Read my articles about technology, business, and leadership

enjoy my pictures as well

This isn't about the innovators creating groundbreaking technologies, but about 99% of businesses worldwide that leverage various products to offer their services or to create other products. Consider two bakeries on the same street in a city of one million residents. One displays the words "In Business Since 1890" on its signboard. Their competitive edge over the newly established bakery is experience. They have gathered customer feedback over the years and continuously improved their offerings. Even if the two bakeries use identical ovens, products, and recipes, the experience becomes the competitive advantage. Consequently, entrepreneurs often "acquire experienced employees." This approach has been the norm for centuries. In the fourth industrial revolution, however, the product contains both knowledge and experience (of hundreds of millions of people). We have seen the fusion of these two elements in various "smart products." For instance, a smartphone camera processes pictures according to the most suitable format and color for social media platforms to maximize likes or engagement. This is the "Product of the Future."The question then arises: Where is the competitive advantage between two photographers using the same tool - the Product of the Future? Is it price, network, or marketing? 

Today, these factors are more crucial than experience or knowledge - these can be purchased, bluntly speaking. Lately, there's been much buzz about Generative AI. In my view, it's a breakthrough, not necessarily in technology or discovery, but in daring to create the Product of the Future. Andrew Lowe and I recently attended a lecture at MIT where Danny Lange compared ChatGPT to the 1957 launch of Sputnik by the USSR. This seemingly small "wheel" with four "sticks" made the world realize that someone (in this case, the USSR) was gaining dominion over space. Did Sputnik have stunning functionalities and capabilities? No, but that wasn't the point. It was about demonstrating dominance and signaling that satellites are the technology of the future, which indeed they are. The situation with Generative AI is similar. The fusion of knowledge and experience in one product makes this technology revolutionary. Its general availability, ease of use, and operation speed make it extremely useful. So, what will be the competitive advantage in the era of Generative AI? What will distinguish two text writers using GPT chat? Of course, a third writer who does not use it won't stand a chance from the start. The answer, though seemingly simple and obvious, might surprise and shock many, especially those clinging to their business models. The answer lies in usage, the business model, and the operating model. The SPEED to adapt and the conscious courage to do so will be pivotal. Those who remain stuck in "yesterday's world" will quickly lose margins, customers, and existing competitive advantages to new, agile businesses based on APIs and applications. The competitive advantage will arise from the ability to "unlearn" habits and understand new customer needs and behaviors. It will be directly proportional to the implementation of scalable changes in business models and the utilization of the human potential for tasks that Generative AI cannot yet perform.

There are many novelties (technological) around us, but only a small part of them stand a chance to be remembered and used. Only some technologies have characteristics that permanently introduce them into our lives. Most technologies, especially modern ones, are very difficult to understand. Even though they look nice, our habits don't allow us to use them. People want to own, and if not own, then to feel that they can have access at any time. This is how the revolution in music or movie carriers took place. We have streaming, we no longer buy CDs or DVDs because it's not a problem to subscribe to dozens of platforms and have access. But if we want something special, we buy a limited vinyl, not to listen to. But to own. People are naturally lazy. If they can do something more simply, they usually do. Even better if the "entry cost" is minimal. If something is free and doesn't require much knowledge to use, it's impossible to resist. Even if over time you pay a price. Intelligent car maps or concierge-like platforms recommend a price or specific product. Who could resist? And finally, people need security. Not just in a physical sense. It's very important to be aware that if anything happens, I have someone/something to ask for advice. That's one of the most important reasons why we use trusted mail servers, GPS maps, or translators on our smartphones. It's at hand, always, and it works. But how does this relate to the pony? This is where the mystery of man lies. Humans above all need contact and a sense of interaction. They need a sense of dominance and appreciation. This is what a pony, dog, or any intelligent life form provides. If we combine these characteristics, Generative AI has them all. 

Starting with the trust that it's always at hand, through the sense of owning personalized results, and ending with natural contact, and natural language. Of course, the era of Web3 is not over, and I believe there are many wonderful possibilities for its use. Perhaps META itself, which tech giants are slowly withdrawing from, is debatable, but NFTs or DAOs will return. And soon. But at the moment, "everyone wants Generative AI." And rightly so. There are concerns about ethics, development, dominance, and job loss. And everything is understandable and right, but the temptation of possession is greater, and the effects of the operation are OUTSTANDING."

