Happy Wednesday!
This week I’m thinking about AI enablement. Unlike previous software enablement cycles AI enablement has been coming in waves as each wave of AI has new and different capabilities. I listened to this episode twice. First with an eye on what the endgame is for AI enablement. I actually disagree with the consensus with the guests, I don’t think the outcome of AI enablement is boring. Mostly because of the successive waves of AI enablement as well as the philosophical navel gazing required to fully adopt AI. What does it mean to work? So many people have what they do wrapped up in their identity that even after they adopt AI, the questions and insecurity still persist. Second with an eye for individual tactics and this podcast does a great job of peppering you with tips. Such as do you frame AI as a coworker or an employee. I agree with them, framing AI as a coworker is a bad framing both from an ownership but also an adoption perspective. New coworkers can be aids but also rivals. Who is this clanker taking my job? Versus, I can now spend more time thinking about what we need to do rather than filling out spreadsheets and copy pasting between systems.
Getting past the pilot: Why so many AI test projects have trouble scaling
Fortune bolsters this point. This article suggests you advocate for lots of AI experimentation but with very tight guardrails around governance. You can see that in agentic roll outs at companies recently. Many roll out agentic AI with the general idea that people can use it to replace headcount but that shows a fundamental misunderstanding about what AI is, what automation is and what enablement is. You need to fully understand a workflow before you can automate it and you need to understand that automation before you can improve it, deprecate it or make it work well with other automations. Many business discovered the hard way that automating the mundane part of someone’s job does not obviate the need for their headcount. Getting rid of that headcount prematurely might mean having to take the expensive step of rehiring someone to help understand what the last employee already knew.
This survey shows that the growing concern over “AI slop” and its impact on employee experience is something you should pay attention to. When AI adoption goes up people end up spending more of their time reviewing output and less time doing. Great, everyone is like a manager now. It’s exhausting in a measurable way from a cognitive standpoint. This is leading to Gen Z in particular to be nostalgic about the world before AI. But why? They’re tired, they don’t know why we’re asking them to use AI and it’s leading them to use AI tools less. Providing the why consistently and proactively is the keystone in any change management program and it’s just not being done at a lot of companies. Many people see the benefits as obvious. But are the benefits obvious? Clearly not to everyone. The “verification tax” paid by employees verifying AI work is exhausting. They don’t feel like they have the tools or working style to support this new style of working and I don’t blame them for feeling like that. It’s a real concern and it’s L&D and enablement’s job to help them realize the upside of AI while providing them workflows that give them time to reflect, decide and improve their judgement in a way that the workforce didn’t need before.
This is Stanford’s AI economic indicators report. This is important for people working on AI enablement or in workflow development. Knowing the “canary” jobs and their changes helps you understand how other jobs might evolve in the age of AI adoption. I’ll save you the click, AI adoption is going up and to the right and the US is leading in the speed of adoption. On Canary jobs the headlines have focused on entry level software engineer jobs falling. Software engineering is heading towards the “diamond shaped” workforce while health care is having a broader base. The whole story is more complex though because different tasks are responding to AI in different ways. They’re also responding depending on the kind of job and whether AI is used for augmentation or automation.
This foundational book explores System 1 (fast, automatic heuristics) vs. System 2 (slow, deliberate, logical effort). Cognitive debt essentially occurs when a system or individual over-indexes on System 1 shortcuts or delegates System 2 processes entirely to external tools, causing long-term skill degradation. I think every L&D professional nay, every business professional should read this. Although I will acknowledge this is quite a chunky read.
Nerd Out Gear⚡| Share Your Thoughts 💡 | Connect on LinkedIn 🌐
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.