Technical Trainers: Don't prioritize AI

Okay, I don't really mean that. AI can be likened to past IT innovations, but treating it like smartphones, cloud computing, or blockchain might be a bit na?ve. There is a reason OpenAI CEO Sam Altman said, “What I lose the most sleep over is the hypothetical idea that we already have done something really bad by launching ChatGPT.” Make no mistake, however, that AI shares a common history with previous groundbreaking technologies with good and evil consequences.

Although AI has been discussed since Alan Turing published COMPUTING MACHINERY AND INTELLIGENCE in 1950, it flourished from 1957 to 1974. Reaching its current status and beyond required only the advancement of computer memory capacity and processing power, among other hardware and software advancements in machine learning. The proof of Moore’s Law, the development of “expert systems,” and the IBM-built supercomputer Deep Blue’s defeat of Gary Kasparov in 1997 made clear that AI was here and continued to prove its existence throughout the 2000s.

Among the similarities to past innovations is the process we find ourselves in today: the rapid response in technical training and educational opportunities around AI and ChatGPT in particular. Each day we see more jobs with the titles: Prompt Engineer, AI trainer, AI Auditor, and Machine Manager. Training companies are seeing a surge in training for these positions as well as for executives who need to understand the fundamentals of AI. Tutorials on the Internet abound; everyone seems to be busy learning AI, which is no different than the learning in the early 1990s about the Internet, the world wide web, and how to write web pages. And just as we moved toward Rich Internet Applications and the change from proprietary technologies like Flash to the overabundance of JavaScript frameworks, we learned how to build large enterprise applications with new server-side frameworks, build processes, cloud computing services, and more. So relax, technical trainers; we will all get there just as we have in the past.

However, the topic of this post is, as technical trainers, what makes this technology different, and what might have prompted Sam Altman’s foreboding quote? Skillsets were required to participate in previous technology shifts like the creation of HTML and the World Wide Web. Someone had to teach the fundamentals of the Internet, how to create web pages, and so on. Further advances in programming languages, the evolution of the web browser, cloud computing, and processes such as Scrum/Agile all seemed to necessitate instructor-led training. To be clear, training itself has undergone many changes, from old-school classic sage-on-the-stage lectures to today’s modern techniques grounded in the neuroscience of learning. Additionally, we have seen training on DVD, subscription-based video training, and virtual-reality-based training, not to mention gamification and more.

What makes AI different is the notion that we can use tools like ChatGPT as educational devices. Whether they enhance the current live, trainer-led sessions or replace them remains to be seen, and in my opinion, it is far too early to implement as such. In past technology transformations, workers needed instruction that originated mainly in the form of a human being transferring knowledge; we now have technology that may not require a human being transferring knowledge. Indeed, the technology itself may provide the instruction.

Currently, most technical trainers limit AI’s use to a combination of grunt work and assistance in the more creative aspects of training. Examples include using AI to generate code, courseware, quizzes, and summaries and provide ideas for course outlines and learning objectives. The main caveat when doing this is to remember that you know more than ChatGPT; more about your client, your learning outcomes, and your students. Thus, the results you get as output will be exponentially more helpful if you first “train” the AI tools and provide them with the knowledge you have as input. This fact may seem obvious, but many of us fail to consider the vast amount of knowledge we acquire about our learning environment before the training session. An hours-long phone call with attendees puts so much data in our brains that we take it for granted. The hesitant tone of an insecure learner tells us volumes. The pain points that management informs us of have significant implications for our course objectives, outlines, etc.

So, knowing that as technical trainers, our knowledge of AI will increase, we will do what we always do when faced with new technologies – we will learn what we need to learn with the benefit of our prior knowledge domains. We will help our clients grow in productivity and profitability. Bear in mind, however, as other technologies have changed the employment landscape, so will AI. In a process called “creative destruction,” where old technology is replaced by new technologies,?the prepared and flexible worker plans accordingly. Just like typesetters, video store workers, and switchboard and punch card operators are jobs that all disappeared, they were ultimately replaced by higher-paying and more knowledge-intensive careers. So too, will the technical trainer’s job change or be eliminated?

While that remains to be seen, we’d be well-advised to remember and prioritize our skills outside of AI. For starters, don’t use AI to do things you already do well. Look for the pain points in your training and use AI to improve the learning experience. If AI will enhance our knowledge and problem-solving skills, first consider what we bring to the table that AI lacks. In the words of Po-Shen Loh, a professor at Carnegie Mellon University, “Think about what makes humans human, and lean into that as hard as possible.” @poshenlog Loh’s advice to focus on creativity, emotion, and the stuff that makes man different from a machine should remain at the forefront of the technical trainer’s mind because it’s these qualities that ChatGPT cannot duplicate. Like Loh, @RayJimenez tells us, trainers, to “look for pain points” and “deepen coaching with the aid of AI.”

Find the tedious aspects of training preparation, courseware development, and so forth, and yes, you can reduce the time they take by order of magnitude, but don’t leave to AI those challenges that have already been solved. There are many challenges that AI cannot see or solve. Back to Po-Shen Loh as an example, he improved his public speaking skills by taking a year-long improv-comedy class. Another human can best solve the human pain points we encounter during training — with the assistance of AI.