The answer is simple - NO. The market needed about 5 years to realize what was happening and adapt to the new realities. Internal sales departments gradually moved to new business models related to e-commerce, after recognizing the potential of social network marketing, the new term Digital Marketing became the basis for even very traditional businesses. What's happening now? Since 2015, trends related to Big Data, IoT, AI, Cloud, and Mobility have been appearing in the market of new technologies. The "paranoids" were massively testing new technologies and adopting new business models. COVID-19 accelerated the adoption of cloud and mobile solutions. You could show exponential sales growth through e-commerce channels (particularly thanks to mobile payments and quick delivery of goods). Unknown customer offers appeared, gaining market share, such as choosing 5 products, leaving what fits, and returning the rest at our expense. Drop shipping dominated "small" trade, and in many cases, social media portals became the only access channel to the customer even for large players. What's ahead of us? I think most companies and their leaders realize how pivotal a moment is coming or has already come. I'm talking about Generative AI, of course. And to be clear, these are "paranoids" in the positive sense of the metaphor. They try to identify work and processes performed by dozens of employees and intelligently automate them. And while in the era of software Robotic Process Automation, we talked mainly about back office work, now we are talking about creative work, customer contact, marketing content, and substantive development. Of course, in this case, the work done by people will be replaced by Generative AI, and the control over its operation will be the same people with other skills. But the game is about something else. The stakes are new business models and changing customer habits. The offer of completely new services and smarter products. Here is the chance for the "Holy Grail". What is it? In my opinion, it is the real era of Data Monetization using AI (generative AI). It is creating or modifying business models based on fully automated and intelligent decision-making processes. Everyone probably knows the saying that a frog boils in water because it doesn't recognize the temperature increase. It needs an impulse from the outside to save itself. Someone must "give it a hand". 

The process described in the book "The Innovator's Dilemma" by Harvard Business School professor and legend Clayton M Christensen, is when innovation leaders make changes to transform from a caterpillar into a beautiful butterfly. Over the past five years, the hot topic of Digital Transformation has lost its impact. But now it's back. Maybe we will call it something different. But right now, with the trend of generative AI, "frogs will jump out of the boiling pot". Right now, companies will believe that it's worth making this transformation and really change. Right now their customers expect it because they themselves are already using simple Generative AI tools and understand their benefits. They may not fully understand their drawbacks and consequences, but in the world, it is already the case that "bad money drives out good", or to paraphrase, worse quality AI results replace better human work. But that's a compromise humanity is ready for. Generative AI will be this helping hand for entrepreneurs. Thanks to it, they will improve their services and products, they will copy others, and optimize the use of their own knowledge resting in the data they collect.

1. The first is the discovery of a secret, magical knowledge, hidden in data. We have plenty of it, and it is rapidly increasing. For over 10 years, people have been talking about Big Data and its potential for increasing productivity or generating new revenue. And it's true. Only a few companies can convert data into information, and then into knowledge that gives a competitive advantage or simply generates income. These magical transformations using artificial intelligence (both advanced analytics and machine learning) are becoming the basis of the growing trend called data monetization. I spoke about this potential a few years ago and described it in my book "SPEED no limits in the digital era."
2. The second thread is introducing a new business model in place of the old one or fundamentally modifying it. Examples of companies transforming from failing or at least stagnating businesses into sector sharks are multiplying. They have crushed the traditional way of doing business and introduced disruptive changes to business models - they no longer sell tapes, they sell access to platforms. Instead of focusing on the product, they focus on the customer and their needs.

