I spend a lot of time in meetings talking about how companies can get the most out of artificial intelligence. These conversations almost always include a deep-dive into technology features and requirements, as well as long digressions into the intricacies of data management. Although tech and data are critical to any AI implementation, what’s often overlooked in these discussions is the third essential ingredient for success with AI: human curiosity.
Today, companies are using machine learning and AI applications to save time and money by automating tasks once done manually by humans. However, the transformational potential of AI goes far beyond operational automation. The challenge facing senior management teams today is elevating AI from an efficiency tool to a source of strategic insight and competitive advantage. That’s where human curiosity comes into play.
The recipe starts with the right tech
It sounds funny to say, but technology might be the most straightforward element of a corporate AI initiative. Companies and consumers have access to a wide and expanding list of generative AI (GenAI) applications, from free web-based platforms that individuals can use to costly corporate versions.
When choosing among these options, companies should select providers with enterprise-level tools that are robust and secure enough to meet their needs. For example, when LTX and Broadridge built BondGPT, a GenAI-powered app that answers bond-related questions, we chose OpenAI GPT-4. Our partnership with OpenAI and use of GPT-4 helped us gain access to a large language model (LLM) well-suited for researching the bond market in a matter of minutes, down from the hours of manual work it took before.
No matter which LLM you choose, the company you partner with will have to provide a custom implementation to fit your business goals and existing tech stack. Your business will also incur new costs as a part of the process. A recent Harvard Business Review article found that the biggest challenges and costs companies encounter when adopting AI include fine-tuning the LLM, fine-tuning the amount of prompt engineering the model requires to answer questions, and expenses related to operations and talent, among others. And of course, companies will have to feed their new AI solutions with massive amounts of reliable data.
Fold in real-time data for a smooth base
GenAI runs on data. Providing AI applications with the volume and quality of data they need to generate reliable results can be a challenge. At the organizational level, bad data doesn’t just diminish the quality of the AI application’s output, it also introduces huge risks in the form of inaccuracies and hallucinations. For financial services firms and other companies operating in fast-paced markets, the need for real-time, up-to-the-minute data is paramount with AI. Monthly, weekly, or even daily updates often won’t cut it.
For this reason, large-scale AI implementations often require a preliminary overhaul of a company’s data management system and the creation of a new platform that efficiently collects, processes and normalizes data from across the organization. In most cases, that real-time data will include proprietary data that must be protected, and customer data that must be anonymized.
Add the secret ingredient: human curiosity
Companies able master these technology and data requirements are starting to benefit from the huge efficiency gains possible with AI. I’ve become a strong proponent of AI as a time-saving device. Recently, I received three reports from different business teams on a particular issue. I was asked to present an overview of the issue to our management team. As an experiment, I uploaded the reports into a GenAI application and asked it to create a summary. The result was quite good and, after a bit of editing, I was able to use it as the basis of my presentation.
But AI isn’t just a time-saver. This tech can help identify changes and trends in demographics, technology, financial markets and a host of other realms with the potential to impact business, so long as we tap into our human curiosity. A recent Fortune article documented how U.S. government policies during the past 40 years helped shift $129 trillion dollars into the pockets of baby boomers and older Americans at the expense of both younger generations and the federal budget. A wealth transfer of that magnitude has implications for virtually every type of business. That’s precisely the type of trend AI can uncover—if we present it with the right questions.
In the television series “Ted Lasso,” Ted wins a bet against his rival, Rupert, in game of darts. He beats Rupert — a ringer who uses his own custom darts set — by pretending to know nothing about the game. After Rupert insults him throughout the contest, Ted reveals his dart skills and, just before hitting the winning shot, reminds Rupert of the quote often attributed to Walt Whitman: “Be curious, not judgmental.”
That scene resonated with me because it illustrates the importance of human curiosity. If Rupert had been curious, he would have asked Ted questions, such as “have you played a lot of darts?” and could have avoided a humiliating loss. In the same way, business executives who approach AI with curiosity and ask the right questions can generate truly transformational results.
Technology and data provide the operational foundation for AI. Once that’s in place, the question is: what do we ask of it?
Read the full article here