Image courtesy of Giovana Fleck/Global Voices.
The debate on AI, like anywhere else in the world, was brought to Turkey with the naivety that accompanies most modern innovations. Could AI drive a car? Could a robot talk to a human? Could AI cure cancer? Would all social problems be solved in seconds now that perfect, honest and unbiased technocrats were born? In other words, every existing problem was framed in AI. The possibilities seemed endless! AI: the king of the jungle, the mighty lion, who would eliminate the flawed hypocrites and save us with his perfection.
But while discussions about AI are widespread and accessible in other parts of the world, that is not the case in Turkey. At least not yet. While there are many AI experts, they rarely interact with the general public. Because of this, the understanding of AI is very broad: for most people, AI means machine learning and related technologies. So there is a noticeable lack of overall knowledge about the technology, including its limitations and growing concerns about its broader impacts. Looking at examples of AI use in other regions, we can already see that the noise surrounding the technology is far from the majestic roar of a lion, but more like the annoying buzzing of a mosquito.
It is less than 40 years since the triumph of neoliberalism, which proclaimed the end of epic narratives, politics and history. Today, these statements are considered even more absurd than they were decades ago, because history manifests itself clearly and violently in every instance. And yet, the idea of a depoliticized and enlightened autocracy with an impartial and incorruptible technocracy survives in many places, including Turkey. Perhaps this quest for incorruptible rule is not shocking, given that Turkey ranks 115th out of 180 countries in Transparency International’s Corruption Perceptions Index. The king of the jungle, our AI overlords, have arrived, solving every debatable issue with the certainty only a science fiction android would have.
The use of AI in elections and political intimidation
So where and how is AI being used in Turkey? Recently, AI “magic” was used in the latest city council elections. When answering questions about the controversial list of candidates ahead of the local elections, Özgür Özer, leader of the main opposition Republican People’s Party (CHP), pointed out that the culprit was AI. Apparently the party leader had relied on the technology before announcing his candidates. Meanwhile, Murat Kurum, Istanbul mayoral candidate of the ruling Justice and Development Party (AKP), vowed to use AI to solve the megacity’s zombie traffic problem. Successfully guessing a candidate’s chances of victory and balancing it with the party’s values is a difficult task. As a city council candidate in a megacity, it is also difficult to express one’s position on the status quo of car-centric infrastructure and public transport. Whatever decision one makes means political sacrifices for rejected possibilities, unless, of course, there is a perfect machine to make those decisions. AI benefits both sides.
But as the election results show, even AI couldn’t help the ruling party in the local elections. The AKP received the lowest vote share ever in municipal elections in Istanbul and across the country. AI also didn’t help the main opposition Republican People’s Party (CHP), which won a historic victory (with its controversial candidate Ezer chosen by AI) but actually lost to the AKP candidate. To be sure, neither outcome was due to AI, even in part. But you can be sure that these examples will be brought up whenever someone mentions AI in a political context again.
When AI can actually impact and influence people’s lives, the stories about it are less implausible than the purported uses of AI in politics and its hypothetical uses.
Most real-world examples and case studies present AI as far from an all-purpose technologist capable of solving complex problems from local politics to city infrastructure. These examples and cases show the nuisance, if not the misuse, of AI technology at best.
Search for “yapay zeka” (artificial intelligence in Turkish) on any online search engine and the first results you see are praising AI, similar to sentiments shared by politicians. The potential and (sometimes inaccurate) actual uses of AI, mainly abroad, are shown through rose-colored glasses. Keep scrolling and another side of the AI debate emerges, where there are legions of people complaining about AI, describing it as annoying and frustrating. Devoid of empathy and capabilities, AI technology is beginning to resemble the late cultural critic Mark Fisher’s notion of a “boring dystopia.”
The dark side of AI technology
AI’s worst feature isn’t just its frustrating nature. It has a darker side. Even before the Taylor Swift deepfake alerted the US Congress, there were cases in Turkey where deepfake techniques were used for sexual harassment. A young woman sparked a public debate when she warned the media about a stalker who used deepfakes based on her social media images. Just before the 2023 general election, Erkan Baş, a member of parliament and leader of the Turkish Workers’ Party (TİP), also had AI-generated pornographic photos of himself circulated on Twitter. The images have since been completely removed from social media. These two examples make it clear that relatively unknown people and public figures alike are not immune to this postmodern method of sexual harassment.
Another way deepfake technology is being used in Turkey is through voice imitation, where scammers have tried to imitate the voice of President Recep Tayyip Erdogan to extort money from victims. The mosquito metaphor should be clearer now than it was at the beginning of this article. For many, the AI buzzing around in our heads is annoying, irritating and threatening to carry old diseases more efficiently.
Of course, the debate that exists will follow the actual usage. For example, there is no open debate about using AI to replicate public discrimination, as happened in the Netherlands, because in Turkey AI is not yet used in public services. As the use of AI continues to grow, so will the debate. But this growth and debate should be supported by an expansion of critical debate by people who actually understand the technology and its good, bad and ugly aspects.
Ultimately, while the technology has potential applications, the current state of AI in Turkey is more like a pesky mosquito than a majestic lion, an apt analogy given the lack of expertise even in public debates: after all, lions don’t really live in the jungle.
