ChatGPT: The Study on How We Use It
The OpenAI study reveals how AI has overturned demographics and habits in less than three years. For the first time since the birth of ChatGPT, OpenAI has lifted the veil on its users. As if Willy Wonka had finally decided to open the doors of his digital chocolate factory, the San Francisco giant has published a detailed study that portrays who, how, and why they use its most famous creation.
The results are not just surprising: they are revolutionary. And they tell a story that goes far beyond cold statistics.
The official paper, based on data collected from millions of conversations and in-depth surveys, represents the first systematic attempt to understand the social impact of a technology that has already changed the way we work, study, and communicate. But above all, it reveals an unexpected phenomenon: what we might call the "great demographic reversal" of artificial intelligence.
The Numbers That Change Everything
If in 2022 ChatGPT was perceived as yet another toy for nerds and developers, today the reality is completely overturned. The most striking data to emerge from the OpenAI study concerns the gender distribution: as reported by the Washington Post, the majority of ChatGPT users are now women, a percentage that marks the definitive overtaking of the traditional male dominance in the tech world.
This demographic reversal is not accidental but reflects a deeper change in the approach to technology. While the first users were mainly programmers, researchers, and tech enthusiasts, the second wave of users has brought ChatGPT into the homes, schools, and offices of people who see AI as a practical tool rather than a technical curiosity.
The geographical distribution tells another fascinating story: according to official OpenAI data, the United States still represents a significant portion of global traffic, but adoption in emerging markets such as India, Brazil, and West African countries is growing at a dizzying pace. It is yet another demonstration of how digital technologies can leapfrog traditional geographical and economic barriers, just as mobile phones did with landlines in the 1990s.
Generations in Comparison
If there is one generation that has embraced ChatGPT more than any other, it is undoubtedly Gen Z. With almost half of the conversations studied coming from people aged between 18 and 25, digital natives have transformed conversational AI into a sort of always-available personal consultant. But the real surprise comes from analyzing how different generations use the technology.
Millennials (26-35 years old) represent 35% of the user base and show a more structured and productivity-oriented approach. They use ChatGPT mainly to optimize work, write professional emails, and manage complex projects. It's as if they have found the personal assistant they have been dreaming of since the days of Spike Jonze's "Her," only this time the AI doesn't fall in love with them.
Gen X (36-50 years old), which makes up 15% of users, shows a more cautious but deep curiosity. They are often the first to ask existential questions to the AI, to test its ethical limits, and to use it to delve into hobbies and personal interests. Baby Boomers, although representing only 5% of the user base, show surprisingly sophisticated usage patterns, focusing on historical research, financial planning, and support for health problems.
Each generation has developed its own conversational "dialect" with AI: Gen Z tends to use informal language and treat ChatGPT as a peer, while older generations maintain a more formal, almost reverential, approach. It is fascinating to observe how a technology can adapt to such different communication styles while maintaining its own consistency.
From Code to Kitchen
The most significant change revealed by the OpenAI study concerns the transformation of use from professional to personal. If initially ChatGPT was perceived as a tool for developers and professionals, today 70% of interactions take place in personal contexts, while only 30% are work-related.
This migration from the office to the home has brought with it a creativity of use that not even the creators of OpenAI had foreseen. People use ChatGPT to plan dinners, create smart shopping lists, interpret dreams, write personalized greeting cards, and even as a sparring partner for family political debates.
Educational use represents a particularly interesting chapter: university students use ChatGPT for brainstorming and text revision, while parents rely on AI to help their children with math homework or to explain complex scientific concepts. This has raised important questions about academic integrity, but has also opened up new pedagogical possibilities that educational institutions are still exploring.
The kitchen has become one of the most active laboratories for conversational AI: personalized recipes based on available ingredients, suggestions for specific diets, advice for wine pairings. ChatGPT is becoming the digital sous chef that many did not know they wanted, transforming meal preparation from a domestic chore to a guided culinary experiment.
The Productivity Paradox
While OpenAI celebrates the massive adoption of its technology, independent researchers are beginning to raise more complex questions about the cognitive impact of conversational AI. A study from MIT has revealed worrying data: prolonged use of ChatGPT for writing seems to reduce the activation of brain areas associated with creativity and critical thinking.
The research, conducted on 300 professional writers and students, showed that after six months of intensive use of AI, participants showed a 23% decrease in the activation of the prefrontal cortex during autonomous creative tasks. It's as if the brain, by getting used to delegating complex creative processes, begins to "atrophy" in those specific functions. A phenomenon reminiscent of that observed in the 1990s with the introduction of GPS navigators, when many people began to lose the ability to orient themselves autonomously.
However, the same study highlighted positive aspects: ChatGPT users showed greater efficiency in solving routine problems and reduced "blank page" anxiety. The paradox is evident: AI makes us more productive in the short term but could compromise some of our most precious cognitive abilities in the long term.
Parallel research suggests that the impact critically depends on the mode of use: using ChatGPT as a brainstorming and revision tool seems to maintain brain activation, while completely delegating the creative phase leads to greater technological dependence. The key, according to researchers, is to use AI as an "amplifier" of human capabilities rather than as a substitute.
