Consider the scenario of an attempted break-in at a warehouse this year. Four suspects with crowbars and other tools are jimmying the doors of a commercial storage area at 1am before they are detected by security cameras and then scared away by the guards on duty. No arrests are made, but of course, the guards on the premises give first-hand accounts as soon as feasible. That information will be valuable to both the company that was targeted by the would-be robbers and, most especially, the security firm that was hired to protect it. Â
These guards are now using tools powered by generative artificial intelligence (GenAI), provided by their employers, in their reporting. They can speak as naturally as addressing colleagues over the phone as they make their onsite reports. The recordings of their voices can be cleaned up and professionalized with a click. Curse words removed. Vocabulary clarified. And there are prompts to make sure critical data is captured: “Do we have the exact time? Location?” And so on. Â
With standardized incident reports that capture critical information in clear language, consider what a security company can do. It can offer a better service, price more appropriately, and optimize the number of cameras and guards in place. Better living through data. Â
This is clearly a positive use case for AI. And yet, I warn my students and the companies I work with around the world that the power of this case comes from the ground up; it’s an important but incremental improvement in operations, not a radically new way of conducting business. The security company was already collecting data to improve its service, and the latest GenAI tools have simply improved its collection process. Â
Boring is beautiful with today’s GenAI. The biggest disappointments I have witnessed have been when big legacy companies have gone all-in on ambitious and expensive GenAI projects where the end goal is unclear. One company I know has a sophisticated, proprietary GenAI system that I dare say is seldom used beyond summarizing meetings or checking transcriptions. That is not earth-shattering. Â
To reap the advantages of GenAI without falling prey to disappointment and overspending on the hype, here are my four pieces of advice:Â Â
Even the best GenAI tools today are not going to unlock much value for your organization right out of the box if your data systems are not in great shape. To do that, your internal data must be clean and governance infrastructure in place. AI projects require thoroughly thought-out process engineering. I can’t emphasize this enough. Take time to think through your systems and ways of working. Your organization’s infrastructure should allow its data to speak with other relevant data from different starting points and systems. Do not expect positive results if you dive in without doing your data (and governance) due diligence first.
Take it easy, especially at the beginning. What I have learned from working with organizations on three continents is that GenAI implementations are turning out to be more difficult than expected. To avoid disappointment, start humbly. Create quality transcriptions of recordings, summarize long, complex texts, translate documents, and be on the lookout for other sensible use cases.
Some companies have found out the hard way that their new AI-powered tool is not ready for prime time.
This is not a technology that’s going to work if it’s pushed from the top down. You are not going to implement AI by diktat. Get the tools in the hands of the people who are working in the field. This is how more value will be captured.
First, offer AI training to improve your staff’s understanding of the technology and its limitations. Then, make sure you have oversight and risk mitigation in place. Some companies have found out the hard way that their new AI-powered tool is not ready for prime time. I’m thinking of a car dealership in the United States whose GenAI-powered chatbot, acting as a sales agent, ended up selling a car worth $58,000 to a savvy customer for $1.
As to risk mitigation for content creation, keep the following three actions in mind. First, ensure you align your organizational values with AI principles by including transparency, fairness, accountability, and safety in your AI use. Second, make it mandatory that all entities creating content with AI use a watermark to label that output as such. Third, create a controlled environment within the organization to fine-tune the use of AI and avoid leaking damaging information.
To capture value over the longer term, be cautious.
The companies that I see achieving real efficiency gains and extracting positive outcomes from GenAI are the companies that have their bases covered and start with a portfolio of simple problems. In a democratized approach to tackling problems from the ground up, they offer access to the tool to a broad swath of their employees.
Boring is beautiful is my mantra because I see too many cases where the latest GenAI tools are lowering operating efficiency in the near term. To capture value over the longer term, be cautious. Think of the security services company I mentioned before. Higher quality and more accessible data from incident reports improve its offering. With good data governance and AI tools in the right hands, your company can avoid the crime of wasting GenAI’s potential.
AI x 9: This article appears in a nine-part summer series that examines how AI is impacting leadership and business, produced in collaboration with Expansión.
IMD’s Simon J Evenett, Johannes Fritz, and Tommaso Giardini argue that the Trump administration will take a more sophisticated approach to AI regulation than some expect ...
GenAI is becoming a more common feature of the supply chain function, but significant areas of opportunity remain, suggests IMD’s Carlos Cordon. ...
AI is rewriting the rules of hiring – job seekers and recruiters alike must adapt to a process driven by algorithms but grounded in human connections. ...
Breakthroughs in agentic AI, open-source development, and closer collaboration between AI companies and the defense sector have significantly raised the AI risk profile. These developments have compelled us to advance the AI...
Explore first person business intelligence from top minds curated for a global executive audience