Trying AI is easy. Figuring out how to make it useful to a business is a different story.
Companies have more AI tools available than ever. Employees are using chatbots to get through everyday tasks. Marketing teams are creating content with AI. Developers are turning to coding assistants, while customer service teams are testing ways to automate routine requests.
The experimentation has been useful. It has also raised a bigger question: What happens when the experiment is over?
Businesses are now trying to figure out where AI actually belongs. They have to decide how much control to give it, how employees should work with it and whether the money being spent on AI is producing anything meaningful in return.
Those questions will be part of a September 9 panel discussion at Pera Soho in New York City. Business and technology leaders will talk about what they have learned from putting AI to work, what has not gone as planned and the decisions companies are facing now. Jason Rosenfeld, Chief Growth and Alliances Officer at NewRocket, will be among the panelists. The conversation comes at an interesting point in the AI boom. The excitement is still there, but businesses are becoming more realistic about what it takes to make the technology work.
Moving Beyond the Experiment
The early days of generative AI were filled with experimentation. People asked chatbots to write emails, summarize documents and brainstorm ideas. Marketing teams found new ways to create content. Developers started using AI to help write code. Customer service teams looked at ways to automate responses. It was a fast-moving period, and there was plenty to learn.
One lesson has become pretty clear: A successful AI demo does not automatically make a successful business solution. The strongest projects usually start with a problem rather than a piece of technology. What is taking too much time? Where are customers getting frustrated? Which repetitive tasks are slowing employees down? Those questions can lead to better AI projects than simply looking for somewhere to use the latest tool.
A company does not need dozens of AI initiatives to make progress. A few projects that solve real problems may end up being much more valuable.
Measuring What AI Is Really Worth
Figuring out the return on an AI investment can be tricky. An AI tool might help an employee finish a task in 10 minutes instead of 30. That sounds like a win. Then the employee spends another 15 minutes checking the work and fixing mistakes. Suddenly, the savings look a little different.
There are other costs, too. Employees need training. Systems need to be secured and monitored. Someone has to maintain the technology when things go wrong.
The price of the software is only part of the equation. Leaders need to look at whether AI is actually cutting costs, improving customer service, increasing revenue or giving employees more time to focus on work that requires experience and judgment.
Without clear ways to measure those results, it becomes difficult to know which projects are worth expanding.
Data and People Still Matter
AI may be the new part of the equation, but some familiar challenges have not gone anywhere.
Data is one of them. Many businesses have plenty of information, but it is often spread across different systems, outdated or difficult to access. That can quickly become a problem when a company tries to expand an AI project. People matter just as much.
Employees still need to understand what the technology is doing, check its output and know when to step in. The opportunity is not always about replacing people. Sometimes it is simply about giving them more time to focus on customers, decisions and complicated problems.
How Much Control Should AI Have?
There is a big difference between AI suggesting an action and taking that action on its own. A company might be comfortable letting AI schedule a meeting or sort incoming requests. It may feel very differently about letting an AI system approve a payment or access sensitive information without human review.
AI agents make this question even more important. They can take actions across multiple systems with less human involvement. That could make businesses more efficient. It could also create more risk. Companies need to decide what AI can access, what it can change and when a person needs to take over.
The first stage of the AI boom rewarded companies that moved quickly. The next stage may reward companies that know where to focus. The companies that succeed may not be the ones using AI everywhere. They may be the ones that know where it can help, where people still need to be involved and how to tell whether it is actually working. The question is no longer whether AI can change business. The harder question is what businesses will do with that change.
