How to Use AI to Break Down Organizational Silos

GenAI allows everyone to do a little bit of everything.

That fact can be especially valuable for startups, where employees typically wear many different hats. Several small business owners recently told The New York Times that genAI is helping them operate by filling in their skill and knowledge gaps. But it can also help bigger companies, where roles and functions are typically more defined, by eliminating bottlenecks and breaking down silos.

Here’s how:

1. Workers can create things they never would have otherwise.

Before genAI, people who didn’t know how to code were limited in their ability to create bespoke tools to assist them with their work. “The sign of an engineer is, if I have to do something three times, I'll get annoyed and I'll write a script to do it,” says Ely Greenfield, chief technology officer of Adobe’s digital media business. He argues that LLMs put that ability into more people’s hands. Now, if an employee or team has a pain point, they may be able to come up with a solution on their own, as long as it meets the company's security, quality, and ethical standards.

A good example of this comes from Amazon Web Services, where a team in one of their data centers used Amazon Q Apps to build a simple app that pulls information from documents to help troubleshoot HVAC error codes. “No software team is going to spend time building that app,” Deepak Singh, vice president of next-gen developer experience at AWS, told me. “They've got other things to do. So it never gets done.” 

The ability to create assistants for specific yet repetitive tasks is why some companies have leaned into GPTs, customized versions of ChatGPT. Moderna employees, for example, had created over 750 GPTs by April of this year, which perform tasks like recommending doses for clinical trials and writing responses to regulators’ questions, according to The Wall Street Journal. 

2. AI can unblock parts of employees’ workflows.

Rebecca Hinds, head of Asana's internal think tank, The Work Innovation Lab, told us last year that members of her team use ChatGPT’s data analysis feature to analyze large surveys, a use that was more helpful for the team’s less technical employees. “All of a sudden those employees don't need to go to a data analyst and say, ‘Do we see this in the data?’” Hinds explains. “They can start to take on some of that role themselves and be a data analyst.”

Adobe’s Greenfield says he’s looking into use cases that help employees gain insights faster without needing specialized technical skills. He gives the example of interpreting clickstream data on Adobe’s products to understand where they’re helping versus failing customers and why. “Today, I may have to go to an overloaded analyst and hope that they can get around to writing the query and pulling out the insights that I'm looking for,” explains Greenfield. Instead, an employee in product marketing could “work with an LLM to say, ‘Okay, you're connected to my data source. You're seeing how customers are using our products and where they're succeeding or where they're bouncing. They weren't able to make the flyer or add text…Why? What's different about these customers?’”

3. Workers can develop and implement ideas faster.

Sal Khan, founder and CEO of Khan Academy, recently told Charter that while it’s important to have deep expertise in your domain, it's increasingly the case that “you can’t just stay in your lane.” The people who will succeed are those who are entrepreneurial and use the tools available to see their ideas through. “Right now…in big organizations, you have a marketer who has an idea but really can't do anything until they get buy-in from the designer and the engineer, and then it's going to take four weeks,” he says. Entrepreneurial employees can now try to move forward on their own using genAI. “‘Okay, I have a marketing idea. Let me create some mocks really fast.’ Maybe eventually you'll hand it over to design and they can refine it, but let me just do it.”
 

Written by Jacob Clemente.

© 2024. This article is reprinted with permission by Charter Works Inc. All rights reserved.