The revolution brought about by ChatGPT in February 2022, then in version 3, was not primarily due to any significant new capabilities in terms of technology but to the fact that this OpenAI product did three things critical to its impact on society. Firstly, it created a service that is easy to use and accessible enough that you don't usually need special knowledge or skills. While it was possible to work - in many ways - with more exciting models with various expert systems, accessibility became one of the critical parameters of the revolution.

The second important dimension was working with language. Turing imagined that intelligence and language formed an inseparable whole. Suppose a system can communicate in language so precisely that we cannot distinguish between man and machine. In that case, we can say that Turing's thought is intelligent, opening up a great debate about how language and thought are related. After all, Ludwig Wittgenstein's famous quote, "The limits of my language are the limits of my world", goes in the same direction - language and the means of language determine how we can think, what the world looks like, which we can instrumentally grasp through language. Working with language is thus a crucial issue in our occidental cultural circle because we have become too accustomed to identifying thought and speech. ChatGPT's focus from the beginning has been on the fluency of dialogue, not on factual correctness or concrete usability.

The third aspect associated with the ChatGPT revolution is the ability to work with natural language at the input level. While searching in databases, working with specialized tools, or even searching in a library, we use and encounter specific language different from how we usually speak. However, with ChatGPT, users talk to it like a person - some assign it a gender, say thank you, goodbye, please. These features have made artificial intelligence more accessible to imagine and grasp, though at the same time also frightening to some.

We will be interested in how these language models can be used for creative work with information. We face a situation where the technology is so new that examples of good practice form a limited circle of inspiration. Part of what we call creative is the search for further options and solutions. If we try to offer certain avenues of invention or answers to specific problems below, these should always be seen as starting points for further creative work.

To illustrate the extent of ChatGPT's capabilities and creativity, the results of the Torrance Tests of Creative Thinking, a standardized creativity test conducted by the University of Montana, can be cited. It showed that ChatGPT 4 scores better than 99% of the US university population, at least in terms of novelty, and overall, it scores 95% in creativity. Thus, it is more creative in the standard test. The question, of course, is whether what we are testing with the test is creativity or only part of it, and possibly what kind of creativity.

The first area we'd like to mention is the ability to discuss an issue with ChatGPT. One of the essential tools of critical thinking is to work with counter-arguments or arguments that take a position opposite to our own. This is where critical and creative thinking clash interestingly - on the one hand, we focus on the detail in the form of an argument (critical thinking). On the other hand, there is a transcendence of our existing views, so there is a kind of out of box thinking process (which is typical of creative thinking).

For example, let's consider writing an essay on regulating car traffic in the city. Our position is that cars do not belong to the town and people should use other modes of transport. In such a situation, it is pretty relevant to try to use ChatGPT to ask for arguments for and against:

**Title: The Necessity of Regulating Car Traffic in Cities: Embracing Alternative Modes of Transportation**

Introduction:

The rapid urbanization and population growth in cities have led to numerous challenges, with traffic congestion being prominent. As the debate over the role of cars in urban environments intensifies, it becomes evident that regulating car traffic is a crucial step towards creating sustainable, efficient, and livable cities. This essay aims to present four compelling arguments in favor of regulating car traffic in cities, followed by an exploration of four counterarguments against such regulations.

