To successfully hire and retain AI talent, employers must start by taking a step back. What do you want to do? Do you want to retain talent because AI is the next big thing and everybody’s telling you that you need to have AI in your product? That may be true. From a sales perspective, telling prospective customers your organization doesn’t have AI built into its product is a good way to end the conversation. However, if your starting point when talking to talent is “Yes, we’re going to have AI. It’s going to be big. It’s going to be huge. Everybody says so,” that’s another good way to end the conversation. 

This is because people with AI skills aren’t attracted to companies that have no idea what they’re doing. There must be a direction; that direction must be clear; it must make sense. Therefore, the first step in attracting and retaining AI talent is knowing what you want to do with AI. You need a vision that’s compelling; it’s surprising how few companies have that. 

Beyond the Vision: Prioritizing Essential Skills and Internal Talent

Once you establish a vision, you can start to think about what kinds of talent you need. Very few companies need the sort of AI expert who can build and train models. They’re few and far between – they work mostly for Google, OpenAI, and a few other companies. Your AI will either use a service like OpenAI or be built on a smaller open model like Llama. So don’t start thinking that you need AI Ph.D.s. Likely, your needs are more modest. You’ll want people with a solid knowledge of statistics and who have experience collecting and working with data. We’re seeing increased interest in topics like data and AI governance, particularly as regulation starts to shape the industry. Whether they call themselves statisticians, data scientists, or something else, they will be easy to find, and most of them are already thinking about AI.

You will also need people who can fine-tune models, implement RAG pipelines, and evaluate whether the model is behaving the way you expect. You will need people who are looking into AI security. AI operations – deploying, monitoring, and managing AI systems in production – will be a significantly different specialty within AI operations. It’s difficult to hire precisely for these areas of expertise. You’re best off looking for candidates who work in specialties like data science or IT operations and probing them to see how they’re thinking about AI. Don’t try to hire someone with experience “building RAG applications.” Look for data engineers who build and manage data pipelines and ask them how AI fits into their future. 

You’ll also need to look at your own staff. Who is interested in AI? Who’s looking for a new role? Some of your top people are no doubt thinking about AI already. Some of them may even be looking for jobs elsewhere. Give them the opportunity to be loyal to you, and you’ll be more likely to retain them.

Standing Out in the AI Talent Pool 

Beyond possessing this technical expertise, prospective employees must realize that AI is a new area that’s changing very quickly. The best thing you can do is develop a portfolio: Some projects that use AI to do something different or interesting. They can be personal projects. They don’t have to align with the interviewers’ needs–the interviewer may not know what they need. 

But more important than that, if you’re on the job market, make it clear that you can work alongside AI, and that you understand collaboration between humans and AI. That’s how the workplace is evolving; make it clear that you are part of that future. Humans working with AI will be more productive than either humans or AI working by themselves. And humans who understand that kind of collaboration will be especially valuable in the talent pool. 

Whatever a company wants to do with AI, thinking of AI as an oracle that gives humans answers is a dead end. If you understand that, and if you have insight into how humans and AI can work together, you will be well positioned to help a new employer build the AI-enhanced products they need.