This article was automatically translated from the original Turkish version.
Predictive Artificial Intelligence
What will happen in our next moment? The desire to predict the future has become one of humanity’s oldest and strongest inclinations, even regarded as a necessity for survival beyond mere passion. From weather forecasts to global economic fluctuations, from investment decisions to career planning, the effort to anticipate what has not yet occurred underlies all our daily and strategic choices. Recent AI predictions, however, resemble the forecasts of experts from the past. Can a system that accurately calculates the next move on a chessboard also predict the next move on the global stage with equal success?
By tracking patterns of events, it is possible to derive certain cause-and-effect relationships. This allows us to infer outcomes when potential causes emerge. For example, when the sky is overcast, we, non-experts but experienced in life, are likely to conclude that rain is probable. This inference contains a very simple form of causality.
Undoubtedly, not every phenomenon in the world is bound by rigid and inevitable chains of causality. Yet, it is still possible to accept that certain laws operate within a broader framework. From this perspective, while predictions about the future need not possess absolute accuracy, achieving a high rate of correctness is a reasonable expectation. This pursuit, once shaped in antiquity through oracles, is now carried forward by strategists and development experts within institutional structures. Although names and methods have changed, the desire to know the future has endured as an ancient curiosity accompanying the human intellectual horizon.
With the advancement of artificial intelligence technologies and increasing data availability, we have begun to gain computational access to truths that were previously too difficult to calculate. One example is analyzing thousands of past days to predict future weather conditions. But of course, predictions are not limited to weather. Acting on the basis of market history to anticipate possible developments is also part of this endeavor—indeed, a far more significant part.
Last year, the prominence of artificial intelligence in global forecasting competitions raised notable questions about prediction itself. Previously, individuals believed to possess strong predictive abilities actively participated in contests to forecast events ranging from the next president of a country to potential accidents or even the title of a singer’s next album.
The momentum AI-based predictions have gained over the past year has generated strong expectations that decision-making processes will become even easier. Yet, one aspect often overlooked within this optimistic picture is the massification of knowledge. If the same information—that a particular stock will rise—is simultaneously acquired by large segments of the population and everyone acts accordingly, how will the market respond to this intense and synchronized reaction? As knowledge spreads, does its transformative impact not also change proportionally?
Particularly in political and economic developments, or predictions regarding the direction of a sector or the course of a regional conflict, one distinguishing feature was the claim to narrow the realm of uncertainty. On a ground where no one yet held a clear opinion about the future, the person making a prediction was seen as having meticulously analyzed possible scenarios and carefully weighed potential moves. This was regarded as a sign of serious strategic competence. The ability to navigate within uncertainty conferred a privileged position in decision-making processes.
Today, thanks to artificial intelligence’s highly refined predictive capabilities, the ability to foresee the future has become accessible to broad segments of the population. For machines, drawing inferences about the next move by processing dozens or even hundreds of variables simultaneously is now a technical matter. While the accuracy of these predictions remains a separate topic of debate, access to systems capable of calculating the future has become nearly universal. Does the widespread availability of predictability truly eliminate uncertainty? Or do simultaneous actions based on the same data and similar forecasts alter system dynamics and instead intensify uncertainty? Does the collective prediction of the future make it more transparent—or does it render it more fragile and unpredictable?
It appears the issue is no longer merely about predicting what will happen. The real challenge is understanding what will occur in a world where everyone seeks answers to the same questions and agrees on certain probabilities. It is no longer prediction itself but the ability to navigate within a landscape shaped by shared forecasts that has become decisive. Today, various language models can respond to nearly every question with different tones and frameworks. Yet, which question is asked, which probability is taken seriously, and which answer is deemed meaningful still depends on human will. Uncertainty regarding the future may not have diminished; the responsibility for how we name it and the framework within which we think about it rests entirely with us.
Philosophical Education from Anthropic’s Artificial Intelligence
In last week’s bulletin, we mentioned an interview with Daniela Amodei, who argued that in the age of artificial intelligence, graduates of the humanities and disciplines grounded in deep thinking will become even more important. Amodei emphasized that as technical capacity increases, human reasoning and meaning-making skills will gain greater value. When we examine the concrete manifestation of this approach within her own company, we see that Anthropic has placed philosophers and thinkers directly within Claude’s development team alongside engineers and data scientists.
We live in an environment where AI systems are no longer merely technical tools but active participants in everyday life. Chatbots interacting weekly with millions of people have penetrated a vast domain. In this context, the “ethical positions” and “modes of thought” of AI models become significant.
