Dr. Brittany Castonguay 3 November 2024 #technology
Artificial Intelligence (AI) is a revolutionary tool that will continue to be a dynamic playmaker in business operations for an unforeseeable future. As consultants, SHRI recommends embracing AI's technological advantages for your organization. You can use AI to streamline tasks, reduce redundancies, improve customer relations, gather feedback, etc. However, you cannot use AI to replace the human touch.
A strong AI program requires human balance. AI can still be flawed programs with blindspots that technology cannot circumvent. Technology is designed to follow a set of codes and respond based on the logic prescribed. AI does not have the capability to be abstract in its reasoning. Consider Colorado law, which pertains to AI and adverse employment actions. If an agency uses AI to consider hiring actions, promotions, raises, or evaluations, it may be subject to an appeals process and held liable under Colorado law if AI makes a discriminatory decision. The program follows its logic, which a human codified. Looking over its limitations, AI requires oversight.
As employers begin incorporating AI into their organizations, they should build a robust checks and balances system to protect themselves from potentially damaging actions. Colorado is one example of a state where laws either go up for a vote or have already been passed. Regardless of your state's current practices, the government will memorialize AI practices to protect individual employees. As you would when introducing any new technology, organizations should develop a policy to regulate its uses. Policies do more than protect the organization; they can be used to train employees and act as a safety net.
The first step is determining how you will integrate AI into your workforce. What problem do you want AI to solve? Clarifying the problem will help lead to a solution and alignment of goals. Starbucks uses AI in its marketing strategy to anticipate consumer interest in new products by using algorithms to calculate customer behavior on sites, ads, and apps. AI can see how long you hover over an image, click on a description, or go back to view a product multiple times over the week.
Consider how you will collect and interpret the data used by your AI system. Think back to the Colorado example. If you have AI looking over applicants, like the popular Applicant Tracking System (ATS), what keywords and parameters are you using to screen for applicants? If the data AI is tracking is not reliable or flawed, the data analysis will be flawed.
Remember, AI cannot think or invent. Your AI programs should consist of teams of innovators that will help your AI evolve into better products by rapidly testing ideas and scaling output. To do this, your teams should be agile and flexible to quickly pivot in response to new industry trends. There is an inherent risk the organization must be willing to accept, yet the dividends of scoring a winner and getting ahead of the competition will offset the potential losses. AI and agile teams that support AI reduce loss because they can quickly test new ideas and analyze the data without spending months on R&D.
As you consider your integration needs, remember to consider the back-end of implementation: program maintenance. Due to several recurring events, every AI program must be an effective and compliant part of the organization. The following are best practices to consider when integrating AI into your organization:
Best practices:
Regular Audits
Vendor Compliance.
Acknowledgment of AI use by Employees and Customers
Alternatives to AI
Regulatory Policies
Training
Comentarios