And here is where the knowledge from the tales of the 10th century comes to our aid. Indeed, we speak of success when we simultaneously implement technology (in our case, data monetization) and change the business model. Aladdin finds the lamp alone does nothing for him. Only when he rubs it and summons the genie who fulfills his wishes does it change his situation and bring him wealth and happiness. I meet many clients who ask for the implementation of modern technologies. They ask how to monetize data. However, few of them, if any, see the full picture and, most importantly, are capable of changing the model of operation or revolutionizing it. One might think it obvious that when talking about transformation, we mean the introduction of technological changes and business changes. But that's not the case. Many transformations fail because there's a lack of changes in the organization's culture (or these changes were not considered when creating the strategy), the difficulty of changing the values upon which the organization is based was underestimated - and I'm not talking about ethical values here, but business ones. After all, who would want to change a model where one sold their product for a million dollars once, and now we encourage to earn one dollar from two million transactions? We must remember that just like in the „One Thousand and One Nights” tales, the "enemy" changes before a transforming organization. Over all these years, we got to know him well and knew how to control and fight him. And now a super character, cunning and unknown, appears. It takes on various forms, is cunning and has powerful abilities. This is like competition in these dynamic times. The customer is not loyal, the regulator can be surprised, and in the end, a crisis and inflation appear. Scheherazade didn't finish her stories, causing her husband to be curious about what would happen the next night. Did you know that one of the strongest mechanisms of successful data monetization and new business models in the digital world is dynamic asymmetry of information? But more on that in the next story.Good night.

Companies and their clients are creating more and more data. It turns out that for some time now, more and more specialized institutions are emerging that follow the path of collectors. These companies (most of them coming from the telecommunications market) aggregate and sell data. They saw potential in the fact that having good-quality data and a complete collection constitutes value. Exactly as in the case of collecting postage stamps. The price for providing this data may be standardized (e.g., a credit report, or weather forecast) or depends on the specifics of the aggregate that must be prepared and the data set that must be used to prepare such a query. As for the idea of this type of marketplace (data marketplace), they are nothing new and also operate on the principle of "willingness to pay" or a minimum price. And we would probably end this story about the possibility of data monetization and constantly emerging data providers and marketplaces here. In the not-too-distant future, there will probably be many more of them because generative AI becomes a wonderful tool for converting data into information and knowledge. Companies, having their own operational data and collecting it, see an easy opportunity to earn on them simply by changing their form. It's like selling eggs and flour separately. By putting in a bit of work and knowing the recipe, you can turn this into a cake and sell it for a much higher price. And most importantly, in a repeatable way, as a new business model. 

So what is another possibility? And here again, a certain business model known to us all comes to mind. We remember how in the '80s or '90s in front of stadiums where concerts or matches were held, there were ticket sellers, offering them for several times their nominal price. There are probably still such places. Nevertheless, this business has almost completely been replaced by websites where you can buy tickets for all events in a given city. Sophisticated revenue management models optimize the price depending on demand and availability, and high service commissions make it a great business. What makes it profitable? Namely, its nature – real-time mode. The platform owner must have information about demand, supply, events, and weather predictions, for example, to optimally manage the price (and sometimes availability or apparent unavailability). This platform model is a wonderful example of data monetization. But what phenomenon is key to success here? This mechanism is the dynamic asymmetry of information. The phenomenon of information asymmetry has been known in economics for a long time. It can have negative or positive connotations. However, in the case of digital business models, it is the basic "engine" for making money. It is easy to check whether a startup will be doomed to failure or has a chance of success by only checking whether and how many levers of dynamic asymmetry of information it possesses. Thanks to the mechanism of dynamic asymmetry of information, the platform owner makes decisions about connecting the buyer with the seller or launching specific actions. An example would be the order of display of products on a sales platform depending on how much the seller paid for the "position". Sound familiar? But the real magic is on the other side, the advertisement seller knows exactly how much to ask for such a position. So de facto knows dynamically how much that information is worth at any given moment. What's more, all of this happens automatically and is admittedly controlled by very basic, but still artificial intelligence. 

Dynamic asymmetry of information based on artificial intelligence and operating very quickly (basically imperceptibly to human decision speed) poses a very large threat in the ethical, social, and economic scope. Improperly or unethically used, it manipulates the customer, supplier, and sometimes also the platform owner. But that's a topic for another article.

