The use of artificial intelligence technology in the workplace are making employees both nervous and excited.

The powers (and limitations) of the headline-grabbing ChatGPT platform from OpenAI are raising questions about authenticity and creative autonomy, while Microsoft’s GitHub Copilot promises to help programmers write and fix computer code.

“Technology can be positively disruptive to the current workforce, and AI technology has the potential to help reverse the overall downward trend of US labor productivity that we’ve experienced for decades,” says Atif Zaim, national managing partner, advisory, for KPMG.

He says while fear is only human nature, now is the time for leaders to think about how it can positively impact their workforce -- and the skills that will be needed to capture the potential productivity gains.

“Leaders can help alleviate employee fears by clearly outlining their vision for the future of their workforce and how they can harness AI to help in their daily jobs and deliver better outcomes for employees and customers,” he says.

Forrester vice president and analyst Diego Lo Giudice calls it the “early days” of AI applications in the workplace, noting it is not yet a super mature space for large-scale leverage of AI on the IT side.

“For software development, for example, the fundamental premise is that AI can augment the efforts of different professionals in different stages of the of the development lifecycle,” he says. “I don't think IT organizations are going to run ChatGPT to write their code right now because there are some risks and challenges.”

If used in the right way, however, Lo Giudice say it can help skilled developers speed up their work, using platforms like ChatGPT to help with knowledge management activates that take up precious time.

“ChatGPT translates code from one language to another, you can give it a piece of code and ask it to explain what the code does, or ask it to write documentation for code,” he explains. “You can give it a piece of code written in Python and ask it to provide an alternative version or a better optimized algorithm -- and it will do so.”

AI Helps with Offense or Defense

Mika Aalto, co-founder and CEO at Hoxhunt, explains secure coding behaviors have been a challenge for many software engineers, especially in very large code bases.

From his perspective, AI-based tools like Copilot can help “auto-complete” programmers work and ease the burden on data scientists to accomplish more impactful goals.

“There are different approaches to adopting AI depending on whether you’re looking to use it for offense or defense, meaning innovation and growth or for security purposes,” he says.

If an organization is looking to create disruptive innovation for competitive advantage, the CIO or CTO should first do a landscape review that examines existing challenges and how AI can be leveraged to create a new solution or to do things better, faster, cheaper.

“If your company stands to be disrupted by AI, the CEO and Board should be involved from the outset as key stakeholders to defend against or to design disruption,” he says. “Look at the investment Microsoft recently made into AI. They learned their lessons about late adoption the hard way when they came late to the party of the mobile revolution.”

AI Lets Developers Spend More Time Developing

Muddu Sudhakar, CEO and cofounder of Aisera, explains AI can be used for developers in Github or GitLab for account creation, code check in issues, code merge, debugging, configuration issues and what-if analysis.

“Developers spend a lot of their time on configuration, debugging, and maintenance,” he says. “These are not particularly interesting or a good use of a developer’s valuable time.

Instead, these functions should be automated by systems like Copilot, allowing developers to spend their time on what they like to do -- creating great apps.

“This is why Copilot is considered to be about pair programming,” Sudhakar says. “Not a replacement for a developer.”

Angel Borroy, developer evangelist at Hyland, agrees automated workflows and AI technologies can streamline processes, boosting employee productivity and happiness in the long term.

“Copilot for example, may help junior developers to produce code faster, since it can complete code with the right syntax and it’s able to suggest popular algorithm implementations,” he says.

However, it’s important to note that when dealing with complex developments (including multiple files instead of a single codebase) and custom logic implementation, Copilot may have some flaws, proving that AI and automated technologies perform best when paired with employees.

“We’ve found when it comes to administrative tasks, data storage, and developer

tooling, AI can automate data quality checks, make recommendations for improving data

integrity and uncover hidden trends and surface insights that enable workers to be more

productive,” he adds.

Please Use AI Responsibly

Sreekar Krishna, KPMG's US national leader for artificial intelligence, explains generative AI can produce synthetic test cases for developers and QA activities or translate code written in one computer language to another.

“The technology can also automatically check the quality and interpret data where metadata is not available, interpret tabular data and summarize them with natural text and jointly interpret image, text, and tabular data,” he says.

Krishna cautions that while generative AI has exciting potential, the recent focus on the technology has also reinforced the importance of responsible AI.

“Going forward, organizations will be using AI methodologies to make decisions for their customers, employees, vendors and everyone associated with them,” he says. “A responsibility charter needs to be sponsored by C-suite leaders and developed through dynamic and consistent discussions led by the leaders in compliance, risk and data analytics.”

Lo Giudice adds it is important for organizational leaders and IT workers, for example software developers, to come together and decide which AI-based tools could be deployed and the strategy behind that deployment.

“Developers are influencers of this, because if they get excited about it, it will win,” he says. “The senior developers I've spoken with are very excited about this because at a minimum, it takes away the repetitive tasks.”

What to Read Next:

IBM’s Krishnan Talks Finding the Right Balance for AI Governance

Why the US Risks Falling Behind in AI Leadership

5 Ways to Embrace Next-Generation AI