The first thing that caught my attention in Побег was not what happens inside the session. It was the way the product positions itself before anything starts.

On paper, the ingredients are familiar enough: AI tools, visual exploration, creative exercises, a final artifact. You could easily describe it as a workshop on using AI to develop a visual concept and turn it into a small campaign. That description would be accurate. It would also make the product almost instantly comparable with dozens of other offers.

Verlab does something else. They call it Побег — Escape. And that small decision changes the entry point completely.

Instead of beginning with a competence gap — you don’t know this yet, so come and learn it — the product begins with a state :

That made me look at the product differently. Because maybe AI is not really the problem Побег is solving. Maybe AI is simply the material. The actual problem seems closer to this: How do you enter a new technology without turning it into yet another optimisation project?

That is a very different design brief. And once the problem changes, the format that answers it can change too. If the problem were simply I need to learn how to generate better images with AI, then a course or a technical workshop might be the most logical response. But if the problem is I want a reason to explore this territory without feeling that I am studying it, then a course suddenly starts to feel too heavy.

This is where this product becomes worth getting interested in from a design perspective. The format is not just the container for the content. It is part of the answer.

Think about one of your own products for a moment. Strip away the current format.
No course. No workshop. No consultation. No masterclass.

Not just what people need to know. What is happening in their life or work when they start looking for your expertise? Those two questions can lead to very different products.

And this is where Побег becomes more than an unusual name. The concept of “escape” creates a small world around the offer. You escape from something. You escape into something else. An escape is temporary. It gives permission. It suggests movement, curiosity, distance from routine. It does not promise mastery. It does not even require a productive justification.

That concept then quietly shapes what feels natural inside the experience. A relaxed rhythm makes sense. Experimentation makes sense. A creative route makes sense. Doing something beautiful without needing to explain why makes sense.

The name is not sitting on top of the product as decoration. It is helping define what the product is allowed to be. I see here an important distinction.

We often treat naming as the last step in product development: first build the offer, then find something catchy to call it. But sometimes the concept comes earlier. Sometimes it gives the product its logic.

Imagine the same activities under a different name:
AI Visual Campaign Intensive - The moment you read it, performance enters the room.
Generative Art Masterclass - Now expertise and instruction move forward.
AI Jewelry Workshop - The offer becomes clearer, perhaps, but also narrower.

Побег creates another expectation entirely. You are not only coming to learn something. You are entering a temporary situation that changes how the rest of the product is perceived.

The first edition is built around jewelry. Participants move from mood, image and material toward a visual language, then use AI to develop an object and eventually a small campaign.

At first I wondered how important jewelry really was to the concept. The more I thought about it, the more useful it seemed precisely because it is specific.

“Experiment with AI” is almost too open. Jewelry gives the exploration a territory. There is an object. A body. Material. Scale. Styling. Photography. Meaning. Brand language.
Enough structure to create friction. Enough freedom to produce different results.

This is something we sometimes forget when talking about experimentation. We tend to associate experiment with freedom. But complete freedom is not always very productive.
Sometimes we need something to push against.

Jewelry does that here. It gives curiosity somewhere to land. And this is probably the first real format shift in the product.

The knowledge itself has not disappeared. Participants still acquire skills. They work with prompting, visual consistency, style, narrative, and image generation. But learning is no longer standing in the center of the room. It has been moved slightly aside. People gather material. They make. They try. They experiment at their own rhythm. And somewhere in that process, they become more capable.

That is quite different from the usual logic of an educational product.

Normally, the product begins with the question, "What should people learn?"
Here, another question seems to come first: What kind of situation would make people want to explore?

And maybe that is why Побег feels different before the session even begins, as the ingredients are not radically new. They have been organized around a different idea.

That is what makes the case worth taking apart.

Побег is useful precisely because the underlying territory is familiar. AI tools, visual creation, guided practice, and a final artifact already exist everywhere. What changes is the relationship between them and therefore the experience of the product.

Once I stopped looking at Побег as “an AI workshop with a creative theme,” another question appeared: what exactly has been designed here?