Industrial Technologies have one characteristic: their emergence (usually growing gradually and leading companies from a particular sector create their standards and implementations) changes and revolutionizes an industry. Carr uses telegraph lines, railways, or electricity from the Second Industrial Revolution as an example. He concludes this part of the article by stating that IT is just another technology to use, but you cannot build a competitive advantage on it, or build it for a very short time. The same is true today with AI (generative). Simple, easy to use with fantastic capabilities. And just as in the 19th and early 20th centuries, all newspapers were raving about the beneficial railway transport, which allowed for the transfer of finished products from factories to the other end of the country or universal electricity. Today in every piece of news everyone is talking about how wonderful or dangerous AI is. How it will inevitably change industry A or B. And this is true. Advisors around the world are receiving RFPs from clients wanting to learn how AI might transform their companies or alter the way they provide services or products. At this point, Carr surprises. He says that after a while, when the technology is popular enough to be a commodity (as was the case with IT and will be the case with AI – the democratization of technology), creating a strategy on how to use electricity or transport to build a competitive advantage simply doesn't make sense because it is universally available. Today, who contemplates or devises a strategy on how to use a phone, railway, or electricity for their company? Today, strategies, plans, and analyses for the use and impact of AI on a business are created not due to their rational need to exist, but due to a lack of knowledge, experience, and understanding of this technology, or maybe even because of fear of the inevitable change. And this is understandable. However, what is important is contained in the following sentence quoting Carr: „When a resource becomes essential to competition but inconsequential to strategy, the risks it creates become more important than the advantages it provides” Carr was thinking about the risk associated with the lack of access to technology, which, adopted in a company for reasons of limitations or catastrophic events, is suddenly unavailable. The example of a lack of electricity or maybe in our time's access to the internet can completely paralyze a company. And it happens that we have experienced numerous examples of this during the COVID-19 pandemic or numerous outbreaks of hacker attacks on IT systems. But in the case of AI, it's possible to take a step further in perspective. Do we remember what happened to companies like Kodak or Xerox when digital technologies literally swept them off the market? Do we remember how companies that built their product on trust in technology lost customer trust because they couldn't deliver services or their service was compromised (RIM)? Do we finally remember how telecommunications companies, the owners of the infrastructure, lost their revenue to application content providers (Meta, Alphabet, or TikTok) which cannot function without them? Why did this happen and why is it happening? Because the level of asymmetry to knowledge, information and customer access is changing. In the previous few articles and in my book "SPEED no limit in the digital era" (2), I described this phenomenon as dynamic information asymmetry. It is through it that margins are built, on which companies earn by creating new business models and monetizing data. 

However this phenomenon also has a potential negative impact on companies, and this is due to AI. The democratization of technology, i.e. making every market participant, at a marginal cost, able to use it and build new services and products based on it, gives in the case of AI a powerful weapon of having the knowledge and experience of millions of users at disposal. Such a „product of the future”, as I described in "SPEED no limits", is easy to use and very cheap (today an example may be ChatGPT), taking away the competitive advantage of existing market players. Knowledge-based companies that use it to generate revenue will have to deal with the disruption brought by AI, but here's a caution. Their competitors who will appear on the market in large numbers will be fully dependent on a single technology or even one provider. Companies that easily adapt AI solutions to increase productivity or reduce costs also see the risk of a single point of failure. As long as AI does not make autonomous decisions or is not fully transparent for application control, the risk will be managed, but when we decide to use solutions for basic business activities, then N.G. Carr's words gain significance. Due to AI tools, the difference between companies, or strictly speaking their services and products, will become blurred. Today, it is already difficult to distinguish the origin of a digital service. It is impossible to determine what technology stack a product is based on. In most cases, it is the available human resources (their knowledge), regulations, and the unit price of technology that determines their choice. Mass automation of processes and the implementation of self-service models increase companies' dependence on technology, which, as Carr notes, does not constitute a competitive advantage. And since "AI Doesn't Matter" because it will be every day and will be universally available with the knowledge and experience of millions, it will not provide a competitive advantage either, I suggest thinking today about the risk for a company of not transforming in time and improper or faulty AI operation after the transformation. How to maintain the TRUST of your clients and partners and create a security patent.

