As of October 2018, this is what Alexa is used for…
It would be interesting to see statistics from other devices such as Google home and Siri, however, I am sure they are very similar. I think this is the emergence of a new digital ecosystem, similar to how the iPhone established the Apple Store and blossomed across the platforms into Google Play or the Microsoft store for Apps. Similarly, like the Apple App Store which started with rudimentary Apps like the calculator and calendar, Google and Amazon are now allowing third parties to integrate onto their platform.
A pharmacist the other day said to me that the these AIs will one day listen out for our coughs as a keyword much like “Hey Alexa” or “Hey Google”. In theory, if it had all of your coughs recorded it could compare the pitch of the sound, i.e. if it is of a higher pitch or contains more bass than normal this could indicate inflammation in the airways or the converse COPD. At the other end, it could record your gaseous emissions, distinguish the type of sound and diagnose gastric irritation.
Where I belive the pharmacist got it wrong was the fact that he said a prescription would then be created as a consequence. I think this may be taking things a little too far. I think it would need a more holistic approach, not just diagnosing from one metric, but a clinician would use all of the data available to them and draw a picture.
I think many people oversimplifying the diagnostic procedure, you always want more data, in fact, you would be negligent in not having a look at the extra data if it is available. The question is can AI get better in the amalgamation of this data. i.e. your google home finds that there is an increase in pitch of your cough, it then checks your fit bit to see if you have exacerbation asthma this then links into your phones ECG recorder and finds that you have a cardiac issue instead of a respiratory issue. Similarly, with flatulence, it could identify the sound which related to liquid stools, check your Amazon/Ocado/Tesco account and find out what you have been eating recently and record a data point of potential food intolerance.
