What will the best artists, engineers, and narrative designers of tomorrow have in common? They’ll make AI tools and workflows an integral part of their output. A recent survey of 1,000 AI users found that 50% get little or no formal training from their employers on generative AI. This is a huge miss for the games industry, as formal training programs lead not only to more effective use of the things most employees are already using, but it also provides a much needed avenue to train on ethical use of generative AI.
What Effective L&D Programs Actually Look Like
Structured learning and development programs for generative AI don’t mean rigid control. They mean channeling inevitable adoption into lasting capability. The most successful initiatives share several characteristics:
- They tier fluency development
- They match learning infrastructure to reality
- They capture and share gained knowledge efficiently
- They address genAI resistance
- The make L&D a core part of work
Tiered fluency development: Everyone will need baseline AI literacy. One might argue that such literacy is needed today, but few compelling arguments can be made that baseline AI literacy will not be necessary in the very near future. I’m not talking about being fully steeped in technical architectures or math that has no numbers. I’m talking about a practical understanding of capabilities and limitations. Any successful L&D program should start here, then branch into discipline-specific applications.
Learning infrastructure that matches reality: Gone are the days of classroom-style instruction for workplace L&D. No professional wants to be regularly lectured at, and so great L&D programs find ways to couple learning with new experiences. At Series, we’ve done limited time game jams, an AI fair, short form (as in, no more than 10 minutes in length) teach-ins, and more. We’ve also given our team members the tools to become instructors on their own. Show-and-tells and company-wide updates during All Hands meetings utilize time employees are spending together in better ways than just reciting what every team is working on.
Knowledge capture and sharing: When someone discovers a better way to “make a wheel” – a better prompt, a better workflow, etc. – that knowledge needs to be captured, documented and disseminated. Some studios maintain internal wikis of “AI recipes,” proven prompts, workflows, and use cases specific to their projects. Others use Slack channels with good threading discipline. The mechanism matters less than the commitment.
Addressing the Resistance: Not everyone will embrace AI tools enthusiastically, and that’s worth addressing directly. Some resistance is philosophical; some resistance is practical. Don’t shy away from resistance or criticism. Steer into these headwinds and address them directly. Better yet, have the participants in the instruction address the resistance themselves. We like to use the “how might we” framework here – “how might we accomplish X,” or “how might genAI solve this problem.” This gets people out of the negative “it’s impossible” or “it shouldn’t be done” mindset and gets them thinking along positive lines.
Recurrence is Key: Treating AI L&D as one-time training rather than ongoing development might be the worst mistake a company can make when it comes to improving AI use and fluency. The frequency of L&D engagement at Series is as important as user growth, daily time spent, and other “traditional” KPIs. Why? Because things move fast in this brave new world. What was true of the state-of-the-art six months ago is archaic today. AI is improving at an exponential rate, and humans are generally bad at recognizing exponential growth. We know we need to stay on top of new developments, and so we make frequency of L&D engagements a core part of our company goals.
Why This Matters
Twelve months from now, basic AI fluency will likely be an expected baseline for creative and technical roles in gaming. They won’t be differentiators; they’ll be requirements. We’ve already seen an uptick in younger prospective employees wanting to evaluate us on our approach to AI use, adoption and ethics (which is a topic for another post). From our POV, a game company that trains teams to use AI tools effectively and thoughtfully signals cultural maturity that generic “AI-first” positioning doesn’t. And right now, much of the “AI-first” rhetoric is just that – generic. For us, having a robust L&D program is one way in which we make Series the opposite of generic.