IDEO: AI should only be added to physical products when it solves a real user problem
Designer Savannah Kunovsky from the firm IDEO describes, in an interview for IEEE Spectrum, when integrating AI into physical products helps the user and when it is merely a reaction to pressure from investors and competitors.
Savannah Kunovsky, head of the industrial design team at the design firm IDEO, describes in an interview for IEEE Spectrum how the company decides when to add AI to a physical product. According to her, part of the demand for AI integrations stems from pressure from investors and competitors – companies want to have an "AI story" for investors ahead of results, or are reacting to a competitor's product rather than to a user's need. The test IDEO uses is simple: does the given feature help a person do what they want to do, in a real context of use.
According to Kunovsky, companies often reach for large language models because they are the most visible, even though a broader range of technologies exists – computer vision, sensor data fusion, or models running directly on the device. Choosing the appropriate technology for a specific task is, in her view, just as important as deciding whether to use AI at all. Because AI introduces a new interaction paradigm, it must, in her opinion, be simple – as an example she mentions the first Apple mouse with a single button.
As a problematic example, she cites the Humane AI Pin, a wearable device without a display that replaces a smartphone with a projection and an AI assistant – according to Kunovsky, it was an interesting provocation, but the technology was not ready and was slower and less reliable than an established phone interface. Another example is a client who wanted a stove without knobs, controlled by voice and AI; during testing, smoke started rising from the pan and the user reached for a nonexistent knob before the AI managed to react, so the firm recommended keeping the knobs. Kunovsky also mentions the shift of air conditioning and volume controls in cars to touchscreens and voice menus, due to which the organization Euro NCAP now rewards physical controls in its safety ratings.
According to Kunovsky, for tasks requiring an immediate reaction "by feel" or under stress – a volume control, a light switch, a stove burner, an emergency stop – it is advisable to keep mechanical controls, because they are faster and more reliable than a system that must first interpret the user's intent.
Why it matters
The interview gives companies and designers concrete criteria for deciding when to add AI to a product and when not to, instead of relying on a fashionable trend. The practical benefit is a set of test questions and concrete counterexamples (Humane AI Pin, a stove without knobs, touch controls in cars) that show when AI worsens usability instead of improving it.
Two audiences, two different impacts
What this means
For individuals
For people designing products or deciding on their features, the interview offers a concrete test: assess whether an AI feature helps the user in a real context of use, or merely replaces a working solution with a new and less reliable one.
For a business
Companies developing physical products risk worsening the user experience and incurring unnecessary costs if they add AI under pressure from investors or competitors without verifying a real need; the recommended approach is to keep physical/mechanical controls for tasks requiring an immediate and reliable response, and to choose the specific AI technology (LLM, computer vision…
StrategyCheck the original
Event sources
only one source so far · 1 publisher, 1 independent. We count feeds from the same owner only once.