Digital Geography
The geographical analysis of the OpenAI study reveals disparities that reflect broader global inequalities. While developed countries show high but stable adoption rates, emerging economies are experiencing a real explosion in the use of conversational AI. ChatGPT was downloaded 10.2 million times in India in August, a huge leap from the 2.5 million downloads in the same month of the previous year.
This phenomenon has created what analysts call the "digital leapfrogging of AI": countries that have skipped generations of traditional software are directly adopting conversational artificial intelligence as a primary productivity tool. It is the same pattern observed with mobile telephony in Africa, but applied to natural language processing.
However, this growth also exposes a new type of digital divide: while access to ChatGPT becomes more democratic, the quality of the experience varies drastically depending on internet connectivity and the level of digital literacy. Users in rural areas or with unstable connections often cannot take full advantage of AI's capabilities, creating a stratification in access to artificial intelligence that could amplify existing inequalities.
The linguistic analysis is just as revealing: while English remains the dominant language for interactions with ChatGPT (68% of the total), languages such as Hindi, Mandarin, Spanish, and Portuguese are growing rapidly. OpenAI has invested significantly in improving the system's multilingual capabilities, but quality gaps persist that penalize non-English-speaking users, raising important questions about global technological equity.
The Economics of Artificial Intelligence
The economic impact of ChatGPT goes far beyond OpenAI's revenues and touches sectors that no one had anticipated. According to the latest McKinsey report on the state of AI, generative artificial intelligence is radically transforming productivity in sectors based on the transformation of informational inputs (data, texts, briefings, research) into informational outputs (documents, presentations, analyses, recommendations), with benefits that vary significantly depending on the type of activity and the implementation methods.
This dichotomy is creating a new stratification in the intellectual labor market. Professions that benefit most from AI automation - such as routine copywriting, standardized data analysis, customer service - are seeing downward pressure on compensation, while roles that require genuine creativity and strategic thinking maintain or increase their economic value.
The phenomenon has given rise to what economists call the "AI dividend": workers who master the strategic use of artificial intelligence earn on average 18% more than colleagues who do not use it. It's as if knowing ChatGPT has become the modern equivalent of knowing how to use Excel in the 1990s: no longer a plus, but a fundamental skill.
Simultaneously, a parallel economy of "AI trainers" and "prompt engineers" is emerging: professionals specialized in optimizing interactions with artificial intelligence systems. Startups dedicated to training on the use of AI raised over $500 million in 2024, demonstrating that the technological revolution always creates new market niches even as it destroys others.
The macroeconomic implications are still being studied, but preliminary data suggest that countries with high adoption of conversational AI could see a 2-4% increase in per capita GDP over the next five years, mainly thanks to increased productivity in knowledge-based services.
Towards What Future?
The OpenAI study, while providing an accurate snapshot of the present, raises more questions than it answers about the future of human-machine interaction. The rapid growth in the adoption of ChatGPT - over 500 million monthly active users as of September 2024 - is setting precedents for the integration of AI into daily life that will go far beyond chatbot conversation.
This study from Florida State University documents how the analysis of lexical trends before and after the release of ChatGPT in 2022 revealed a convergence between human word choices and the patterns associated with LLMs.
This suggests that we are witnessing the emergence of a "hybrid AI personality": users who, after months of intense interaction with ChatGPT, begin to incorporate the AI's linguistic and logical patterns into their human communications. This is not a simple stylistic influence, but a deeper change in the way information is processed and presented.
This phenomenon has significant sociological implications. If millions of people begin to communicate using similar structures and logics - those learned from AI - we could witness a kind of "linguistic standardization" that, while facilitating communication, risks reducing human expressive diversity. It's as if everyone started singing the same melody, losing the richness of individual variations.
Ethical issues multiply with use. Data from the study shows that 23% of users have developed a form of "conversational addiction" to AI, interacting with ChatGPT for more than 4 hours a day. Behavioral psychologists warn of the risks of replacing human relationships with artificial interactions, especially among adolescents and young adults who may prefer the predictability of AI to the complexity of real interpersonal relationships.
On the other hand, the same study highlights unexpected benefits: people with social anxiety disorders report improvements in their communication skills after "practicing" conversations with ChatGPT, while individuals with linguistic or cognitive disabilities find in AI a support that significantly increases their communicative autonomy.
The future that emerges from OpenAI's data is neither utopian nor dystopian: it is profoundly human in its complexity. Conversational artificial intelligence is not replacing human intelligence, but is reshaping it in ways we are still learning to understand. As in any great technological transition, from printing presses to personal computers, the true impact will be measured not in the first few years of adoption, but on the generation that will grow up considering conversational AI not as a novelty, but as a natural part of the human cognitive landscape.
The great demographic reversal of ChatGPT represents only the beginning of a broader transformation. The fact that AI has moved from research labs to living rooms, from programmers to housewives, from PhDs to high school students, demonstrates a technological plasticity that surpasses all initial predictions. But it is also a warning: any tool that becomes truly ubiquitous ends up shaping those who use it as much as it is shaped by its use.
The real question is no longer who uses ChatGPT or how they use it, but what version of humanity will emerge from this increasingly intimate coexistence with artificial intelligence. And perhaps, that is a question that not even ChatGPT can answer.