Content

Hiring Artificial Intelligence Talent: Beware Of Unicorns

At the dawn of big data, data scientists were the unicorn of the hour. Talk to any one of these rarities, and they would say there were only about 100 of them in the world.

When executives woke up to the potential of big data, it was also at the same time we were dealing with financial collapse. Balancing between the data economy and keeping regulators at bay created a new unicorn, the chief data officer.

Digital has had its own unicorn story. Adopting technology to automate, augment, and scale businesses is hard. Usher in the chief digital officer.

Today, AI talent challenges inside enterprises are causing firms to assess their approach to data science while also recognizing deep data deficiencies and the impact on DevOps. The new unicorn? The AI engineer.

Why? What made these roles and individuals unicorns? There is a recognition at each technology leap that enterprises need a sage to plug a number of gaps between digital experience, analytics, and data.

  • Chief data scientists were expected to have highly advanced statistical skill, computer science skill, and some business expertise in the area they would work.
  • Chief data officers were expected to have deep information science, data architecture, data governance, digital solutions, and business expertise.
  • Chief digital officers were expected to lead business transformation, solution architecture, and have strong analytics and data experience.
  • AI engineers are expected to be able to develop expertly within information science, data management, integration, application development, and have knowledge of the use and application of machine learning in solutions.

Do you see the problem? Too many skills and experiences are expected from a single person. In an emerging market, to expect savants are available or findable is asking a lot. Even the digital disruptors — Amazon, Facebook, Google, or Tesla — know these roles are mythical. An early MVP for each wasn’t a market viable product; it was a proof of concept (POC). There was enough there to tell a story of what could come and what could be achieved. Each took the next step of breaking down the talents and experience necessary to scale and create the MVP that potential customers might actually use and pay for.

Enterprises need this same strategy for two reasons. First, you may get a unicorn after significant searching and coercing of these rare creatures. But the trick will be how you can keep them. Money only gets you so far if unicorns are bored, unchallenged, disconnected, or frustrated. Someone always has more money. Second, if you can get unicorns to stay, you can’t scale beyond your own pilot and POC. You live in the experiment and not the desired outcome. You plugged the initial talent gap to get started, but that only delayed the real challenge of broader talent and resource transformation to run an intelligent cognitive business.

Learn from those digital disruptors:

  • Break down your talent needs irrespective of a role to understand what you really need.
  • Organize these talents into logical centers of gravity for a role and career path.
  • Design the operating model around a natural collaboration effort where these roles enhance each other rather than compete.
  • Suppress unicorn mentality to avoid superhero hires that mask the underlying issue of poor business practices.
  • If a unicorn is needed, strategically source and place for short-term transitions and ensure unicorns translate skills, processes, and operational best practices for long-term scaled business gains.

The role names may still persist, but the expectation of the individual in the role is rightsized rather than a kitchen-sink job description.