Personality predictions based on LinkedIn are already scarily accurate

Roel Castelein takes a look at the future of personality analysis and what it might mean

At a recent tech startup reception I bumped into an old colleague. After catching up he showed me, as nerds love to do, something ‘new’ on his smartphone. The app analyzed my LinkedIn profile on the fly fairly accurately and matched my profile with others to provide insight on how we could collaborate together. The app, called Crystal Knows, is scary, not in what it does today, although impressive, but what it means for the near future.

Install Crystal Knows free as a Chrome extension and you can check any LinkedIn profile in seconds based on the DISC (Dominance-Influence-Conscientiousness-Steadiness) personality classifier. Additionally you can compare two profiles and Crystal will suggest how you can work with potential contacts. Upgrade to a paying subscription and Crystal, like an experienced PA, will suggest how to best approach people based on their profiles or how to write them effective emails.

Under the hood of Crystal are clever ‘text analysis’ algorithms that analyze the content of your communication, meaning what it finds on LinkedIn, and then with statistical modeling, judges your personality according to the DISC personality classification system invented by W. M. Marston, a physiological psychologist. As with all statistical models, the more data Crystal collects, the smarter and more accurately it will judge you.

This is compounded by other systems. Have a look at Facelytics, much more then scanning your face as a replacement for the security on your iPhone X, it also detects gender, age and emotions through your selfie camera. Next step would be to use the app in recruitment processes.

All combinations and suggestions need to be pre-programmed, but this technology is in its infancy. Anybody familiar with Artificial Intelligence, Machine Learning and Data Science should see where this is heading. Our smartphones and the stuff we post on social media are such rich data sources that Crystal and Facelytics are just the beginning of a wave of intelligence apps that will weave every bit of information together and based on underlying models – be it personality, marketing, or fraud detection models – determine who you are, how to treat you or what to sell you. And the outcomes will become more accurate over time.

Scary? You bet. ‘Interface’ is a clever novel about a biochip implanted in the US President’s brain that lets him receive instant feedback of what the impact is of his speeches through colors. This enables the President to instantly adapt his wording according to what is received well. It raises his manipulation skill to a superpower.

Mix this sort of interface with our social media sharing and existing technology and we have the potential for many types of analysis or classification; health risks, credit ratings, job applications, propensity for fraudulent activity or likelihood to buy certain products or services, all within easy reach of clever algorithms.

Should you be worried? Yes and no. Like any instrument ‘public data and statistical models’ can be used for good as well as for evil. But just remember the next time you post something nice on Instagram or share professional content on LinkedIn, Crystal and her peers are not only watching and analyzing you, they are growing smarter by the day. They know where you live, what personality you have and, just like in the Minority Report story of 1956, they will predict your behavior in a scarily accurate way.