Why Entrepreneurs Still Need To Leverage People (Even With AI)

This story originally appeared on Under30CEO.com.

The internet is getting excited about the prospect of AI. After the release of ChatGPT 3.5 last November, we’re still very much in the early stages of people working out what’s possible with the technology. Everyone wants to know how far it can be taken.

The prospect of AI becoming a superintelligence is leading many to wonder whether it could present a new form of entrepreneurship. Historically, business leaders would leverage a combination of people and capital to push their enterprises forward. In the last twenty years, that’s expanded to “networks,” but in the future, it could involve AI, too.

You can understand why AI has so many people excited. As some commentators are putting it, it’s like having legions of highly intelligent grad students at your fingertips, ready to work for you for free.

But, of course, it’s not quite that simple. While AI might be impressive, it can’t replace people yet. Plus, there’s an argument to suggest that many consumers don’t want it to replace people. Having a human in the loop is beneficial and makes them feel like they are getting a better service.

For this reason, we’re still very much back at square one. While AI might be useful in some circumstances, it is not a wholesale replacement for labor.

Emotional intelligence

One of the reasons humans still need humans is the emotional intelligence they offer. People tend to be excellent at getting on the same wavelength as other people and understanding their needs.

That’s what live virtual receptionist services, such as Go Answer, specialize in doing. These services believe in the power of having a real person available to speak with customers. It’s a key differentiator.

“The whole point of our services is to ensure businesses have somewhere there to take their calls, even when the owner or senior management is asleep,” the company says. “Having a human in the loop enables firms to display their emotional intelligence even in the middle of the night instead of forcing customers to speak to robots, which they might not want to do. The great thing about human agents is their flexibility. General intelligence lets them understand even the most convoluted customer requests. A machine might not be able to do that, or use its executive function to contact someone who knows the answer.”

While machines are developing a theory of mind and getting to grips with human emotions, they tend to lack the intelligence that a person could offer. Knowing what to say and how to say it remains a challenge.

Complex decision-making

Related to this, human beings can also better make complex decisions. Members of staff can weigh up all their options and ask which is the best from a range of alternatives, given some available options.

Machines can do something similar, but, again, they may lack relevant context. They might also misunderstand various stakeholders’ objectives and requirements which could skew their decision-making in the wrong direction.

Related: 7 Ways to Use ChatGPT at Work to Boost Your Productivity, Make Your Job Easier, and Save a Ton of Time

Again, entrepreneurs require humans for these purposes. So-called “agentic” AIs simply aren’t at the stage where they can direct production or manage business processes. Bottlenecks in their functionality and lack of embodiment mean that we’re still years away from these kinds of applications.

Corner cases

Entrepreneurs still need human beings for corner cases, too. AIs depend heavily on a normal distribution of outcomes, but the real world doesn’t always behave in that way. Instead, there are extremes of the distribution and corner cases with qualitatively different solutions.

Humans are great at these “exception to the rules” scenarios, but machines are not. While they can handle the former, they do a poor job of the latter.

“We found that when customers talk to robots, they break easily,” Go Answer says. “Even minor deviations from standard questions produce frustrating replies that don’t really deal with customers’ concerns. Eventually, many become so frustrated that they simply give up and quit the service entirely. But with humans in the loop, this simply doesn’t occur. A person can tell that a customer is becoming frustrated and they feel emotionally compelled to resolve it. If a query isn’t being resolved, they find ways to approach it from different angles or arrange callbacks once they have more information. AIs don’t have this proactivity or short-term memory.”

Personalized customer experience

Human agents may also be better at providing customers with a personalized experience compared to AIs. Colleagues can better understand the types of interactions clients want based on their character profiles and conversation history.

“This is something our agents focus on from the start,” Go Answer says. “It’s why we offer a personalized service to the businesses we work with. Virtual assistants understand that they should apply a flexible and almost improvised approach to customer conversations depending on subtle cues and their requirements. Simply reading a script and saying the same thing every time isn’t an endearing approach and it doesn’t serve companies well.”

Better innovation

Entrepreneurs also still need people to spearhead innovations. While AI might be cutting-edge, it tends to regurgitate the best of human ideas instead of developing its own concepts and novel approaches (though this might be changing).

Innovation validation is a critical aspect of a high-functioning business that AI simply can’t replace. You can’t hand software a new prototype product and gather meaningful feedback, simply because AIs don’t derive any pleasure or services from using most consumer products designed for humans. Therefore, entrepreneurs will always need people in these circumstances. Having individuals with whom they can bounce off ideas is essential for iterating on products and making them better.

Making ethical decisions

Finally, entrepreneurs need people to make ethical decisions. Unfettered, machines will often make psychotic recommendations to achieve certain ends if they are the most rapid approach for doing so.

Making ethical decisions requires moral judgment and nuance. While some engineers believe AI might be conscious, most agree that it can’t and doesn’t view morality through the same lens as people. As such, it is up to humans to make the best choice for the direction in which to send a business.