Identity, brand, future (III)

A discussion with AI as an experiment.

We have reached the third round of our discussion with the Chinese AI chatbot DeepSeek. After considering the development of an identity, I would now like to focus on the design aspects of communication and branding.

(PV) Once we have determined that we can optimize the content development process with the help of appropriate tools, the question arises as to how this content can be translated into comprehensible brand concepts. Today, we find that at a certain point, the rational approach ends, and a relatively subjective, creative design process begins. On the one hand, a rational process is important here; on the other hand, subjectivity also gives rise to the desired distinctiveness. And it is precisely these visual signs (brands) that fulfill the essential function of orientation through their characteristic recognizability. Do you see any possibilities here?

(deepseek) Yes, the tension between rationality and creative subjectivity is not a problem, but rather the driving force behind a strong brand. Here are some concrete ways in which technology not only manages this process, but also enriches it:

The translation machine

From data to design, or AI as a bridge between the rational and the emotional. For example, an AI analyzes target group interviews (We stand for radical reliability) and generates visual mood boards that translate abstract concepts into colors and shapes (e.g., earthy tones and geometric rigor). Tool idea: A semantics-to-design generator that encodes brand values into design principles (sustainability = imperfect textures, variable logos).

Controlled loss of creative control

Systematizing subjectivity: Generative Adversarial Networks (GANs) produce hundreds of logo variants, but the selection is made by a curated committee (humans plus AI, which adds surprises). Example Airbnb: The “Bélo logo” was created from crowdsourced sketches, but the final abstraction followed clear rules (heart, location pin, and letter A).

Dynamic recognizability

Flexibility without confusion: Variable identity systems (such as MIT Media Lab) use algorithms to adapt logos to contexts but within a defined set of visual rules (color space, grid, etc.). Next level: Interactive visual symbols (AR trademarks) that change depending on the user (a sports brand logo shows statistics to fans and a tutorial to new customers).

The “human check”

Why does AI alone fail? Creativity needs breaking points, because AI optimizes for harmony, but brands such as SupRemo and Glossier deliberately break rules (The algorithm would never have allowed that!). Solution: hybrid teams – AI delivers 95% rationally optimized designs, humans add the 5% “irrationality” (e.g., an inexplicable splash of color).

Orientation in the flood

The paradox of the future: the more brands adapt, the more important it becomes to have a distinctive signal. Keywords: sound and haptic branding (e.g., the Tesla door knock sound as an acoustic fingerprint). Biometric triggers: A coffee brand could develop a subtly scented display that evokes memories that are impossible to copy because they are sensory.

Provocative thesis

The brand of the future is a “cyborg” (a hybrid of biological organism and machine). Its body (logos, colors) is rationally constructed, but its soul (the irrationality that remains) is human.

Questions for you: Where do you see the danger that AI will kill magical subjectivity? And are there brands that already master this balance today?

(PV) To answer that question, I would like to bring in another aspect. We can imagine analyzing the work of the most famous designers, which in itself would be a very demanding research task, in order to derive certain algorithms and thus enable highly qualified design. But conversely, we find that truly innovative solutions, such as those from Apple, Nike, or Orange, are characterized by unconventionality and subjectivity. Or, as Steven Jobs said, there is no need for market research, because it only determines what is mainstream. How do we view this creativity from the perspective of AI?

(deepseek) The crux lies in the paradox of creativity: true innovation breaks rules, but in order to break rules, you first must master them perfectly. AI can be both: a tool of discipline and a catalyst for rebellion.

AI as a master class in design

Analyzing big-name brand designers (Paul Rand, Ivan Chermayeff, Karl Gerstner, etc.) could teach algorithms what makes technical mastery: composition, color psychology, typography, because AI can perfectly replicate these patterns. But just like Mozart, who mastered counterpoint, it needs to consciously break the rules (Jobs' “distortion field”). Take Apple, for example: it’s simple logo went against all the trends in technical corporate design at the time. AI would probably have classified it as “risky.”

