There are two formative experiences I learned from the early Lynda.com instructors that could never be replaced by an AI-first instruction model. The first was from photographer Chris Orwig. In his early Adobe Lightroom courses, he would sing shortcut key combinations, and it was fantastic. He made up a little rhyme that tied the keystrokes to the action, belted them out in a lyrical voice, and instantly created a little ear-worm that would play back in my head when I was working later. Was it goofy? Yeah, kinda. But was it effective? 100%. The second was in a Photoshop course from Taz Tally, probably from the early 2010s. I distinctly remember one video where he just said, "Sit back and watch this." Taz then proceeded to work through an entire editing session at his normal, professional working pace, entirely on the keyboard. Fingers flying over the shortcuts, cropping, zooming, setting color corrections. All blazing fast, and all very impressive. The lesson wasn't to immediately impart that skill on the learner, but it was to show what was possible if one took the time to build up a muscle memory. Again, super effective, and a learning memory that doesn’t come from prompted responses, but from watching a human at the top of their game. Tom Green just posted an article that outlines his thoughts on the future of education platforms in the face of AI. The loss of lived experience and human connection is going to have a profound impact on the way we learn foundational career skills in the future. You can read it here, it's worth your time if you're in this space: https://lnkd.in/gq9QV5GU My main takeaway: AI is great for meeting people in the moment, to solve an immediate need. It's perfect for giving away fish. But for building a foundation of knowledge in support of a lifetime of learning, people are still going to need the human connection that will impart knowledge in creative, memorable ways. Instructors that connect stories and threads. That are able to share not only the best practice when everything is working, but the pitfalls when things get messy. Instructors that are able to be honest with learners when a flashy new feature is only half-baked, and to provide workarounds or cautionary advice that goes against the documentation's rosy depictions. AI can't teach you how to fish, not really. It won't ever be able to share the knowledge needed to read a river, to watch the water's flow and whorls. To understand the complex vegetation and weather patterns that direct a seasoned angler to the best spots to cast a line. And this is exactly the type of instruction that we're tossing aside in the rush towards instant answers and 30 second tips from a machine that, at best, can only give an average solution to a known problem found in its training data.
