Series 15 – Episode 4 – The impact of AI in the electronics industry

The survey, titled ‘AI in Electronics: Balancing Promise and Pragmatism’, aimed to explore the practical realities of AI adoption beyond the prevailing hype.

Power shared that the motivation behind conducting this survey stemmed from the burgeoning presence of AI in the news and the significant investments being made in the field.

He noted: “AI is obviously in the news every day, it really burst out in the last couple of years. I mean, it's obvious that AI has been around longer than that. But really, it's these last few years, sort of post-COVID that it's really been in the news, and there's huge investments going on.”

The survey gathered responses from around 400 industry professionals, revealing several notable insights:

Security risks over job losses: one surprising finding was that security risks were the top concern among respondents, surpassing worries about job losses. This highlights the critical importance of addressing cybersecurity in AI implementations.

Popularity of ChatGPT: ChatGPT emerged as the most widely used AI tool, significantly ahead of other tools.

Power remarked: “ChatGPT is still by far the most used AI tool in the market, way ahead of all the others in terms of the respondents of our survey.”

Rapid adoption: the adoption of AI has accelerated, with the majority of users starting to use AI tools within the last 18 months. However, a quarter of respondents reported not using AI at all, suggesting room for further adoption.

AI's influence on design and development cycles was another focal point of the survey. Faster design and development cycles were identified as key opportunities provided by AI.

Power explained: “We've seen AI hit software development much earlier than electronics design. GitHub, in particular, provided a great training set for AI models, making it easier for AI to model how to write code.”

While AI has made significant strides in software development, its impact on hardware design remains limited due to the lack of comprehensive training data. Nevertheless, AI is gradually making inroads in areas such as PCB layout and chip design.

The survey also highlighted lower costs and better customisation as benefits of AI.

Power elaborated: “If you can use it to accelerate things like code development, it makes it easier to customise software for people. It's a lower incremental development cost, which can make previously unfeasible customisations affordable.”

Power noted that large language models, such as ChatGPT, continue to dominate AI usage, particularly in document generation and chatbots.

He observed: “We're seeing AI use on a practical level more around the edges at the moment on things like document generation, chatbots, even things like marketing.”

Looking ahead, Power predicted that as confidence in AI grows, there will be a shift towards more specialised tools. The industry might also see a shakeout, with a few major companies emerging as leaders due to their extensive resources and advanced models.

At ByteSnap Design, AI is being used cautiously, particularly in software development, due to concerns about the provenance and quality of generated code. However, the company sees significant potential in support functions such as technical writing, document creation, and customer support.

Power noted: “We're keeping a very, very close eye on the developments of things that we can use to part development cycle.”

The survey ‘AI in Electronics: Balancing Promise and Pragmatism’ offers a valuable snapshot of the current state of AI adoption in the electronics industry, highlighting both the opportunities and challenges. For those interested in delving deeper into the findings, the full report is available on ByteSnap Design’s website.

To hear more about AI in electronics and much more, you can listen to Electronic Specifier’s interview with Dunstan Power on Spotify or Apple podcasts.