**Arguments in Favor of Regulating Car Traffic:**

1. **Reduced Traffic Congestion:** The presence of a high number of cars in city centers is a significant contributor to traffic congestion. Congested roads lead to wasted time, increased air pollution, and heightened stress levels for commuters. By implementing regulations that encourage alternative modes of transportation such as public transit, biking, and walking, cities can alleviate congestion, making urban travel more efficient and pleasant.
2. **Improved Air Quality:** Cars are notorious for emitting harmful pollutants that degrade air quality and pose serious health risks to city dwellers. The implementation of car traffic regulations can lead to a substantial reduction in air pollutants, resulting in cleaner and healthier urban environments. This not only benefits the well-being of residents but also contributes to the fight against climate change.
3. **Promotion of Sustainable Transportation:** Focusing on alternative modes of transportation promotes sustainability by reducing carbon emissions and dependence on fossil fuels. Regulating car traffic can pave the way for investments in public transportation systems, cycling infrastructure, and pedestrian-friendly streets. These changes align with the global push toward more sustainable and eco-friendly urban planning.
4. **Enhanced Quality of Life:** Excessive car traffic negatively impacts the overall quality of life in cities. Noise pollution, road accidents, and the limited availability of green spaces are some of the consequences of car-centric urban planning. By regulating car traffic, cities can create quieter, safer, and more attractive spaces for residents to enjoy, ultimately enhancing their quality of life.

Conclusion:

While the debate surrounding regulating car traffic in cities is complex, the arguments in favor of such regulations highlight the potential for improved urban living conditions, reduced environmental impact, and enhanced sustainability. Nevertheless, it is essential to address the valid concerns raised by opponents, ensuring that any regulatory measures are balanced, well-planned, and accompanied by investments in alternative transportation infrastructure. Ultimately, the path forward involves finding a harmonious balance between personal freedoms, economic considerations, and the urgent need for more sustainable urban environments.

This (relatively long) text does not need to be read carefully to easily see the fundamental issues that need to be dealt with if we want to convince anyone of the merits of regulating car traffic. We have four primary benefits and four possible negative attitudes that must be addressed. This example is a bit of a stretch in that probably anyone could similarly make these arguments, but it is nevertheless valuable in that:

• It clearly names the possible counter-arguments, and the essay writer can critically and systematically refute them one by one and take them seriously. This usually positively affects the quality of the whole text and its argumentative depth.
• For some topics, the benefits or problems may not be so obvious. In the Czech environment (and not only in the Czech environment), for example, there is repeated talk about free public transport and its benefits and problems. And keeping track of these aspects is not nearly as easy.

Suppose we get down to the actual writing and devising of the essay. In that case, we will probably come across other points in processing and reading the literature (or we can have ChatGPT generate other issues) that are possible and appropriate to work with. The critical point is that this person must accept responsibility for their position and the arguments they wish to make. But suppose we want to use this tool to its full potential. In that case, it's worth discussing with them (e.g. "Okay, you're describing the economic impact of regulating car traffic in the city. Are there any examples of how cities have dealt with this? Or which have reverted to less regulation?"). Discussing with ChatGPT is very practical because the discussion is an essential tool for thinking and critically considering alternatives. The more and more carefully we talk with ChatGPT, the more depth of thinking we can get into on a particular issue. In this way, we can quickly achieve that the resulting text will work with different arguments and methods of thinking than the one we would have written without ChatGPT.

Here are some more ideas to improve the results of the creative process:

ChatGPT is not the only tool that is built on large language models. Another much-discussed one is Google Bard, which, more than dialogue, follows the plane of an expert tool that allows one to search for information in a broader context. Typically, it can answer questions or requests like, "I need the Houellebecq novel. Where is the main character actor." or "Name of the main character Houellebecq's first novel". What is typical about these questions is that they are more akin to that particular language of search than natural language (but it works with that, too) and that the answers you get are more clearly structured and challenging. They are, therefore, primarily used to support the creative process when you don't want to spend time searching for specific information about which you have only limited ideas. Typically, a character with a particular description from a Houellebecq novel is relatively easy to find.

Claude.ai is another exciting tool that can efficiently work with data, such as custom files. It is. Therefore, a tool that can help to analyze data and thus greatly help in a creative approach in science where the first analysis would take too much time or where we are looking for new and unexpected connections and at the same time do not have sufficient mathematical programming apparatus.

For academic purposes, we can certainly recommend the Perplexity AI tool, which does not aim for fluent dialogue but for answering complex questions by searching for and citing expert sources. Especially for a basic orientation in an expert topic, this tool makes sense to combine with ChatGPT, which can be more creative and work better with text.