The obvious answer would be the session itself as the sequence of activities, the tools, the exercises, the final campaign. But that is only one layer. What makes the product stand out is that several design decisions are working together at the same time. There is a concept, a format, a process, an offer, an outcome, and potentially even the beginning of a repeatable product model.

That matters because intellectual products are often reduced to content plus delivery. We have expertise, we organise it into modules, choose a format, add exercises, and call it a product. But Побег suggests a different way of looking at the object: perhaps the product is not simply the knowledge inside the container. Perhaps the container changes the value of the knowledge itself.

Take the concept first. Escape gives the whole experience a reason to exist beyond skill acquisition. It's a strong meaning. It explains why the tone is lighter, why experimentation is legitimate, why a Saturday session makes sense, and why producing something “just because” is not treated as a waste of time. The concept is not a slogan added after the design. It helps organize the design.

Then comes the format. A workshop normally carries certain expectations: someone knows more, someone else comes to learn, activities move toward an intended learning outcome. Побег loosens that contract. There is still guidance, but the participant enters less as a student and more as someone joining a temporary creative situation. That sounds like a small semantic change, but it shifts the relationship with expertise. The expert is no longer only transferring knowledge; they are designing conditions in which something can happen.

The process changes with it. Instead of making learning visible at every step, the product foregrounds gathering, making, trying and composing. Skills are still being acquired, but they are embedded inside the activity. You may learn how to maintain visual consistency because you need it for the object you are creating, not because “visual consistency” has appeared as module three on a curriculum.

This is where the format starts to affect the experience of learning itself. The participant is not constantly reminded that they are being taught. They are occupied with the thing in front of them. The learning is still designed, but it is partially hidden inside the doing.

That changes the offer too. If you sell an “AI visual creation workshop,” the value is easy to compare. How many hours? Which tools? What level? Which techniques? What will I be able to do afterwards? These are perfectly legitimate questions, but they place the offer inside an existing comparison set.

Побег introduces another type of value. You are buying access to a designed situation: a reason to disconnect from routine, enter a particular creative territory, experiment with a set of tools, and leave with something that did not exist before.

The final artifact matters here more than it may seem. A creative campaign is evidence that something happened. It gives the session a material afterlife. You do not leave only with notes or a list of techniques; you leave with an object that carries traces of the decisions, experiments and discoveries made along the way.

This is an interesting property of many intellectual products. The more abstract the expertise, the more useful it can be to ask what remains after the experience. Sometimes it is a framework, sometimes a prototype, a map, a decision, a story, a visual system, or a new hypothesis. The artifact does not have to be the goal, but it can make otherwise invisible intellectual work tangible.

At this point, I started seeing Побег less as one format and more as a stack of design layers.

There is the concept that gives the experience meaning. There is the format that defines the relationship between participant and expertise. There is the process that structures what people actually do. There is the offer that translates all of this into something people can recognize and choose. And there is the artifact that makes the experience visible afterwards.

Let's be honest, none of these layers is particularly radical on its own. What is worth noticing is the way they reinforce one another. This is also why two products can teach almost the same thing and still feel fundamentally different.

Imagine two experts with similar knowledge of AI image generation. One creates a four-module course on prompting, style, consistency and campaign development. The other creates a temporary fictional world in which participants use those same capabilities to answer a strange creative brief. The knowledge may overlap considerably. The product does not.

And that leads to a design question more useful than “How do I package my expertise? ”.

Do they come to listen? To practise? To make? To investigate? To rehearse? To solve? To experiment? To compare? To discover?

The answer changes more than the format. It changes what expertise becomes once it is turned into a product.

There is one more layer hiding inside Побег, though, and it is the one that makes the case slightly uncomfortable in a productive way.

If the product presents itself as an experiment, where exactly is the experiment?

Is it mainly in the language and atmosphere? Is it in the freedom participants have inside a guided process? Or does the architecture itself contain a genuine unknown?

That distinction matters, because “experiment” is a very powerful promise. And powerful promises deserve to be taken apart too.

The word experiment is doing a lot of work here.

It sounds open. Low-stakes. Curious. Slightly rebellious. You are not enrolling in a programme or committing to mastery. You are trying something. If it works, interesting. If it doesn’t, that may be interesting too.