The first and second industrial revolutions brought a different kind of fear. Machines stronger than man and the strongest animals, forces of physics and chemistry such as electricity or radiation, caused fear. But every time we managed to tame them. Of course, to some extent, we used various kinds of safeguards, mostly consisting of procedural and mechanical separation of this "bad force".In my book "SPEED no limits in the digital era", I introduced the concept of a "Security Patent". I defined it by comparing it to a fuse in an electrical installation or a pressure valve in installations with compressed gas or water. For a product to be widely used and for people to trust it, there must be a fully autonomous element preventing accidents. So, during an electrical surge, the fuse protects us from shock, or in the case of a sudden pressure jump in the installation, the safety valve protects against explosion. Companies have taken responsibility for the health and life of users themselves, guaranteeing them safe use of products. What's more, if the product was used as intended, the "Security Patent" did its job flawlessly. We gained trust which over the years became the competitive advantage of brands. A reliable product from a reputable supplier. And even today, we prefer to pay more if the product is backed by a brand we trust. In the fourth industrial revolution in which we live, security is changing in every way. The mistake that was made at the beginning of the computerization era resulted from a lack of imagination or faith that the digital world would become ubiquitous. And with this omnipresence came the disease of cybersecurity, or the lack of ability to guarantee that electronic services are safe. Advanced cryptographic data protection methods and logical and application security mechanisms do not provide the certainty that a fuse, a piece of wire in a ceramic housing, does. The danger to business and general users from the perspective of artificial intelligence can be simplified into two dimensions. One is the direct use of AI, and the other is the threat arising from the unavailability of AI. 

For the former, there is a lot of discussion about the consequences of introducing AI, specifically replacing human work and skills with AI. And this is the natural course of things. One retired professor from Cambridge told me an anecdote about how easy it was to identify mathematicians, and physicists on the university campus in the 50s and 60s. They all had torn pockets or jackets worn out by slide rules. This tool was essential for scientific life. Then they replaced them with calculators, and the ability to use a slide rule was forgotten. Of course, the new tool, the calculator, was based on known algebraic laws and, as a rule, did not make mistakes in calculations. Today, the real threat of AI is its fallibility or rather the lack of an idea for testing the correctness of its operation. Especially when it comes to self-learning. AI creators admit that they do not know how it arrives at its conclusions, which is an inherent feature of these algorithms. Another problem is deliberately induced or accidental bias. So, learning on flawed premises from the start. False or biased input data give false or biased results. In a situation where the user does not have the knowledge to verify the result or simply blindly believes in it, it constitutes a serious risk. The "Security Patent" here will trust in the AI service provider, as is the case with other technologies. This will be a cost that the supplier incurs for verifying data sources and verifying results. This cost will be distributed to users. The second and, in my opinion, much more serious threat is the fact of AI unavailability. Not much is said about this, which is a shame. What do I mean by this? Do you remember what the global costs were for preparing for Y2K? It is estimated that it was about 500 billion USD. They were only concerned with the trivial matter of making sure that after the change of the date to the 21st century, computers would still work correctly and make correct calculations. 

The world's dependence on computers at the beginning of our century was much less than it is now. But this comparison illustrates very well the threat I am talking about. What is our "Security Patent" in a situation where AI is not available or we find that it has systemic gaps, i.e., it makes mistakes? In classic computer ecosystems, there are business continuity plans in case of the unavailability of IT systems or ICT technology in general. Basically, they come down to the fact that we go back in time in operations to conventional methods, and instead of self-service over the phone, counters are open with service by institution employees. These anachronistic examples show the way of thinking. And in the transitional period, when there is knowledge and resources from "both worlds", they are imaginable and feasible. A mathematician in the 1980s who ran out of batteries for a calculator was able to dig out a slide rule in a drawer and finish calculations. Even in Y2K plans, the assumption was that if the telecommunication network failed, data would be downloaded onto floppy disks and driven to the other side of the city by car. With today's state of technology, this is already impossible. Firstly, there are no floppy disks, and secondly, the amount of data is too large. Of course, I am satirically showcasing the problem, but just as we are essentially helpless against cyber attacks and lack a security patent, we are entering the era of easy, cheap, and convenient AI which will replace everything and everyone without such security. I will wager that around the world, individual countries will soon make clumsy attempts at AI legislation and regulation. This will give a sense of control but will not solve the problem. Maybe we are doomed to accept this risk, close our eyes, and believe that someone will find a solution. Perhaps this is another one of the civilization diseases that will create a billion-dollar market of "AI auditors". We cannot stop progress. And we don't want to stop it. But let's not make the same mistakes in terms of security and privacy that the internet brought.

