Al & Machine learning
Facebook 10 year Challenge- Welcome to Artificial Intelligence
In my coaching practice, there is a technique that I use mostly to put my clients mind in a future frame. This technique helps motivate you to drive towards your future goals. On the flipside, you mind mostly powers your motivation in reverse when you think about the past, more so if the past is undesirable. I have been conjecturing about this technique and the motivation behind Facebook’s “10 year challenge” frenzy that has taken the world by storm. From a glimpse of the comments on friend’s timeline, I observed that it has elicited a feel good spirits elevating mood and not a bad way to start the year, isn’t it? But hold your horses, that challenge isn’t facebook happy new year gift to you. As you will soon find out, facebook, whose business model revolves around understanding human insights, might just have pulled off one successful Artificial Intelligence AI campaign to tune their facial recognition algorithm. What this means for real people rather than data scientists is that the data Facebook is storing could be used to open your phone, identify you in certain databases, or even used to make purchases.
Artificial intelligence is just but a fuzzy term with no agreed definition and can pretty much mean anything, so I will choose to focus on the technology driving it. When faced with daunting challenges like image recognition, the particular technology that’s deployed is machine learning simply, when computers learn from experience. For example, Facebook would want to automate the tagging of your photo posts every time you post without asking you tag. This will require machine learning algorithms to glean through your existing verified images and the list of records that data is experience. It’s history from which to learn. After which the computers to do what computers do best, number-crunching and optimization, uncovering the patterns and trends so you can classify new incoming photos and classify them according to the features that are matched on the database.
So to overcome the limitation of us trusting that the machine has discovered something valid, that what it learned will hold true in new situations never before seen we have to train the data against a real situation, and that is what facebook has done by crowd sourcing the “10 year challenge” so as to tune their prediction algorithm. Your data is only as good as the training the machine learning programs receive. Thus it only seems natural that the Facebook “10 year challenge” that has been extremely popular over the last few days, could actually be a viral ploy to get users to categorise data showing how their facial features changed over ten years.
Potential Benefits
We live in a world full of events that are interconnected, this connections are reflected in all the tons of data being collected. That’s what makes data the worlds most potent, flourishing natural resource. The world has enthusiastically dived into the world of golden age of data discoveries. A frenzy of number crunching is churning out heaps of insights that are colorful and often surprising. After all, its said that machines are really taking over. Computers decide or at least help decide a number of real life applications. Facial recognition and other biometrics are starting to be used for many types of identification. The iPhone X uses facial recognition to open the screen and verify purchases, and other types of digital ID rely on this verification as well.
For the case of facebook, the potency of the facial recognition algorithm has immense applications such as in field of predictive policing, where law enforcement institutions can glean social media data and pin point crime prone individuals and apprehend then before they commit the crime. There are also advance facebook is making in the medical care study by analyzing images to predict your aging rate and this same algorithm does indeed predict when you will die. That data is used to fuel research in healthcare, credit scoring modelling where condition and character are key parameters of affordability of a loan facility and also creating a profile of a fitness for a certain job helping human capacity professionals to hire the right talent for the vacant roles within the organization.
There are thus several connections found in data and each one helps predict things and these serve as foundational building blocks with which machine learning builds predictive models. Facebook has indeed demonstrated a major paradigm shift of traditional research of forming hypothesis then test it to validate future occurrence’s. So as you take part in the Facebook ten-year challenge, know that you not just doing it for fun, its business!
More about this is covered under the topic Raising your Data Quotient in my upcoming book- Big Data and Predictive Analytics
By Timothy Oriedo: Author and Data Scientist at Predictive Analytics Lab Timothy.oriedo@predictiveanalytics.co.ke
Hi! I simply want to give you a huge thumbs up for the excellent information you have right here
on this post. I am returning to your web site for more
soon.
We appreciate the feedback. Keep on checking our blog for more information