#37 Society and Technology

AI-powered chatbots, which we now rely on for information and conversation on nearly every topic, are on the verge of profoundly reshaping our consumption habits. One of the most concrete signs of this transformation emerged in a visual leaked onto social media in recent days. It is claimed that code embedded in the beta version of ChatGPT aims to convert its chat interface into an advertising platform. This leak has generated serious concern among many experts and users regarding ChatGPT’s fundamental reliability and neutrality.

When the internet first emerged, its core mission was to enable data sharing. Over time, however, one of its revolutionary transformations was the adoption of shopping practices onto its own platform. Today, the logic of commerce and advertising reshaped by the internet is entering a new era in the age of artificial intelligence. Now, there is growing debate over whether recommendations and responses generated by AI-based chatbots will directly become advertising content.

A critical example of this hypothetical scenario emerged recently through social media leaks pointing to “AdTarget” codes within ChatGPT. However, to fully understand this development, one must examine its background—the institutional structure of OpenAI. Since OpenAI announced the ChatGPT model, one striking fact has drawn attention: approximately 620 employees hired by the company in recent months previously worked at technology giant Meta. As is well known, Meta is a company that stands out primarily for its advanced advertising algorithms and massive social media platforms.

Key figures such as App CEO Fidji Simo, Head of Technology Vijaye Raji, and Head of Marketing Kate Rouch have brought to light the central claim that some of OpenAI’s recent strategies and initiatives clearly resemble tactics previously used by Meta to transform itself into a corporate behemoth.

OpenAI itself has also shifted its focus toward adopting a similarly massive commercial model, aiming to satisfy investors and justify its staggering $50 billion valuation acquired just a few months ago. This strategic direction naturally raises concerns that Meta’s advertising-driven business model will be transferred to OpenAI. These leaks and corporate histories prompt users to ask: Will ChatGPT, originally designed as a multi-purpose tool, abandon its role as a trusted source of information and become instead a new medium for delivering targeted advertisements?

The latest information leaked from ChatGPT’s Android beta version indicates that OpenAI is actively working to transform its chat interface into a fully functional advertising platform based directly on user requests and intentions. Internal code directories discovered in ChatGPT’s Android app beta version 1.2025.329 provide the clearest evidence of this shift. Expressions such as “bazaar content,” “search ad,” and “search ads carousel” confirm that OpenAI is actively testing sponsored product cards and scrollable ad carousel templates for search-engine-like responses.

Initial signals suggest that advertising mechanisms will initially focus on search results. This means advertisements will not be randomly inserted as banners into every conversation, but rather integrated as sponsored responses or product rankings within answers generated by retrieving information from the internet. Thus, ads will be presented to users not as classic promotional formats, but as natural, embedded replies within the conversation. This approach aims to make advertising an inseparable part of the conversational experience.

The financial scale of this move is enormous. With approximately 800 million weekly users, 2.5 billion daily prompts, and an estimated 5 to 6 billion monthly visits, ChatGPT commands massive user traffic. Even a very low advertising load across such a vast volume could generate significant revenue streams for OpenAI. These leaks strongly reinforce expectations that artificial intelligence will become the next major marketplace for digital advertising.

OpenAI CEO Sam Altman’s apparent ambivalence on this issue is also noteworthy. In a previous program at Harvard University, he described advertisements as a “last resort” and characterized them as “uniquely annoying.” However, his recent statements suggest a shift in his position.

In a podcast, Altman stated: “I find ads a bit unpleasant, but they are not a barrier… I like some ads. I give Meta a lot of credit—Instagram ads provide me with clear added value. I enjoy Instagram ads. It’s as if I didn’t even know I wanted this thing.” These remarks gain significance when considering that many of OpenAI’s employees were transferred from Meta.

Elon Musk continues to shape the technology agenda. In his latest announcement on X, he stated that Grok will face T1, the team of esports legend Faker and reigning champions of League of Legends, next year. This event, which could be seen as either a promotional stunt or an escalating AI competition, will demonstrate the power of artificial intelligence and, specifically, the technique of Reinforcement Learning.

