Column: Embedded design
A hollow world By Myk Dormer, RF Engineer and Director, Smallwireless Limited I
can’t be alone in noticing that what is generally referred to as “AI” is creeping into practically all areas of human activity. AI-generated art seems to be fl ooding the Internet, every search engine or browser
seems to be sprouting “assistants” or prioritising AI-generated results, and you can hardly read an editorial or a forum without tripping over someone extolling one AI tool or other. Looks as if it’s crept in engineering, too. Initially, when this whole subject fi rst
cropped up – and it’s not that many years ago (things are moving fast!) – I can’t say I was particularly concerned. My thoughts could probably be described as “generally positive”. I could see the use in image-processing tasks, including much- vaunted MRI image analysis and screening, in-vehicle navigation and industrial automation. We had seen buzzwords like “machine
learning” and “heuristic systems” fl oating about for some time, and this didn’t feel
AI will have a signifi cant impact on engineering skills development, which may not necessarily be a good thing
What gets referred to as “large language models” began to subsume the whole “AI” label, and the
fi rst of the “chatbots” and “AI art tools” appeared, initially with some fairly comedic results
much diff erent. Computers being used for highly repetitive, computationally intensive tasks is pretty much the defi nition of what they are. However, things seemed to shiſt fairly rapidly. What gets referred to as “large
language models” (LLM) began to subsume the whole “AI” label, and the fi rst of the “chatbots” and “AI art tools” appeared, initially with some fairly comedic results. But, these systems (almost by defi nition) refi ne themselves over time and, in eff ect, learn by experience, to the point that as of the end of 2025 we are seeing systems capable of plausible conversations and able to produce images that take some eff ort to identify as not human generated.
AI in engineering So far, so good, even though I’m not really here to examine the recent history of AI, or attempt to cast a crystal ball prediction of its eff ect on future society. Rather, I am concentrating on its infl uence on engineering, and the strange little RF related niche that I inhabit. So far, that infl uence is aptly described by one word: “Malign”. T e thing is, it shouldn’t be at all
surprising – and I’m talking “in an engineering sense” here. It’s in the inherent nature of an LLM to regurgitate content that it has “found” (or is that “stolen from”?) elsewhere, cunningly edited into a form that seems to mirror what the user is expecting to see or hear. T is may be fi ne if the user wants a synthetic friend who can tell them where to buy good (or rather: well regarded) coff ee, but as an engineering tool it is terrible. If you were trying to fi nd a defi nition of
what engineering isn’t, then “a completely uncritical re-capitulation of the existing state of the art, without any innovation” would come pretty close. If we move from specifi cally the design
aspect of the engineering process and look at some of the more peripheral activities, such as researching information, or writing documents, we hit another AI landmine.
14 December 2025/January 2026
www.electronicsworld.co.uk
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48
