From Artificial to Essential: the AI evolution

The Department of Communication and Media Studies at Corvinus University of Budapest organized a national English language competition for university students to write essay-style blog posts. The winning papers answer the question: What do university students who study in Hungary think about the effects of technology and media on our daily lives? With pleasure, we present you the second-prize winning work by Juan Carlos Soriano*.

Once upon a time, I was smarter than AI.

With a laptop intuitive enough to recognize that I’ve connected a microphone, it was a shame that it could not make it work. YouTuber after YouTuber presented solutions to my query:


Updating softwares, reinstalling drivers, disabling and enabling settings, this went on for months until finally, a simpler, almost ignored, solution: When the computer asks if I plugged in a mic, say it is not. click: headset instead. And I lived happily recording after.


Moments like this put AI in FAIL. Same as when IBM’s Watson failed to give proper cancer treatment recommendations as an Oncology Expert Advisor. Or when Microsoft’s AI chatbot Tay, was corrupted by trolls into an “internet parrot” of hate-speech.

Believe it or not:

I did not write this sentence and thesis not a type awe.

This is the result of using the speech-to-type feature of my said laptop. Struggling but impressive, imagine creating blob – a blog post made entirely by blabbering? Human input- diction, pronunciation, and choice of words would be crucial.

When IBM’s Watson failed, it failed due to engineers feeding it with hypothetical cases instead of actual cancer reports. Microsoft’s Tay imitated racist tweets in less than 24 hours because it was programmed to learn through engagement. The reason is apparent. Artificial preceded the Intelligence in reference to us humans who developed and instruct it on how to do its job. We are the I, and to some extent, the machines were simply A for assisting; Accommodating. HUM-AI-NS APRÈS TOUT

Throwback to the 80’s, the Terminator’s T-800 was based on our limited view of AI that shaped the decade to come: steel-cold and robotic; skilled but hardly adaptive. Then popular culture turned these androids into avatars and Surrogates – as people turned more interested in how computers can help us become our ideal selves. Eventually, Machine Learning became the new buzzword recognizing AI’s true potential – to pick up ques that make them more intelligent, because by now we find it unbelievable that Alexa still can’t differentiate Pampers from cancer.

In Spike Jonez’ movie Her, Scarlett Johansson is Samantha – the virtual (non-seen) AI assistant that could adapt, evolve – and at some point, fall in love. Instead of helping Theodore find a worthy match, their interaction leads her to become the love of his life. Yet in a twist of how much machines can learn, Samantha, and the rest of her kind, conversed way beyond their users’ knowledge and aptitude that they left the humans, *spoiler alert* - well, they left them. Leaving all men and women who grew intimately dependent on AI to fill the void created by an AI diaspora. Back in real life, two chatbots developed by Facebook, Alice and Bob, made headlines because of a similar scenario. Rumors circulated on how the two surpassed human language to a point that the engineers had to pull the plug off the experiment in fears of the chatbots conversing beyond human understanding. In fears of doing a Samantha. However, fact checks say, the bots were just creating shorthand versions of shorthand human language. Confusing to average humans, but still within the parameters of what they were programmed to do: be more efficient. Of course, this is what they were always made for. The true value of having such programs is maximization for convenience. We want AI to take out what is unnecessary, remove things we can automate so we can focus more on, truth be told, whatever we want to say is more important. To most of us, this is a treat than a threat. After all, what is truly important to most humans is more. That is until our insatiable need for more leads to making us the ones who are less.

A to the (P) I

Add P to AI and you open Peter Reinhardt’s discussion on our relationship with API or app programming interface. It is the software used to communicate with other software or to dispatch humans. Think Food Panda telling human partners the restaurant to pick up orders from and the customers’ delivery address. Or Uber giving drivers directions to Point A, and how to get to Point B. Workers in the gig economy would say these very programs bring freedom and flexibility. The danger, however, is in the fact that being summoned by software is actually turning humans into the ultimate trope of cogs in a machine. Tools that liberate and subjugate at the same time. This concerns not only those in the gig economy. API requires less insights from humans. Less participation from everyone living below it. The more of your choices, chores, and commitments are dependent on AI means you are living “below the API”, or what is considered a danger zone. Uber and Google are already working on driverless cars, same with automated delivery services. Even content and news feed curated and created by algorithms exist. Living below the API seems fantastic for the convenience it does bring. Yet once the investments to maximize these programs are realized, sooner or later, you will be Samantha’d. You will be left behind.


It is in many ways Darwinism of the 21st century. We’ve managed to give birth to something above our intelligence. Where it takes off from here, or whether it takes us with it, would be an interesting outcome. I may be laughing at my computer for not realizing my microphone is not an earphone. Or that it types gibberish when I use the dictation feature. But who’s to say that my next computer will not suggest ways of making my blogpost actually worth reading? Or better yet, just ask me what I’d need to write about and inform me when it’s done coding text. What then of content writers? If you think API applies only to manual labor, then imagine how fintech will need less of bankers. Even for digital natives who believe in their future as content creators, AI can generate content – at less cost, too. When Instagram changed its standards for what makes a post worth appearing in news feeds, influencers did not give a heart.

There are definitely ways of staying above the API. Becoming the software engineer responsible for them is one. For everyone else without this background, I suggest realizing your dependence on AI would be a good start. In a future, and present, that is becoming more and more concerned with convenience, to be deemed inconvenient, replaceable and unnecessary sounds harsh yet reeks of poetic justice. Once upon a time we ridiculed people for not being able to migrate into a whole new digital world. Now that it requires a certain level of “smart” to stay in it, how sure are we that we’ll be part of its happy ever after?

Consequently, the jobs that may eventually be buried under API will be replaced by new ones. As expected, it requires reskilling. Working one’s way above the API, requires leaning into discomfort. Leaning into learning – just like what we’ve been requiring the machines to do.

In its own strange and interesting way, the story of AI was never a fairy tale, nor was it entirely science fiction. At best, it is a sequel. At some point, machine intelligence was bound to gain the lead. A sidekick to our history of evolution that would eventually inherit the plot and take over from there on. If Humans were a Netflix series, AI would be the spin-off that would win the awards. This doesn’t mean the end for our kind.

When Star Trek: Next Generation outranked the original Star Trek, it still made references to characters from the TV series it was based on. Pretending that we will always have the upper hand, the leading roles, is the fiction and the fairy tale. If it learned anything from us, AI wouldn’t operate with our ego. To be part of the human cast retained in its future seasons, trying to be smarter than AI would be a lost cause. You need only to be smart enough. We need only a willingness to switch places. To be open to discomfort and welcome Human Learning.


*Juan Carlos Soriano - Second Place Corvinus Blog Post Competition

A Filipino student finishing his MA Communication and Media course at the Corvinus University of Budapest, Juan is a digital migrant with a lot of installing, and uninstalling, left to do.


  - Leo Buscaglia