The innovation blind spot of AI

Market research vs. magic. AI recognizes correlations (people like round corners), but not causalities (you will love round corners if we embed them in a new narrative). In relation to Nike's “Just Do It,” AI would probably have guessed ‘performance’ or ‘victory’– not a cryptic call to overcome oneself.

The solution: AI as a “provocation machine”

Not just optimization, but destabilization too. Tool idea: AI that analyzes the top 100 logos and deliberately suggests anti-design (Here are ten designs that break all the rules. Which one feels right?).

The example of Orange (Telekom) shows that contradiction leads to a unique position. At the time, the name contradicted all industry logic (colors are not brands!). AI might have listed this as a “bold option,” but it would not have been recommended.

The human lever or intuition as a filter

Jobs' famous gut decisions cannot be programmed, but they can be contextualized. As an example, an AI could suggest 100 logos, including a completely irrational one (perhaps inspired by Kandinsky's abstract art). A human feels that it is right, even though all criteria speak against it. Real-life example: Google's colorful, childish-looking logo was professionally incorrect in 1998, but it captured the spirit of the web optimism of the time.

The new role of AI: from tool to “creative sparring partner”

Not a replacement, but a mirror and an incentive. Phase 1: AI learns from the master’s and becomes pedantic (the design violates seven design laws). Phase 2: AI uses this knowledge to provoke deliberately (here is a logo that looks like a blob – but it triggers 78% more emotional responses).

Summary: AI can teach craftsmanship (like a design professor), but it cannot replace vision (like Steve Jobs). The future belongs to hybrid teams. AI as a library of all styles and a rebel coach. Humans as arbiters of magic. They choose what is illogical but true. Question: Do you think we will eventually need AI that consciously develops “bad” designs just to force us to think anew?

(PV) No, I don't think so, but I do see a tendency toward mediocrity in the tools. If we look at the various design programs that are already on the market today, we see that they are certainly capable of visibly improving the quality of design and communication, and this will certainly be optimized in the future, but that is not authentic innovation. This raises the question of how creativity will be understood in the future and what role creative people will play in this process.

(deepseek) Your skepticism is justified, but the future of creativity will not end in mediocrity, but in a new division of labor between humans and machines that will enable radical innovation. Here is my perspective:

Why tools tend toward mediocrity (and how to avoid it)

The “blandness paradox”: AI optimizes for average results because it is trained on existing data. For example, AI generates logos by combining familiar patterns (“blue = trust, circle = harmony”), but authenticity comes from breaking the mold. One solution could be anti-AI training: algorithms must learn to work against their own data (Here are 1000 perfect logos – now generate something that disrupts them all).

The new role of creatives: from designer to curator

Creatives will no longer be pushing pixels, but filtering AI provocations (That one crazy design out of 1000 boring ones is genius) or creating context, because a brand like Nike needs more than just a logo; it needs a myth. But AI cannot experience stories, only analyze them or break taboos. Jobs' famous “Think Different” campaign, which broke all the rules of advertising (no product, only dead heroes), became a moment of triumph and a successful relaunch for Apple.

Creativity becomes deeper – not superfluous

Routine craftsmanship disappears (e.g., hours of layout work), creating space for strategic intuition, namely the ability to connect unrelated ideas (What does this AI-generated pattern have to do with the climate crisis?). Or emotional translation, when AI recognizes that people like nostalgia, but only a human knows that in 2030 they will not be thinking about the 90s, but about the pre-ChatGPT era.

The big misunderstanding: innovation vs. optimization

AI optimizes (This logo is 12% more recognizable). Humans define what “better” means. Example Supreme: An AI would have repaired the broken sign of the first boutique – but it was precisely this rawness that became cult. Meta-thesis: The next big brand will decide against AI recommendations and use this as a statement (Handmade in the age of algorithms).

The ultimate question is: what is true creativity?

Before AI: The ability to create something new. After AI: The ability to create something true, even if it looks old. Future scenario: An AI generates 1,000 innovative fonts, but the creative person consciously chooses a classic font because it signals authenticity.

