The Haunting Side of A.I. Technology

Do you believe you have a fairly strong grasp of the reality that surrounds you, whether in real life or in the digital world? We are bombarded with visual content on a daily basis, from art, to high fashion ads, and we know that Photoshop is pervasive in advertising of course. But let’s say you were shown a series of standard-looking photos, such as the following examples…

Here are some people:

And here is some contemporary art:


And how about some pictures of the internet’s favourite thing, cute kitties?

Do you notice anything… “off” about any of these photos? Well, you may or may not be surprised to know that none of the subjects of these images exist in reality. The people, the art, and the cats were all generated by an A.I. That’s right, Artificial Intelligence generated all of these images by learning from a library of real subjects, and then spitting out brand new images that look very much real.

Wild! And perhaps a little disturbing. But the technology doesn’t start off this sophisticated, it has to progressively be improved upon and new iterations are developed on an ongoing basis just like any other tech. The A.I. that generates the fake people for example has gotten pretty darn good. You probably would not be able to detect that the people are not real in the vast majority of instances if you were presented with the images without prior knowledge of the technology behind it (in this specific case, the A.I. being used is called generative adversarial network). However, just by looking at some of the cats being generated by the same A.I., we can see how bad the initial results can be before they reach much more realistic results.

Just for fun, here are some examples of less-than-good “cats” generated:

A little wonky but not terrible, still looks like cats! Let’s see another one:


Oh boy, so here we are getting into nightmare-inducing territory:

Ok last one, promise:

As we have seen, the A.I. needs to learn over time in order to get better as its given task. Basically how machine learning works.

So what can this type of technology be used for in real-world applications? One use for this type of A.I. is to generate images of fashion /makeup “models” in order to eliminate the need to hire an actual person for your photoshoot. Of course there have been other more worrisome potential purposes that this technology could be used for, and some parallels have been drawn to Deepfakes.

Other Less Haunting Uses

If you want to learn more about another fascinating (yet less disturbing) application of A.I. technology (chat bots and self-driving cars are pretty cool, too, but everyone already knows about those) look no further than the automated data analysis done on the interests of people in order to better understand specific demographics, and ultimately target them based on what would end up being the most relevant marketing content.

For example, here at Mapendo we leverage A.I. to assist apps in the realm of marketing. Collecting (anonymous) statistical data, it’s possible to determine if someone who is in Milan, and is using an LG phone and is playing Sudoku is likely going to buy a pizza in the next 24 hours using their mobile device. Neat!

These results are obtained by analysing statistical data on millions of users every day, while at the same time adhering to privacy as the users’ personal information, such as name and last name, which is never revealed.

The algorithms build profiles which establish the most likely behaviour of each individual. All of this data is continuously collected and studied by a proprietary platform (called Jenga) which spits out the most relevant correlations. For example a user “profile” that is likely to order that pizza, can also be one that is likely to order sushi, interact with an eye glasses ad, or respond to a new app promotion.

This can become useful especially looking into the future as we become more and more connected to our mobile devices, which we rely on to complete a multitude of tasks from ordering food, to getting a car ride, entertainment, but also health and fitness especially in the most recent months as a worldwide pandemic has made the digital world even more important to our daily lives. Not only for allowing us to stay connected to the outside world, but also to keep us sane (Headspace and Calm are two examples of apps which encourage users to meditate, and remain level-headed during a difficult time). Useful apps are everywhere but of course the level of usefulness is unique to each individual, and if A.I. can help us get to the most useful apps for us that would save us a lot of time and effort.

The Future?

All in all Artificial Intelligence and new technology will continue to be pervasive in our societies and will continue to grow and evolve. Hopefully we will also heed Elon Musk’s warnings and try to put in place some sort of framework and ensure the technology is strictly regulated.

Receive updates on the most interesting and important Industry topics by subscribing here.

Mapendo is a tech platform for app user acquisition. We deliver huge volumes of CPA & CPI conversions to the best mobile apps available with A.I.

If you are looking to facilitate and automate your user acquisition process, all while getting trusted fraud protection, visit our site or get in touch with us!



We are Mapendo. Curious and innovative. Mobile App Marketing is what we do and w. Artificial Intelligence is how we do it. || Bologna, IT ||

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store