SHRM All Things Work
Avi Gesser on How ChatGPT Can Best Serve Workers
Fri May 12 2023
Generative AI and chat GPT in the workplace
00:00 - 07:07
- Generative AI and chat GPT are being used in the workplace to boost employee efficiency and productivity.
- However, detractors argue that chat GPT can be incorrect, biased, racist, proprietary or copyrighted.
- Chat GPT is a tool that is only as good as the data it's been given.
- Employers should view chat GPT usage in the workplace as a tool with low risk, high value use cases.
- Chatbots are great for customer service and collections.
- Chat GPT is excellent at translating, summarizing and cleaning up grammar.
- It's also good for tech support, brainstorming and generating first drafts of speeches, reference letters and job postings.
- To get better results from chat GPT tools, you have to anticipate all the follow-up questions you'd normally get from a human and put them into the initial query.
Legal risks and policies related to generative AI
06:41 - 13:58
- Generative AI laws do not eliminate legal risks associated with their use.
- Existing rules apply to the use of generative AI tools in certain contexts.
- Companies should be honest with customers about what these tools can do and avoid discrimination against protected classes.
- Some companies allow employees to experiment with chat GPT and generative AI, while others ban open versions due to confidentiality and privacy risks.
- Licensed enterprise-wide versions provide more control over data input and confidentiality, but policies vary depending on company culture, size, and risk tolerance.
- Allowing all employees to use any generative AI tool for any work-related task carries significant risks such as confidentiality, privacy, contractual, cyber, intellectual property, transparency, and quality control risks.
- Three steps for creating an AI policy for employees include outlining prohibited uses that violate company policy or are illegal or disreputable; defining what is not prohibited by providing examples of allowed uses or an exhaustive list; and evaluating undefined use cases through a council or group that balances short-term benefits against downside risks.
AI decision-making and regulation
13:36 - 20:47
- Picking the right person or group for controversial decisions is important.
- HR typically owns employee communications and policies, but there's no one right way to handle AI decision-making.
- Committees with multiple voices can lead to better results and credibility.
- Regulation of AI in hiring, facial recognition, lending, and insurance is coming in a patchwork of state and municipal laws.
- Different countries will have different priorities when it comes to regulating AI tools.
- AI can be leveraged by HR for basic tasks like writing job descriptions and setting up chatbots, but quality control and loss of training opportunities are potential downsides.
- Businesses must balance speed, cost, efficiency, and quality when deciding whether to use AI tools for drafting documents.
HR, payroll, and workforce management solutions from UKG
20:29 - 20:47
- HR, payroll, and workforce management solutions from UKG can help make your dream workplace a reality.
- UKG provides the tools you need to support and celebrate all of your people.