easyDNS is pleased to sponsor Jesse Hirsh‘s “Future Fibre / Future Tools” segments of his new email list, Metaviews
A treasure trove of intimate data begging to be analysed
The problem with data is that it is for all practical purposes an entirely abstract concept. As a word it has become pervasive in our culture, mentioned in all sorts of contexts, and regularly referred to as commodity, property, and fuel.
However just because this word is practically ubiquitous, does not mean that we understand the broader concept that it represents. The current use of the phrase data is arguably subjective, as each person would imagine something different when the word is evoked.
Social media has played a prominent role in our collective awareness of data, as we recognize it as the resource of value, and superficially understand that social media companies are built using our personal information.
Yet this understanding remains largely superficial precisely because we do not actually understand how our personal information is being used, and the phrase data is so wide and inclusive, that it can literally mean anything and everything.
Instead it is the actual applications that employ that data the help us understand its value, and appreciate the myriad of ways in which it can be employed. Even then however, we are just getting glimpses of the versatility and malleability that enables a diverse range of applications arising from having that data.
This is why we should regard the rising digital monopolies with such skepticism and suspicion. The issue is not just what they’ve done with our data, or what they are doing with it, but what they will be able to do with it moving forward and into the future. What is possible and comprehensible changes with each day, as new connections and correlations are made from the patterns and shadows found in our respective data sets.
As a result our mobile devices remain incredibly valuable, and the data they generate or harvest from our activities continues to pay off.
Generally speaking we underestimate what sort of insights or analysis can be done using the information our mobile device is in a position to collect. Our limited understanding or appreciation of the concept of data often constrains our ability to anticipate or imagine what is possible.
If you take a moment to think about the intimate relationship we have with our mobile device, there’s really not much that the device doesn’t know or is in a position to know. What secrets could we hide from the supercomputer we carry in our pocket or purse that not only has a camera, microphone, and internet connection, but also a range of sensors that are sensitive enough to detect a remarkable range of movement and activity.
Just like visiting a mystic or oracle, the issue is not in what they will tell you, but in what you wish to ask, and the knowledge you seek. In this regard, our mobile devices remain relatively untapped resources for revealing some of our most intimate and personal details.
Phones can detect when you've had too much to drink by monitoring your walk, study finds https://t.co/dOH8tiZbRZ
— BBC News (UK) (@BBCNews) August 19, 2020
American researchers used sensors in smartphones to detect when somebody was over the legal drink-drive limit.
Phones were able to do this with about 90% accuracy when users walked just 10 steps in the study by the University of Pittsburgh.
Scientists hope the discovery can be used to develop device alerts, such as asking people not to drive while drunk.
“We have powerful sensors we carry around with us wherever we go,” lead researcher Brian Suffoletto said. “We need to learn how to use them to best serve public health.”
It’s not a stretch to imagine that one of the legacies of this pandemic will be to reinforce the role and authority of public health officials. This is why the question of data usage and analysis for the purposes of public health is presently crucial as it will inevitable create precedents.
Drunk driving is a serious public health issue that results in a high number of deaths and injuries that could and should be preventable. The question becomes how, and the temptation to use data and research like this would be difficult to resist.
“This is a controlled study but in any wider public application, you’d have to consider how this data is collected and used,” Professor Daniel Dresner, a cyber security expert at the University of Manchester, said.
“There is no limit as to the ingenuity of the uses that data can be put to or misused but, it’s important to remember how this science may be adapted in future. Could it connect to the car immobiliser so it won’t let you drive? Will it alert a friend, or the authorities, if you shouldn’t be driving?”
The mobile device is a powerful tool precisely because it helps us navigate our world and organize our information. However in an increasingly connected world, this device will inevitably become an enabler of other connected machines.
The example given in this article anticipates how our mobile device will connect and converse with vehicles we may use or ride in. This is already the case for ride sharing and late model automobiles that automatically connect to our phone for navigation and entertainment purposes. How long before authentication and compliance with local laws also becomes part of that technological handshake?
Here’s the actual research paper cited in the article:
A Preliminary Study Using Smartphone Accelerometers to Sense Gait Impairments Due to Alcohol Intoxication (free full text for @JSADJournal 81/4)@pritikadasgupta @DrErvinSejdic @PittTweet @PittDeptofMed @StanfordEMED#smartphone #alcohol #machinelearninghttps://t.co/qDc3XnOzxB pic.twitter.com/ZAJapZXkp7
— Journal of Studies on Alcohol and Drugs (@JSADJournal) August 18, 2020
Sensing the effects of alcohol consumption in real time could offer numerous opportunities to reduce related harms. This study sought to explore accuracy of gait-related features measured by smartphone accelerometer sensors on detecting alcohol intoxication (breath alcohol concentration [BrAC] > .08%).
The claim that this is possible in real time is also the appeal of the mobile device itself. That it offers instant gratification and information on demand. Should we not be concerned that also includes diagnosis on demand?
In an issue we published in November of last year on digital psychiatry, we noted that algorithms are being developed that will automatically diagnoses mental health issues based on how we use our mobile devices. Here’s a relevant excerpt:
The research that is available overwhelmingly looks at patient facing apps that are targeted towards individuals and are relatively consensual.
However the real threat, as cited in the paper above by Floridi and friends, is the use of digital psychiatry technology in the enterprise. Identifying and tracking down the use of such tools will be difficult, as they’ll be embedded into the design of applications that help human resource departments, marketing, operations, education, you name it.
Anyone wanting to understand the needs and vulnerabilities of people they can surveil will be tempted to use tools steeped in digital psychiatry.
All of this raises the issue that if you could know, should you know? For example should bars be compelled to monitor how drunk their patrons are to prevent drunk driving and other liabilities? This could be surreptitiously done via an app that the bar insists patrons install for loyalty and liability reasons. the need to enable contract tracing in case of the coronavirus is one example of how this could happen.
Although the larger concern in all of this is accuracy. That was the most disturbing aspect of the digital psychiatry tools. Only a tiny fraction were backed by credible evidence and had any claim of accuracy.
Similarly the research around detecting if a person is drunk is also preliminary and basic. Yet that’s never stopped technology companies from adopting and deploying this kind of research as if it were indisputable.
Another reason why being able to dissent or appeal such analysis and conclusions is essential to preserving fairness and human rights.
Insurance companies are increasingly eager to get their customers to install apps that monitor how they drive (or help them make claims if property insurance) and offer discounts in exchange. However no discount will reflect the value and vulnerability that come from them having your data. Who knows what they will choose to conclude in the future?!
This is why we should not underestimate how valuable our information is, and how our data can and will be repurposed in the future.
Tragically privacy is not only something we have to vigilantly protect, but also something that remains elusive and difficult to understand in a world that is constantly evolving. While it is exciting that we’re able to push technology to new limits, it also means we need to constantly (re)assess how to do so responsibly.