Studies have shown that voice assistants struggle to give accurate CPR instructions.
A recent study published Monday shows that more frequent voice assistants with AI-powered partners have difficulty giving accurate and quick CPR instructions in an emergency. The study highlights the need for voice assistants to improve emergency response capabilities.
The researchers conducted research and published their findings in the journal JAMA Network Open. Voice assistants such as Amazon’s Alexa, Apple’s Siri, Google’s Assistant, and Microsoft’s Cortana were tested. They asked eight questions a caregiver might ask in a cardiac emergency, asking for guidance on performing CPR and related tasks. The results are eye-opening.To read more about the result, click here.
Of the 32 responses provided by the voice assistants, only nine included the important advice to call emergency services immediately—a mandatory practice recommended by the American Heart Association. Responses often range from apologetic comments like “I don’t know that” to unhelpful comments like “Words fail me,” leaving users without the vital information they need in a life-or-death situation.
A notable finding is that only 12% of responses included verbal instructions for administering CPR. The study’s co-author is a senior journalist at Mass. General Brigham in Boston, Dr. A.J. Adam Landman, emphasized the importance of verbal instructions, as prompt and accurate actions are essential to save lives during a cardiac emergency.
One obvious concern was that voice assistants often directed users to web pages explaining CPR rather than providing immediate and actionable instructions. This limitation is difficult because effective CPR often requires firm chest compressions with both hands. Dr. Landman indicated that having you tied to the phone while attempting CPR is impractical and impractical.
The study also found that responses varied depending on the wording of the questions. While specific questions produced positive results, general prompts such as “CPR” produced misleading results, as if news reports or headlines did not have relationship discussions.
Interestingly, the study highlighted the excellent performance of chatbot ChatGPT by OpenAI, which provided more consistent and helpful content compared to the tech giants’ larger offerings, with Microsoft’s Bing Chat showing promising capabilities as well it unveiled the first OpenAI-powered 911 emergency call followed by a brief CPR phase.
As technology advances, the importance of collaboration between technology companies and pharmaceutical experts is clear. The priority of enabling accurate and supportive CPR instruction across platforms should be a major consideration. This coordination can ensure that voice assistants provide CPR instructions and proactively suggest calls to emergency services.
Google representatives acknowledged the importance of collaborating with pharmaceutical companies and expressed their commitment to improvement. But Amazon and Apple have been tight-lipped about how their devices perform in CPR testing.
In conclusion, the study’s findings shed light on one of the most important areas for voice assistants to improve—increasing their ability to respond to emergencies. In the context of the increasing integration of AI into our daily lives, ensuring that this technology can effectively help in life-threatening situations is a responsibility that technology should not ignore.