Most HR professionals have at least toyed with ChatGPT. Are you among those who use it regularly or do you only occasionally call it up? Whether you are a regular or occasional user, what almost everyone has in common is that they are still in the early days of pondering the abilities and weaknesses of the tool.

The problem with that pondering is that the abilities and weaknesses of the tools are changing monthly. This blog provides a quick catch-up followed by some recommendations for HR leaders.

Here are the key lessons

• Many competitors. There are now many competitors to ChatGPT: Gemini, Claude, Lama, Grok, and more. For most purposes, the differences won’t be that apparent to the average HR user. It can be fun to experiment, but you are better off just sticking to one main generative AI tool rather than worrying if one is better than the other at the moment.
• Falling prices. There has been a lot of progress on smaller models that are cheaper to run. If you were worried about the cost of the first generation of tools, then you may find these smaller models eliminate that concern.
• Integration into existing tools. AI capabilities are being built into existing tools like Google Search, Microsoft Office, and Adobe products (e.g. anything labeled “Co-pilot” such as Microsoft Co-pilot). This means the tools are at our fingertips, but we need to encourage our teams to recognize that familiar tools have unfamiliar capabilities they need to learn.
• Amazing leaps in multi-modality. The latest AI tools can see, hear, and converse better than ever. GPT4o is almost as amazing a leap as ChatGPT was. You can show GPT4o things with your smartphone camera and talk about them. How will this affect HR is impossible to know.
• Tech is solving the weaknesses. There was a lot of angst about known problems of large language models, such as hallucinations and protecting the privacy of your data. Those problems are being solved.
• AI agents will create a whole new kind of automation. There is a big push for “Agents,” which help you automate tasks. This will open up many opportunities to increase efficiency right across HR—assuming your team teaches itself to use these agents effectively.

Some mistakes to avoid

Here are some mistakes to avoid:

• Falling behind. Don’t assume what you learned a few months ago is still valid. You should not express concerns about data privacy unless you are up to date on how those concerns have been addressed by new technology.
• Getting ahead of the market. When you see the possibilities of AI, it is tempting to invest in building some specialized capability. However, the large companies working in this area are moving so fast that your specialized capability may become a free built-in feature with the next release of one of the major models.
• Trying to learn AI in one go. Sure, it is good to make a push into learning AI, but it is more important to create a sustainable learning ecosystem that will keep up with the insane speed of advances.

What HR leaders should do

Take a deep breath. We are facing an amazing pace of change and it doesn’t look like it’s going to slow down anytime soon. All the while, there is this strange gap between the (accurate) feeling the world is changing fundamentally and the (also accurate) feeling that HR hasn’t changed much in the past year. It’s the old challenge of keeping two incompatible ideas in your head at the same time.

In this strange world, there are two things you should do. One is to build a sustainable learning ecosystem. That’s fancy HR talk that simply means you need continuous learning, and you need it at multiple levels of expertise. You will need access to some people with deep expertise in generative models, good relations with vendors, general awareness among your managers about what is possible, citizen developers who are not AI/IT specialists but are taking the lead on learning about and deploying tools, and finally basic capability for most employees (just like most employees need to know Excel).

The second thing is for HR to be part of the strategy team and thinking about how AI might affect the overall business. There will be lots of small wins with the new AI tools but there also needs to be concern about industry-shaking changes that could catch you unaware.

Feel free to connect if you are, like me, an AI-HR enthusiast.

David Creelman is CEO of Creelman Research. He has a particular interest in helping HR develop the capability to handle AI issues. If that’s a concern, then get in touch. You can connect to Mr. Creelman on LinkedIn or email him at dcreelman@creelmanresearch.com