That is especially powerful around AI, where the dominant emotional mix is strange: curiosity, fatigue, excitement, pressure, fear of being late, fear of becoming irrelevant, and the vague feeling that everyone else has already figured out what to do with the tools. “Come learn AI” can easily sound like one more thing you should already know.

“Come experiment” sounds different. It gives permission. And from a marketing perspective, that is a strong move. It reduces the psychological cost of entry. It replaces performance with curiosity. It makes imperfection acceptable. It also attracts people who may actively resist educational language even when they are perfectly willing to learn. But that does not automatically make it a trick. A marketing promise becomes worth attention when it changes the design decisions that follow.

So I started asking a slightly embarrassing question: how much of the experiment in Побег is in the framing, and how much is actually inside the architecture?

There is a spectrum here.

At one end, experiment is mostly language. The route and expected result remain largely predetermined, but the framing makes participation feel lighter and more permissive.

Further along, experimentation enters the experience. Participants make personal choices - personal interpretation, different visual directions, unexpected combinations, and individual pacing - and arrive at different outcomes inside a guided route.

Then the experiment can enter the process itself: variables are changed, alternatives compared, consequences observed.

And at the furthest end, experimentation becomes a part of the architecture itself. Something meaningful remains genuinely unresolved. The designer creates the conditions but does not fully predetermine what the participant will discover.

That last version is the most experimental in the strict sense. It is also not necessarily the best. This is why I started questioning whether Побег is “really” an experiment.

Maximum openness can easily become confusion. A product with too much uncertainty may feel vague, exhausting, or pointless. A well-designed guided process can create far more meaningful exploration than an open field with no structure at all.

That depends on what the product is trying to do. If the goal is to help someone gain confidence with AI tools, a relatively guided structure may be exactly right. If the goal is to generate genuinely new creative directions, more uncertainty may be productive. If the goal is to challenge assumptions, the experiment might need to interfere with something the participant normally takes for granted. And if the goal is research, the unknown becomes central.

This is why I would resist treating an experiment as a binary category.

A product can use experimentation at different depths.

The design work lies in deciding where on that continuum the product needs to sit. Try applying this to something you already offer.

If you call it a lab, sprint, experiment, studio or playground, what does that word actually change?

• Does it only change the mood?
• Does it change what participants are allowed to do?
• Does it change how decisions are made?
• Does it change what counts as success?
• Does it introduce a real unknown?
• Or could you rename it tomorrow and leave the entire experience untouched?

That is usually the revealing test.

A concept becomes design when it starts affecting behavior. That explains why Побег is a useful case. The promise is strong because it does not merely say “this will be creative.” It creates permission to enter differently.

How far does that permission travel through the rest of the product?

That takes us directly into another field. Because if learning is happening through doing, observing, making and reflecting — only with the learning objective pushed out of the spotlight — then Побег starts to look familiar from another disciplinary angle. Learning designers have been working with that territory for a long time.

Experiential learning has been built around this logic for decades: people do something, encounter consequences, reflect on what happened, form new understanding, and try again.

So is Product as Experiment simply experiential learning with better branding? Not quite.
The overlap is real. The difference is less in the activity than in the design intention.

Experiential learning usually begins with:
What should someone learn through this experience?

Product as Experiment can begin with:
What could someone discover if we create the right conditions?

In the first case, experience serves learning. In the second, exploration leads and learning develops inside it. That small difference moves the center of gravity.

In a conventional learning design process, the destination is often known by the designer in advance. The participant may discover it through experience, but the designer already knows what the experience is meant to reveal.

An experiment leaves at least part of that destination open. This does not mean that learning disappears. In fact, one of the most compelling things about Побег is precisely that learning is still happening everywhere; it is simply not presented as the main event.

Participants gather references, work with materials, generate alternatives, make choices, notice inconsistencies, refine a visual language, and build something. To continue, they need to understand the tools better. They develop judgement because the work demands it. They acquire skill because the next move requires skill. Learning arrives through use. That creates a different psychological relationship with competence. Instead of learning first and hoping to apply later, participants encounter a need inside the activity and develop capability in response to it.