To better understand this matchup, it is useful to briefly revisit 2019, when OpenAI Five defeated OG, the Dota 2 world champions. At that time, OpenAI demonstrated that artificial intelligence could defeat humans in complex strategic environments. However, the conditions Musk proposes for Grok place the challenge in a distinctly different category. Rather than directly accessing the game’s code, Grok will only process camera feeds and will not exceed human reaction speeds. This transforms the test from a pure “processing power” contest into a competition based more heavily on intelligence.

The real focus here is not the game’s score or outcome, but the underlying technology. Reinforcement Learning is a machine learning technique that trains software to make decisions that maximize desired outcomes. Language models like ChatGPT, which we use daily, are fundamentally systems built on text and prediction. Reinforcement Learning, however, learns through trial and error toward a defined goal. It improves its performance by simulating millions of scenarios, receiving rewards for correct moves and penalties for incorrect ones. Games like League of Legends, which involve “Fog of War” and other macro and micro mechanics requiring real-time decisions under incomplete information, offer an ideal laboratory for testing the limits of these models.

For readers unfamiliar with the gaming world, here is why this matters: if an AI can collaborate effectively within a five-player team in a chaotic, dynamic environment like League of Legends and develop strategic responses, that capability can be adapted to the real world alongside robotics. The decision-making mechanisms developed in games can be applied to numerous domains—from managing the chaos of financial markets to enabling autonomous vehicles that adapt to unpredictable human behavior. Grok’s test against T1 will provide compelling data on whether artificial intelligence can evolve beyond being merely an assistant and become a true “agent” capable of navigating uncertainty. Regardless of the outcome, this process is highly likely to contribute significantly to research on Artificial General Intelligence (AGI).

During his first foreign visit to Türkiye, Pope Leo XIV made various assessments on global peace and regional developments. Another notable aspect of his remarks was his commentary on artificial intelligence. Since his election in May 2025, the Pope has consistently addressed AI in his statements. In his official speech during the Türkiye visit, he emphasized the need to examine artificial intelligence from ethical and moral perspectives, warning that AI could reshape human preferences and deepen existing injustices if not guided by an ethical framework.

Pope Leo XIV characterizes artificial intelligence as one of the transformative elements of our era and describes this technological evolution as a second industrial revolution with strong historical parallels. It is said that the new Pope’s choice of the name Leo, following Francis, references Pope Leo XIII, who responded to the social and economic transformations of the industrial era. This further supports the interpretation that the “age of artificial intelligence” will have societal impacts comparable to those of the industrial revolution. The Pope’s statements can be read as an attempt to position the papacy not merely as a theological authority but also as an active participant in broader social change processes.

Although the Pope’s engagement with AI-related topics may initially appear unusual and diverse in scope, careful readers will recognize this as a superficial reading. Given that the papacy is viewed as a religious authority, a more systematic approach would be expected. Yet by addressing AI through the lens of general societal impact, the Pope reveals his alignment with the populist narrative surrounding artificial intelligence.

Overall, the Vatican’s approach to artificial intelligence presents a cautious but not oppositional framework. This stance acknowledges AI’s inevitable transformative power in modern societies while advocating that this transformation must be guided by human-centered values. Although critics may argue that the Pope offers nothing new by merely echoing existing concerns, the mere initiation of these discussions creates space for deeper and more authentic debates in the future. Moreover, the Pope’s engagement with timely, agenda-setting issues like AI can be interpreted as a deliberate effort to expand the papacy’s political and social influence.

Beyond this, it is reasonable to expect that all religious institutions and authorities will increasingly incorporate the impact of artificial intelligence and technological advances on human life and society into their agendas. The crucial question, however, is the depth and authenticity with which they do so. Superficial assessments signal a failure to fully grasp these developments, and their interpretations remain little more than wishful thinking.