Can AI be your “virtual Elmer”? 

Amateur Radio has always thrived at the intersection of curiosity and technology. Think of all the major technical achievements in the hobby from the earliest vacuum tubes to transistors to integrated circuits, and from homebrew receivers and transmitters to commercial transceivers and to today’s Software Defined Radio. Advances in technology have consistently expanded what operators can achieve to better enjoy the hobby for themselves and for the community they serve. 

We now find ourselves at the frontier of a new, exciting, and mysterious technology and I want to make the case that it is up to us to make full and constructive use of it in our daily pursuit of this wonderful hobby. 

Enter Artificial Intelligence, or AI. AI is a broad field focused on creating computers that can perform tasks normally requiring human intelligence, such as reasoning, learning, and decision-making. A key part of AI is Machine Learning (ML), where systems learn from data instead of being programmed with fixed rules. One important area of ML is Large Language Models (LLMs), which are designed to understand and generate human-like text using large amounts of training data. You have undoubtedly heard of, and possibly experimented with, some of the emerging LLMs such as ChatGPT (on which Microsoft Copilot is based), Claude, Gemini, or Llama. Today’s column will focus on LLMs but I’m going to use the more familiar “AI” as shorthand. 

First, let me begin by acknowledging that I think we all have concerns about the way AI is currently being developed, including issues like: 

• Too much control by a few, very wealthy technology companies who have a disproportionate influence over powerful AI which will affect society, jobs, and information, 

• Hard to understand or even trust these opaque AI systems which are known to reflect bias and can sometimes produce misleading or harmful content, and 

• Social and economic risks, as AI may use data without clear permission, disrupt jobs, and be misused for scams, deepfakes, or manipulation. 

That said, by all accounts, AI is a giant leap along an old path of remarkable innovation and technological advancement. It will not replace the ham operator, rather it could be a powerful and deeply helpful companion for learning, experimenting, and understanding this hobby to a greater depth than otherwise possible. It is poised to assist both young and old, newbie and experienced-ham operator alike. Here are a few examples of what it can do: 

• Newcomers and Amateur-Radio veterans alike struggle with topics like impedance matching, antenna resonance, modulation schemes, and radio-signal propagation. There is plenty of folklore and misunderstanding that gets passed on from one ham operator to the next. AI can explain these topics in plain language, offer analogies, and walk through examples. Instead of searching through multiple websites or thick textbooks, the ham operator can now ask targeted questions and get sound, technical explanations tailored to their level of expertise. 

• For those keen on getting their Amateur Radio certificate, AI can generate practice questions, explain wrong answers, and focus on the subjects with which a student needs the most help. It can be a tutor, an assistant, and an “Elmer” (an Amateur-Radio coach or mentor). Its assistance can create an effective, efficient, and friendly learning environment and it can make licensing more approachable for people who might otherwise be intimidated by the technical and regulatory content. 

• For operators who enjoy kit building, homebrew, or troubleshooting challenging technical problems, AI is ready to assist. 

AI can be an infinitely patient, deeply knowledgeable partner that adapts to the operator’s level of expertise and particular interests. AI doesn’t replace the hands‑on experience, problem‑solving, or human mentorship that define this hobby. Rather it can enhance them, offering explanations, examples, and guidance that help operators get more out of their time on the air and on the workbench. The conversational interaction with AI chatbots mirrors the guidance an Elmer might provide, but it’s available 24 hours a day and you control the pace. 

That said, one needs to be very precise in asking questions and describing the task to get the answers you need. Even then, AI can slip up. This is particularly true for vibe coding (AI assisted software writing). AI can incorporate the sort of folklore and misunderstandings I mentioned above. It can be a tremendous helper for getting one going on a block of code, but that code may have errors because of assumptions AI makes about the compatibility with the operating system (OS) environment. It’s a partnership where you and AI collaborate on a solution. AI won’t do everything and often needs supervision to keep it on track. I will provide an example in my next column. 

Many have suggested that AI is little more than a glorified search engine, but I think that understates it potential usefulness. I offer the analogy that AI is to a search engine as a graphical user interface (GUI) was to a command-line interface (CLI). The GUI developed in the 1980s was a huge step in making the personal computer a much easier tool to use. It replaced the clunky CLI, which required typing often-cryptic text instructions at a command prompt. That experience quickly made way for a user-friendly graphics environment with a movable pointing device and overlapping, resizable windows, clickable icons, and pull-down menus. 

I want to encourage you to try using a free version of any of the available AIs out there. You can try entering prompts like these: 

I’ve provided this small subset of examples of how AI can assist and be a “virtual Elmer”. You can likely think of lots more ways it might be helpful to you in your pursuit of the hobby. Put it to work for you. Ask it anything and keep experimenting and keep learning. That keeps it interesting.

Last Updated on 2026-01-03 by Joannadanna