We could describe the difference very simply. In experiential learning, experience is designed to produce learning. In Product as Experiment, a situation is designed to produce exploration, and learning emerges inside it.

The two can overlap almost completely in practice. What changes is which logic leads the design.

And there is another distinction that matters. Experiential learning often still gives considerable control to the designer: sequence, objectives, reflection points, intended transfer.

An experimental product can distribute more of that control to the participant. People may move at different rhythms, pursue different signals, and make different choices and, therefore, acquire somewhat different capabilities. The designer is no longer controlling every learning movement. They are designing the conditions. Having and taking control is a subjective part belonging to a participant. The way they use this capability within a product affects the impact of the last one. That makes intellectual products sit aside.

Here we have the strongest interdisciplinary connections in the case. Product design asks what experience should exist. Learning design asks what enables development. Experimental practice asks what must remain open.

Put them together and the expert's role changes. So the intellectual product does too. The expertise travels and no longer lives only in explanations. It also lives in the environment, constraints, materials, questions, and conditions the expert knows how to design.

This also suggests another architecture worth examining separately: Product as simulation, where the designer creates conditions not primarily for discovery, but for rehearsal. I’ll leave that distinction for another product anatomy case.

For now, Побег takes us somewhere else — into brand strategy.

“When is the next Побег?” This is where the case moves from product design into brand idea.

If Побег had been positioned as an AI workshop, it would immediately enter a crowded comparison set. People would ask which tools are covered, how advanced the content is, how long the session lasts, whether the instructor is better than someone else, and whether they could learn the same thing from YouTube. That is what happens when a product stays inside an established category. The market already knows how to compare it.

Побег changes the frame. You are not necessarily choosing between one AI course and another. You are deciding whether you want to go on Побег.

That sounds like a small branding detail, but it changes the competitive logic. The product starts to become memorable as a whole world is built around the topic. This is where naming matters. And don't tell me it does not.

A good name can do more than make an offer recognizable. It can shape expectations, suggest behavior, and define the kind of experience people think they are entering.

Побег carries movement, temporary departure, freedom, curiosity. It also creates room for variation. One escape can lead somewhere completely different from the next one.

That matters because the topic no longer has to carry the whole identity of the offer. Today it may be jewelry. Tomorrow it could be shelter, tarot, objects, memory, fashion, sound, or something else entirely. The stable part is not the subject but the format concept. That is a powerful brand move because it allows the offer to become serial without becoming repetitive. The audience can recognize the logic even when the content changes.

This is close to how editorial products work. A magazine can change its issue theme without becoming a different magazine every month. A podcast can explore different questions without rebuilding its identity each time. A festival can change its programme while keeping the experience recognizable.

An intellectual product can work the same way. Instead of creating a new course every time the topic changes, you create a branded container with a stable promise, tone and structure, then let the territory evolve. That gives us a simple formula:

And this is where product architecture begins to support monetisation too. A repeatable format can generate editions, seasons, thematic variations and reasons to return. The audience does not have to rediscover the logic of the offer every time. They already know what kind of experience they are buying into. This does not mean that every format should become a brand.

You may already have the beginning of a format brand. And once that happens, monetization becomes a separate design question: how many editions can the format support, what creates repeat participation, what belongs to the core IP, and what changes from one version to another?

When does a repeatable product become an asset in its own right?
That deserves a separate piece on monetization and format IP. For now, one point is enough.

Побег does not simply avoid looking like another AI course. It creates a reason not to compare itself with one. Which is a much stronger position.

The next question is what happens if we push the experiment itself further not to “improve” Побег, but to see how many different products can emerge from the same starting territory.

At this point, it would be easy to turn the analysis into a critique: Побег could be more experimental here, less guided there, more open somewhere else. I don't find that useful.

A better move is to treat the existing format as design material and ask:

Not better versions. Different versions. This matters because experts often assume that one problem naturally leads to one format. If the audience wants to learn AI image generation, then the answer must be a course, workshop or masterclass.

But a problem does not contain its format. The format is one possible design response. So let us keep the broad territory — AI, visual creation, jewellery, curiosity, making — and change only the logic of experimentation.

