs, and Cash Flow can get it for free!

Because generative AI operating at the level of ChatGPT is so new, you need to tread lightly. It would be a dire mistake to read this book and go, “Wow, now I don’t need to pay all these lazy employees/contractors/vendors anymore! You guys are all outta here!” and start upending major parts of your business with layoffs and restructuring. If that’s your approach, you will wind up fundamentally undermining your institutional strength over the long term, and it may actually lead to the destruction of your business. Then again, if your attitude as a leader is one of treating the people around you like disposable cogs in a machine whose chief aim is to maximize your personal benefit, then it’s probably a good thing if you go out of business. But let’s assume that’s not you; that you’re a leader who really does care and wants to do good. What are the risks of moving too fast with this?

First, we don’t know exactly how this technology will develop or what competitors will enter the space. For example, Google is already said to be working on its own generative AI, as well as detection algorithms to penalize other AI-generated content in its search rankings. We also don’t know how regulators will respond, or what it will cost once the technology goes out of public prototype and into a paid enterprise version. For reference, OpenAI CEO Sam Altman recently referred to the costs to operate ChatGPT as “eye-watering” due to the substantial computational power it requires. While it’s probably safe to say that generative AI is here to stay and will only get better in terms of core product, there’s still a lot of unknowns.

Second, AI can make mistakes the same as a person, so just because a bot tells you something confidently doesn’t make it correct. Artificial intelligence learns from the same type of data as a sentient being, and that means it can learn things incorrectly. The technology in its current state especially seems to struggle at handling numerical situations and tends to do a lot better with qualitative material, though even with the qualitative it sometimes responds by stating the obvious (or even the ridiculous). Many times it will try to mimic your conversational tone and end up repeating things back to you. This technology is not an excuse to set aside your own critical thinking, intuition, and experience. The responsible use of AI requires human supervision and judgment. You need to check the material, and sometimes that means challenging the AI.

Third, it’s highly unlikely that this technology will replace traditional search engines, though it will almost certainly pressure them to change in significant ways. Generative AI is currently best if you have a focused knowledge or idea-based query. On the other hand, if you’re looking for something more subjective, philosophical, obscure, localized, or brand new (like news and current events), or if you want to see many different results and research things for yourself, then a search engine is still your best bet. I recommend Brave Search, which contains some of the most important innovation work happening in the search engine space. Due to the computational constraints and expenses that go into generative AI, and the broader nature of how the web operates both technically and socially, I don’t expect that search engines will go away any time soon, if ever.

Fourth, there are major social and human ramifications for the introduction of generative AI. You’ve probably heard discussions of dystopian scenarios where AI puts everyone out of work so people just sit around and starve to death because nobody can earn any money. While there are real dangers to AI which need to be considered by society at large as this technology develops, we can be pretty confident that this particular hypothetical future isn’t going to happen. To illustrate: a thought experiment posits that a hungry dog is positioned between two bowls of food which are identical in every way and situated at precisely the same distance from the dog. Will indecisiveness triumph as the dog dies of starvation? Of course not; the dog’s will to survive will ultimately drive it to make a decision and choose a bowl (though after it finishes, it may try and go back for the second!).

Just because a bot may eventually be able to do things like complex accounting, teaching, lawyering, programming, engineering, marketing, cooking, investing, doctoring, researching, drawing, farming, cleaning, building, selling, or even managing, that doesn’t mean we’ll collectively en masse decide to not find food or shelter, establish communities, or go about our lives; it just means the natural organization of the economy will shift. It is true that breakthrough technology always changes things up: some jobs become obsolete, resources get reallocated, and people will need to re-skill for the new economy. That’s about to happen again.

The use of AI is an emerging business skill that will be one of the most crucial things you need to know, starting now and lasting for the rest of your life. Simply by reading this book and taking the initiative, you’re getting in on the ground floor of one of the next great seismic shifts in the history of civilization. Because you’re learning and preparing, you’ll be able to leverage these changes proactively rather than being stuck reacting.

The best business people will grow to understand both the strengths and limitations of generative AI, and how those strengths and limitations change with time. As we move forward, AI will likely start to replace functions and personnel that depend primarily on possessing static knowledge or engaging in rote tasks. We are moving away from what Peter Drucker famously described as the knowledge worker economy and into what James Altucher refers to as the idea economy. In other words, expect that the executives, managers, professionals, consultants, contractors, employees, and businesses whose value proposition is mostly that they are traditionally credentialed, know about something you don’t, or can do something you can’t (whether by technical constraint or arbitrary decree) to slowly phase out, while those who center their value around the fact that they can create and build and solve will thrive. Make sure you’re in the latter category.

Don't miss any Customers, Costs, and Cash Flow!
Practical business insights delivered to your inbox + access to the article archives. Absolutely no spam.