The Coronavirus Virus or Virion as biologists know it is only about 200 nanometers in diameter and it is spreading across the globe like wildfire.
But how will you know when you get the Coronavirus or even the regular flu?
You could get tested?
Still a far more interesting question is how technology can diagnose you remotely… and sometimes without your consent.
Social Media Mining
Originally when we got sick we had to complain to our family. Then they invented money and work so that we could finally complain to our colleagues. People at work later devised computers and over time the internet evolved. People got bored with the internet and wanted to chat, leading to the monumental triumph of social media.
But the authorities still wanted to know what we were saying. First, the Department of Homeland Security instructed the social media giants to watch the network for chatter on crime and terrorist threats. Then Influenza came around.
Google was the first to monitor our annual battle with the flu with their discontinued Flu Trends service. Google cancelled the public service five years ago after announcing that the figures were spotty and unreliable. However, colleges and universities can still purchase access to the figures today. The decline of Flu Trends opened the door for social media to get into the business of Influenza monitoring. Four years ago Stanford University ran a symposium on the state of health data mining on Twitter. They found that Natural Language Processing (NLP) is good at identifying specific medical symptoms that people complain about in their tweets. Further insights into people’s moods about disease can be achieved by analyzing their sentiment. IBM provides the Tone Analyzer tool for analyzing user sentiment and you could build a social media API with the right expertise. Scientists have already used sentiment analysis to better understand people’s feelings in response to drug use like marijuana, but the effectiveness of Sentiment analysis is limited by the short duration of tweets.
No high-level engineering knowledge is required to build a health sentiment system. IBM already provides the tools.
You are not alone anymore. There is always someone listening. Your Amazon smart speaker is listening and probably your Google speaker too. I wish that I could say that it is a closely guarded secret, but people have known for at least four years after a former Amazon engineer accidentally discovered that the device repeated fragments of commands that he had earlier entered. People have also received audio files in error sent from other Amazon customer’s speakers.
No high-level programming skill is required to implement a health surveillance feature. You just need to train Alexa with a custom skill.
First, create a skill and train it to recognize sickness, “cough!, cough!”. Skip the response. You can also add additional triggers for the illness.
With this skill created you can now create a custom Lambda function to push the response to the skill into the cloud.
Alexa has always been listening to you. Amazon likely knows when you’re sick as soon as you do.
Warrantless surveillance generally gets a bad reputation. You don’t want your government or your employer watching you all the time.
But what if surveillance could remotely diagnose you with a disease and get you help?
The AI startup SenseTime in China is pioneering the application of detecting the compliance and general health of its citizens using surveillance cameras. China’s present population is over a billion people and a nearly impossible task for anyone to monitor in real-time. AI can enhance this task by training the algorithm to identify people’s compliance in wearing masks. With additional training the algorithm can further identify if you have a fever and even your personal identity. China is already harnessing the technology to arrest and identify citizens violating the Coronavirus lockdown.
These automated surveillance schemes represent a very Orwellian prospect for the future of health diagnosis. When these measures are implemented without the consent of customers, they could be seen as a cure worse than the disease. Yet when these tools are inserted into useful applications, they could disrupt the very practice of healthcare by notifying people of sickness before it’s serious.