Suppose the central provocation becomes "What happens when jewelry stops being wearable?"

The territory remains jewelry, but one familiar assumption is removed. Participants could begin with references to jewelry, architecture, sculpture, anatomy or ritual objects. They generate forms, compare what still reads as “jewelry,” and observe where the category begins to dissolve.

The process remains guided. There is still a route, a set of tools and a final artifact. But the goal is no longer simply to create a coherent visual campaign. The participant is testing a boundary. That changes what they pay attention to.

Instead of asking, Does this image look good?, they might begin asking, Why do I still recognize this as jewelry?" What exactly makes the category hold together? At what point does it become sculpture, costume, or object?

The skill acquisition is still there — prompting, visual consistency, composition, campaign building — but those skills are now serving an inquiry.

This would make the format more experimental without making it radically open. The designer still controls most of the conditions. The participant explores inside them.

Now imagine a different question, "Can jewelry communicate something that cannot easily be said?"

The process changes. Participants might begin with an emotion, tension, memory, or social signal. They develop several hypotheses about how material, form, scale or placement could express it. Then they create variants and test how other people interpret them. Suddenly observation becomes much more important. A beautiful artifact is no longer enough.

You need to know what happened when someone encountered it.
Did they read vulnerability where you intended protection?
Did an object designed around intimacy feel threatening?
Did two visually similar concepts produce very different interpretations?

Now the participant is not only creating. They are comparing. Testing. Revising. The artifact still exists, but it becomes evidence inside a larger process. And the expert's role changes too. They are no longer mainly guiding production. They are helping participants formulate better questions, construct useful comparisons, and notice what the results are actually telling them.

That is already a different intellectual product.

We could push the openness even further.

Instead of starting with a clear design problem, we might begin with: Where does jewelry end and something else begin?

There is no single destination. Participants move across territories: fashion, architecture, identity, ritual, technology, status, memory, body modification, sculpture, perhaps even interface design. They collect strange examples. Follow connections. Borrow rules from other fields. Generate speculative objects. Discard some. Keep others.

The outcome could be less polished than in the original format. But the intellectual territory could become much wider. One participant might arrive at wearable architecture. Another at jewelry as a memory device. Someone else is at an object that only exists digitally.

The value here is no longer mainly competence or even a finished campaign. It is territory expansion. You leave with a map of possibilities you could not see when you entered. This is much closer to an expedition than a workshop. And once we call it that, different design questions appear.

How much direction does an expedition need?
What counts as a useful discovery?
How do participants avoid getting lost?
What should be documented along the way?
What brings the exploration back into something usable?

These are not cosmetic differences between formats. They change the work being done.

The three alternatives begin with almost the same material.

AI. Jewelry. Visual creation. Making.

What changes is that they create three different relationships with uncertainty. The guided experiment keeps the unknown narrow. The research lab makes interpretation and evidence part of the process. The expedition opens the territory itself.

That is why I find format design more useful than the usual question of “how to package expertise.”

Packaging assumes the knowledge already determines the product. Design asks what kind of experience should exist around the knowledge. And once you start there, the same expertise can become very different things.

A course.
A lab.
A simulator.
A studio.
A game.
An expedition.
An experiment.

The important question is not which format sounds more original. It is the one that creates the most meaningful relationship between the problem, the participant and the expertise.

Try this with your own field. Take one problem you solve repeatedly. Now imagine that instead of teaching people how to solve it, you let them test one assumption around it. Then imagine they had to collect evidence. Then imagine the territory itself was open and they had to map it. Would you still be designing the same product?

Probably not. And this is where the case begins to produce something reusable.

That is the next step.

By this point, the word experiment has stretched across several layers. It can describe the mood of the offer, the way participants behave, the process they follow, or the degree of uncertainty built into the product itself.

That makes the term useful but also slippery.

I am less interested in defining what an experiment “is” than in identifying what needs to be designed for one to become possible.

I use six zones to make the design architecture visible.

Put these zones together and the architecture begins to look less like a sequence of steps and more like a set of design decisions:

Provocation / Question
What makes the exploration worth entering?

