I recall starting my first job out of college. Initially, I spent six months in a trial by fire. It wasn’t a pleasant experience. Shortly thereafter I landed a job at AT&T Bell Labs in Naperville, IL. It was 1984. Investment in telecommunications was booming. It was a huge project, employing at least 3,000 software developers working on 3 releases with six-month release cycles. The hardware, software and software management was highly complex. Onboarding to this environment required mentorship.
I needed to learn how the software worked on a system could provide telephone service for a small city. The system was incredibly complicated and expensive. Learning how it worked and contributing to the effort wasn’t easy. To improve onboarding and career development, AT&T Bell Labs had developed a mentorship program. I’m not sure how long this practice was in place, but it was incredibly effective. When I started working on the project I was assigned a mentor who was chartered to be my “go to” resource on the project and for the rest of my career at the company.
At AT&T mentoring was a long-term relationship codified into the DNA of Bell Labs.
This system of mentorship works. I used it at another telecom company (Teradyne) and the people who benefited from it had stellar careers. It was the same company where the where Uncle Bob was an engineering manager. I learned a lot from Uncle Bob while he was developing his consulting business. I consider him a mentor as well.
Fast forward to today. Given my background, I should have realized that AI mentorship was necessary in today’s businesses. Having confirmed this with my first client, I’m eager to help more people.
I’ve found evidence that the lack of mentoring is a huge problem. Some articles even indicate it could cause an economic crash worse than the .com bust. This is scary. Having experienced the .com bust myself, I’d much rather be a force to prevent it than let it happen.
Although my experience has always been with developing technology, I also learned what to make to solve business problems. I focused on providing a good user experience (UX) while developing solutions to business problems. Fast forward to what’s happening now. Like it or not, AI is accessible and forces everyone to adopt technology. What’s not accessible is mentoring for when and how to use it.
As a mentor, I teach some basic things about AI to anyone, especially non-technical clients. The first thing is how to craft a prompt. From there I identify the tasks they perform to accomplish their unique job. Everyone’s job is unique to some extent. Given this base knowledge I find the tools they have available and advise on how to use them. The tools may be ChatGPT or CoPilot, but they also include software they use and even basic tools such as Microsoft Word and Excel. I help them use all of these tools more effectively.
One key concept I try to convey is what AI really is. 80% of what AI provides is just a replacement for invoking functions in existing software or software that is knowledge built into the AI itself. This part of an LLM is called a transformer. They understand the language you speak make it possible for the other 20% to do something useful with your “commands”.
Finally, I always, always, impress upon my clients the need to a) inform people that they used AI to do their work and b) verify that the result of work done with the assistance of AI is correct. After all, the human is hired to do the work because they are capable of doing the job and are ultimately responsible for the outcome.