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!
Fictional literature, movies, shows, and games have long been filled with stories of sentient machines, but we are living through the first period in history where that fiction is actually becoming a reality. Granted, we’re still in the very early stages, and many people tend to overestimate the current state and capabilities of AI. Odds are if you’re reading this book in 2023, then in your lifetime you probably aren’t going to see a true technological singularity leading to an ultra-futuristic state where superintelligent bots have automated virtually every part of life. However, that doesn’t mean that AI doesn’t already exist (it does), or that the recent developments haven’t been remarkable (they have).
The industrialized world is already highly-familiarized with some of AI’s precursor technologies. In a very basic search framework, you digitally look for text-based items and a rudimentary matching technology tries to find exactly what you’re looking for (this is primitive search, like when you use Ctrl + F on a document). To get more advanced, statistical inferences and models may be layered in to predict what you’re looking for even when you didn’t write it exactly that way (this is contemporary search, like you find in a search engine). Voice recognition can also be integrated to translate between data that exists as sound and data that exists as text, enabling things like “Hey, Siri” or “Okay, Google” followed by some request that meets pre-defined parameters. The concept of language modeling then attempts to understand how people actually communicate in practice in order to make these types of searches and requests more accurate, seamless, and authentic.
But none of this is really ‘AI.’ These algorithms and models, while highly complex and requiring an advanced knowledge of fields like statistics, math, linguistics, and computer science, are still primarily behaving as programs following orders from the code. The intelligence element of artificial intelligence necessarily contains some aspect of learning, growing, or adapting without direct human intervention and control. True, somewhere along the road a human has to set the bot on the path to learning and maybe correct it from time to time, but is that conceptually so different than a parent teaching a child how to learn and adapt? AI has existed for a while now largely in a field called Machine Learning. It has some impressive use cases, but compared to the futuristic visions and expectations of the artists, it is still very new.
In November 2022, a company named OpenAI released a new technology called ChatGPT (Chat Generative Pre-trained Transformer) as a public prototype. ChatGPT is built upon OpenAI’s earlier GPT-3 language model. The company has founding ties to some of the most noteworthy innovators in tech like Elon Musk and Peter Thiel. While ChatGPT has only been out for a few months as of the time of this writing, the world is already captivated by discussions of what possibilities await. That’s because this class of artificial intelligence known as generative AI can actually generate new content (like a person) based on its knowledge base.
The public release of ChatGPT represents a watershed moment for business because it is now so easy to get fast, on-demand answers to or ideas for complex topics while speaking the way you do naturally. You don’t need to think about how to phrase a query into a search engine that will be SEO-tailored to find exactly what you want, then read through countless websites only to find a fraction of the relevant information. Instead, you can now dialogue with an AI the same way that you would a colleague and it will (hopefully) tell you what you need to know. And because the nature of AI is to learn and adapt, new, more advanced use cases will no doubt be developed through experimentation over the coming years.