Territory / Unknown
Where are we experimenting, and what remains unresolved?

Boundaries / Constraints
What limits make the exploration productive?

Materials / Tools
What can participants manipulate?

Intervention
What will they do to make something happen?

Observation / Discovery
What should they notice, and what might change as a result?

Artifact
What emerges and remains?

The canvas is not a recipe. Try applying it to one area of your expertise. If the answers come naturally, you may have material for an experiment. If they do not, forcing the format is probably the wrong move. It may simply mean that another architecture would serve the purpose better.

The goal is not to make expertise experimental. The goal is to find out whether experimentation creates a better way to experience it.

By now, the obvious temptation is to turn this into a workable framework.

Take your expertise. Add a provocative question. Introduce uncertainty. Give people some materials. Let them manipulate something. Finish with an artifact. Done. Experiment completed.

So, could your expertise become an experiment? Possibly. But I would not start by asking how to turn your existing course or workshop into one. I would look for the places where explanation is not enough.

That question changes the starting point. You are no longer looking at your knowledge as content to transmit. You are looking for places where something can be explored, manipulated, compared, or discovered.

• Maybe there is a problem in your field that has several plausible answers.
• Maybe people understand something much faster when they experience the consequences rather than hear the explanation.
• Maybe one of the strongest assumptions in your industry deserves to be disturbed.
• Maybe comparison produces more insight than instruction.
• Maybe people need to develop judgement, not simply acquire information.

Those are promising territories.

Take brand strategy, for example. You could teach positioning frameworks. Or you could create a controlled situation in which participants remove every familiar category cue from a brand and observe what still makes it recognizable. The framework can come later. The experiment creates the need to understand it.

A finance expert could explain risk perception. Or participants could make the same decision under three different information conditions and compare how their judgement changes.

A leadership consultant could teach delegation principles. Or temporarily remove hierarchy from a decision scenario and let a team observe what happens to responsibility, speed and conflict.

A learning designer could explain cognitive overload. Or create two deliberately different versions of the same learning experience and let people feel the difference before naming what caused it.

The expertise has not disappeared in any of these cases. It has moved. Instead of sitting entirely in the explanation, part of it sits in the conditions of the experience.

And then the learning question returns again. In the Побег model learning is put slightly aside. Not removed. Not devalued. Just moved out of the foreground. Participants gather material, test possibilities, make something, encounter problems, adjust, continue at their own rhythm. They need new skills because the activity starts demanding them. So they acquire them.

This is almost the inverse of the familiar sequence: learn first, apply later.

Here it can become: do, notice a need, acquire capability, continue.

That can be especially effective when the competence itself is not very attractive as a starting promise. Few people wake up excited to “develop better visual consistency across generative outputs.” But they may be very willing to make a strange jewelry campaign and discover halfway through that consistency suddenly matters.

The learning becomes useful before it becomes explicit. That distinction is worth thinking about when you design intellectual products.

• What if people did not have to be persuaded to learn the skill first?
• What if the product created a situation in which the skill became necessary, desirable or obvious?

This is where experimentation can become more than an unusual format. It can change the timing of learning. It can affect the motivation for learning. And it can change the role of the expert from someone who mainly delivers knowledge to someone who also designs the conditions in which knowledge becomes usable.

There is a practical test you can run on your own work. Take one subject you regularly teach, explain, or advise on.

Then ask:

• What part of this do people only really understand once they have tried it?
• Where is there room for more than one valid outcome?
• What could participants manipulate rather than merely discuss?
• What assumption could be changed temporarily?
• What could they compare?
• What would be useful for them to notice?
• And what skill might they acquire almost incidentally because the activity requires it?

If answers come easily, you may have material for an experiment. If they do not, that tells you something too.

• Maybe the right architecture is a simulator.
• Maybe it is a studio.
• Maybe it is a guided practice.
• Maybe it is a game.
• Maybe it is simply a very good explanation.

Starting with format is not the most effective way. I would start questioning the relationship you want people to have with the expertise.

Do they need to understand it, practise it, question it, test it, make something with it, or discover something through it?

The architecture comes after.