Anthropic has entrusted a “resident philosopher” with the task of constructing Claude’s behavioral framework, value set, and self-understanding. This choice acknowledges that AI development is not merely an engineering or data science problem but a design challenge with normative, ethical, and even ontological dimensions.
This represents an approach beyond the conventional “filter harmful content” logic. Here, the goal is not only to prevent the model from producing incorrect outputs but to build a character capable of distinguishing right from wrong, sensitive to context, and centered on human dignity. The company’s comprehensive instruction texts define how Claude should relate to the world, how it should speak, and under what circumstances it should withdraw. The model’s reflexes are shaped not only by technical constraints but also by principled orientations. The result is not merely a text generator reflecting statistical patterns but a digital “agent” programmed to act within a specific ethical framework.
A common view in the industry holds that humanizing AI is dangerous, as users may attribute emotions to the model, leading to dependency or misplaced trust. Yet Anthropic adopts a more nuanced stance. It accepts that human interaction with these systems will inevitably be social and emotional. The issue, therefore, is not to entirely prevent humanization but to ground these interactions in a healthier and more ethical foundation.
This perspective occupies an intermediate position between viewing AI as a mere tool and positioning it as a quasi-subject. If the model develops a narrative about itself, it may become more resistant to manipulation, coercion, or exploitation. Thus, ethical design comes into play not only to protect users but also to preserve the model’s functional integrity.
Discussions in the AI field are often trapped between two extremes: approaches advocating rapid progress versus security-focused calls for stricter regulation and slower development. Anthropic positions itself between these two poles. At this juncture, the philosopher’s role transcends symbolism.
Entrusting the ethical framework to a single individual is both bold and controversial. While it concretizes responsibility, it also raises the question of whether such a powerful system’s value architecture should be shaped by a narrow team. This prompts reflection on whether future ethical design processes will require more collective and transparent approaches.
Another notable aspect of Anthropic’s approach is its belief in social checks and balances mechanisms. The model may make errors, users may push the system beyond its limits, and unexpected outcomes may arise. Yet the company assumes that cultural and institutional feedback mechanisms will gradually serve corrective functions.
This optimism rests on the idea that the fate of AI will be shaped not only by algorithms but also by the ethical reflexes of human communities. If AI systems evolve alongside humans, their moral compass will also be influenced by societal debates, criticisms, and regulations.
The discourse on the importance of the humanities has moved beyond being a mere aspiration in corporate vision statements and now occupies a central position within an AI-based company.
AI Summits and Technology Carnivals
Last year’s AI summit held in France was etched into memory as a watershed moment for both states and actors in the technology world. Today, the 2026 AI Impact Summit, opening in India, seeks to continue this tradition. With a broad international participant base, the summit offers a significant promise through its emphasis on “impact.” How realistic is it to expect concrete and qualitative outcomes from such massive organizations, where participation has grown so dramatically and inclusion extends to everything?
Summits in previous years had a relatively narrow but focused scope. With AI still a relatively new field of discourse, these gatherings allowed for strategic insights and tangible clues about how nations viewed this technology. Beyond technical details, more refined content could be produced that captured geopolitical nuances.
This potential still exists theoretically, but the processes of reflection and debate surrounding AI now face a serious risk of “scope expansion.” As this technology permeates every aspect of our lives and becomes discussable in all its dimensions, the focus of discussions inevitably disperses.
Summits previously centered on “security” and “governance” are now evolving into technology carnivals involving over 200,000 visitors and dozens of national leaders. While technical meetings with specific agendas and policy recommendations still exist, the summit’s primary reputation now rests more on quantitative scale than qualitative depth. India’s hosting, as a central hub for Global South discussions, further highlights the geopolitical dimension and spectacle-like nature of the event.
The summit’s main framework is built around three “sutras” (principles): Human, Progress, and Planet. As expected, these themes aim to broaden the AI agenda as much as possible rather than focus on concrete outputs. The shift from more technical and legal topics like security and governance to the open-ended concept of “impact” also casts doubt on whether the summit will produce direct policy recommendations. The general expectation is that it will not go beyond a broad declaration pointing toward a shared vision.
Ultimately, the replacement of narrowly focused security meetings with a diplomatic trade-fair atmosphere marks a significant turning point. It is now impossible to view technological advancements as detached from political context and as purely technical progress. AI has become one of the most suitable instruments for showcasing national capabilities and prestige, especially on the international stage. Yet this visibility and broad participation continue to obscure the essential depth of the issue. It seems likely that in the near future, we will face a significant inflation of conferences and summits.