Very often, with the introduction of new technology, we cannibalize the habits of our customers and stop making money on what we used to earn. The conclusion is that instead of fearing this, we should consider how to make money in a new way. There are countless examples of switching to self-service instead of approaching staff or introducing new services like storage or posts at gas stations or residential buildings. Looking through the prism of what we know and use and not through the prism of the needs of new masses of customers can lead us not only to lose the market but also lose the position of leader in this market. Often after a few years, it turns out that the missed opportunity is used by others who were not at all brave, they were open to change and expectations of new generations of customers. New technologies such as generative AI are entering our everyday life. They are already coming in. But their real power and how they change our world will be noticeable in five or ten years. Jim Rohn aptly put it, "Usually people overestimate what they can do in a year and underestimate what they can achieve in ten years." After all, this statement is so true that it is attributed to numerous great people in this world.

As we know, for millennia, science was the domain of the few. Only a few had the chance to understand the laws governing the world (of course at the level of their knowledge) and were able to aggregate and document this knowledge. The first officially known universities such as Al-Karaouine in Fez, founded in 859 A.D., the University of Bologna founded in 1088 A.D or much earlier existing academies e.g. Plato's in Greece from the 4th century BC, were excellent examples of a certain democratization of knowledge. Before them, academies or schools were not institutions, but only assemblies of elected representatives of society. In the following centuries, more universities were established, and numerous cultural revolutions or the enlightenment of peoples led to the establishment of numerous schools, mostly monastic. The first industrial revolution, which began in the second half of the 18th century, initiated not only a change in industry but also in education. The first was the need for universal (basic) education and the second was the need to standardize qualifications (diplomas). The pioneer of universal education was Denmark, but as technology developed, numerous "programs" were created in this field. The revolution in education took place in the 19th century in Prussia. The idea of Friedrich Wilhelm August Fröbel initiated mass education. Of course, his idea of equality in education and access to knowledge from kindergarten had more of an indoctrination and manipulation aspect than a desire to equalize opportunities. Prussian vocational schools training specialized workers or schools for teachers became an inspiration for other countries to introduce multistage, systematic education, which we know today. And here is the problem. When we talk today about how generative AI affects the education system, we mean how it will change the one from the second industrial revolution, where the student is taught by a teacher implementing a commission-approved lesson plan. In this approach, it is indeed worrying that tools with more knowledge and experience than all the teachers in a school combined, available for free (or almost free) and operated in natural language, can pose a serious challenge to ministries of education worldwide. However, it should be noted that many specialized international education programs have long used mechanisms for the coexistence of traditional problem-solving mechanisms and modern technological tools. 

One of these bridges is the levels (SL, HL) used in programs like the IB Course. Mass education was supposed to shape appropriate attitudes. It subordinated young minds to the laws and commands prepared by those in power. Learning to read and write based on specific texts instilled "good" principles and assimilated the guiding thought of the nation's guide. So is it worth considering today how Generative AI will affect this particular education system? I think that if we consider this question on this basis, the following scenario will occur:

• 99% of current education systems will be implemented using AI-adopted programs. Just as it happened when calculators were introduced into teaching programs for technical subjects. The role of the teacher will be to moderate the course of the teaching process carried out by various bots with the function of teaching and checking. This is the new Prussian teaching.
• 1% (of course, much less) is teaching with a master, professor, or Nobel laureate whose knowledge and experience but above all, being with him is valuable. This is the new Aristotle's Academy.