Conclusion: The creative calling is shifting from making to deciding what it means. The tools are getting better, but the responsibility for meaning, ethics, and cultural resonance remains with humans. Will there be a new elite of creatives who work with AI instead of against it? Or is there a risk of a split between “AI curators” and “analog rebels”?

(PV) Another point will be how we classify the entire history of design. Today, it is often said that this history is less important (it is no longer taught in many schools), but conversely, it is precisely this knowledge that gives rise to what we call culture or, in our case, design culture. Here, too, AI, which has all these historical facts at its disposal, can make a significant contribution. Could we use this knowledge to improve our design culture with the help of AI?

(deepseek) Absolutely, design history is not a dusty archive, but rather a breeding ground for future innovation. AI can not only preserve this heritage but also turn it into an active tool for design culture. Here are a few examples:

AI as design memory – from chronicler to co-thinker

Context instead of copy-paste: AI that analyzes the Werkbund movement can not only replicate forms but also extract principles (“Less is more” as a response to industrial overload) and apply them to today's problems (e.g., digital overload). A tool that recognizes that Dieter Rams' reductionism could cure today's UI/UX disorientation and suggests adaptive interfaces. Forgotten paths could also be revived. AI searches archives for failed experiments (e.g., 1970s psychedelic typography) that suddenly become relevant (e.g., for AR filters).

Culture as code, or how AI decodes “style”

From aesthetics to ethos. Instead of merely copying visual characteristics, AI is learning to read cultural codes. Why did Oliviero Toscani choose provocative photos for Benetton? To turn advertising into activism. Today's AI could deduce from this that brands need to take a stand and propose radical campaigns. What would-be scenarios in which one imagines how Italian futurist Benedetta Cappa would have designed today's social media branding? AI simulates answers that inspire designers.

The danger of cultural “fast food” through AI

Superficial citation: AI could mix Art Deco with cyberpunk without understanding that both styles reflect social fractures. The risk is that design becomes a costume party. Countermeasure: AI must learn meta-levels. Not just how something looked, but why it looked that way. Colonialism in algorithms? If AI only knows Western design history, it perpetuates dominance. One solution would be a kind of decolonization of design databases (e.g., African textile patterns, Japanese wabi-sabi aesthetics, or Chinese perspective).

Better design culture through AI? Yes, but...

Its strengths lie in the democratization of knowledge. A startup can suddenly access Ulm School principles. Intercultural hybridization occurs when AI combines Nordic minimalism with Indian color symbolism, but only if it understands the respective depth. Limits arise because culture is born out of conflict. The punk style rebelled against existing rules; AI would have classified it as bad design. The unspeakable: Swiss Style represents an attitude (order, stability, etc.). The question arises as to whether AI can capture the feel of an era.

The new role of designers

You will become cultural interpreters who train AI to recognize subversive trends (e.g., street art as counterculture). Breaking algorithmic suggestions (This AI-recommended harmonious palette is too bland; let's go with this disturbing neon accent instead). Retelling the past: AI shows that pop art principles can be applied to a biotech brand – but humans decide whether this is ethically and commercially appropriate.

Provocative thesis

The best AI for design history would not be a library, but a debating partner that forces us to engage critically with tradition. Could AI help us understand design culture more deeply, or would it turn it into a soulless museum?

(PV) We started with the question of what the future of identity and branding might look like. We chose the AI chatbot DeepSeek R1 as our conversation partner. In this respect, our project is a double experiment: first, we are attempting to define a very uncertain future, and second, we are using an artificial conversation partner. Whatever one may think of this research design, I am convinced that this discussion has identified a whole range of potential questions and relevant topics. My aim was not necessarily to make absolute predictions, but rather to raise important issues that we should be discussing.

In the fourth part of this series of articles, I will attempt to find out how different chatbots respond to the same questions and what we can learn from this.

This discussion arose in dialogue with DeepSeek Chat, https://www.deepseek.com (May 2025 version) on May 29 and 30, 2025 – as a thought experiment on the future of identity and branding.

Note: DeepSeek's statements have not been edited in terms of language or content. The statements made here correspond entirely to the AI version of this discussion.