Is AI Helping Or Hindering Productivity?
Advancements in artificial intelligence (AI) technology and the advent of generative AI models like ChatGPT have brought a lot of excitement as well as nervous energy to the business world. AI is the dominant technology trend right now, and organizations of all sizes and industries are racing to integrate it into their products and processes.
AI is indeed a powerful tool, and it is a tool that needs to be used properly. Norms for its adoption and usage have yet to be established, though. When used responsibly, AI can supercharge productivity, employee retention and customer satisfaction. When used poorly, it risks being more of an impediment than an aid.
Below are my thoughts on how to prepare best for the effective deployment of AI-based solutions.
How AI Helps
Organizations already are finding a variety of uses for AI across all departments, with new capabilities and use cases being added constantly. For example, contact centers and customer support teams are using AI-powered solutions to answer calls and provide information, advice and sentiment analysis (often in real time), empowering agents to provide faster and more efficient service. AI solutions also already take notes and log calls, automating time-consuming tasks and allowing agents to focus their energy on engaging directly with customers.
In follow-up to these customer calls, AI-powered solutions provide quality scores and call insights, helping supervisors provide support and training for their agents. These tools also can sort through conversation data to glean new, valuable customer insights, helping businesses to better understand their customers and provide them with higher levels of service.
Automation is a fundamental driver for AI adoption, and in the contact center, a leading use case involves chatbots helping with self-service. First-generation chatbots were not AI-driven, but today’s chatbots are—and they are being used to create smarter customer interactions that can understand natural language inquiries and provide accurate feedback without relying on pre-programmed menu options. Not only does this provide better options for customers who prefer self-service, but it also helps keep agents free to attend to more complex matters.
AI also completes essential tasks outside of the contact center. For instance, businesses that add AI-powered features to their communications platforms benefit from automatic conference notes and summaries, making their daily meetings more efficient. At the same time, AI tools help with tasks by generating presentations, editing emails and reports, and pulling up information when requested.
These are just a few examples of how companies can use AI to boost employee efficiency, enhance business processes and improve both the employee and the customer experience.
How AI Hinders
For all of the good that AI does, there are also potential risks. For example, since AI relies on training data and human-based programming, the saying “garbage in, garbage out” is especially true. AI solutions need to be trained on vast amounts of accurate data and then repeatedly tested and refined to ensure information accuracy that is also free from bias.
Additionally, companies need to ensure that AI solutions are trained on their data and not more generic forms of data scraped from the public internet. Specificity matters, or else companies run the risk of using and providing false, unreliable information.
For example, as it stands today, generative AI risks “hallucinations,” which is the production of outputs that are incorrect. These tools have a long way to go to develop reasoning and deductive skills that can detect bad outputs right away, as humans can. Many found it alarming when Microsoft’s AI-powered search tool, Bing, threatened users harm in early 2023 and provided inaccurate information.
Additionally, human bias can make its way into training data. We’ve seen this in multiple instances, such as in Amazon’s use of an AI tool designed to help with recruiting that ultimately was unhelpful due to its discrimination against women. More recently, researchers have found bias in medical AI that can lead to inaccurate diagnoses depending on the patient’s race. These kinds of errors stand to have serious consequences and must be caught and corrected as quickly as possible.
Another risk factor for organizations when rolling out AI technology is over-reliance on it without adequate human oversight. A “human in the loop” is necessary to make sure that the AI solution is providing accurate information, and oftentimes a human touch is still needed for things like assisting customers with complex or sensitive issues in empathic ways that AI cannot replicate (yet).
Over-reliance on AI-powered technology that minimizes or removes the human element will leave a company in a risky position in which a single error can have disastrous consequences. The U.S. judicial system felt this when a lawyer used ChatGPT to prepare a filing only to find that AI had invented fake case law. Proper fact-checking is mission-critical throughout any organization.
Using AI Technology Responsibly
AI offers powerful solutions for businesses and employees. For now, its best use continues to involve rigorous human oversight as AI learns and grows. With the right tools, technology and training, businesses using responsibly-launched and monitored AI-powered solutions will reap its many benefits, gain new customer insights and empower their workforce like never before.
Originally published on Forbes, April 2, 2024