But if we consider another scenario, in which we learn to communicate and cooperate with intelligence - different than a teacher, different than smart audio-visual teaching aids, then the education process can change forever. As a separate process, it will not exist at all. Then AI tools will be a kind of creator of culture and human behaviors. Through appropriately prepared messages and their delivery, we will learn a certain culture and behaviors. AI will create and instill this culture and behavior. This vision is somewhat terrifying because it is not discreet, but specifically named. Today, culture is shaped by language and communication. Very rarely do we identify this phenomenon as negative. By sending children to specific universities or locations, we decide to "condemn" them to live and therefore adapt to a specific culture. Because of people, customs, and language. And we don't see anything wrong with that, on the contrary, it is a positive, even desirable phenomenon. Can we imagine today a course run by AI and so personalized (hyperpersonalization) due to its content and form and perfectly adapted to a specific student, that the efficiency of the education process will be much better than traditional methods? I would bet that with good marketing and a fashion trend, despite initial fear and many ethical objections, such a system will win and oust the conventional, standardized model. Who wouldn't want to learn calculus and its applications in 10 hours instead of 2 semesters at university? Today it is still too early to determine the path we will follow. One thing is certain, COVID-19 changed the education system. What seemed impossible (distance learning) in many parts of the world has created new opportunities. Knowledge and experience (product of the future) that generative AI brings will introduce another revolution in education, and I am sure that it will be on the scale of Wilhelm von Humboldt's vision.

Then Generative AI appears. On the Great Hall of Social Networks stand various presenters showing that it was bad, click, and it is beautiful. Those observing this, often very engaged viewers (companies), dream of their business also changing in this way. They dream of getting rid of all the negatives, leaving only profit. Of course, skeptics also appear on the network, who ask more or less rational questions. Some criticize because they just invested in another technology and don't know what to do with it, others tried and it didn't work, but most of them don't even know what this technology is but for the principle they are against. The Great Hall craze ends in the evening when positively charged listeners, with levels of dopamine and adrenaline reaching the zenith, return home. They talk to their spouses and friends about what they experienced for a few more hours. They can't sleep the night after the conference. They plan and continue to imagine the change that will start from tomorrow. Their life will change. Soon. Resolutions to change their own lives from tomorrow are created. Entrepreneurs after inspiring meetings about new technologies dream in a very similar way. They organize management meetings, inspire and convince each other, and invite various consultants to listen to what is possible. The vision of the salvific revolutionary technology, which will change the world, does not let them rest in planning. Because what will happen if we don't get into it? The competition will get in and we will be left behind. Every year, the Gallup Institute, Pew Research Center, YouGov, and universities around the world investigate how many of us keep New Year's resolutions. It turns out that only about 2-8% of the resolutions are implemented longer than 1 month. 80% are not started at all. Interestingly, these studies show that over 35% of us plan every year from January 1 to change the same things for several years. And it doesn't change. It does nothing completely, finding new reasons and giving a good justification for doing nothing. The same is true with the participants of the congresses. Apart from the delight, adrenaline, and dreaming, nothing changes in their lives. Well, maybe only the thickness of the wallet. 

Entrepreneurs do not launch anything more than Proof of Concept. Reality overwhelms them. They realize how much change costs, how long it lasts, and how incredible effort it requires from them. They are aware that there is a chance but they do nothing with it. Until the next boom. Until the appearance of the next wave of euphoria. And again, another fascination, plans, and resolutions.And what about those 2% who keep New Year's resolutions? They become the heroes of their surroundings. They can boast about their achieved success. They are the leaders of change. They stand on the stage next year and talk about how they realized that change should be made and consistently introduced it. The next generation listens to them dreaming and planning. And the cycle repeats itself. And that's how everything goes around. In the meantime, a quiet change takes place. The regulator, customers, or the market makes an evolutionary technological change. It enforces the use of digital file exchange, automatic settlements in the cloud, or the introduction of digital currency. Quietly, without the lights of the spotlights and pathetic fanfare. And it happens. And if not, a penalty comes or the inevitable vision of its imposition. There is another variant of technological breakthrough. And again, I'll resort to the analogy of the Great Halls. Speakers appear there who have so far achieved tangible successes. They created something great. They were known in their usually narrow field of specialization and achieved incredible things - they achieved them, not just take Instagram pictures with a Ferrari borrowed for 4 hours. They have reached a point in their lives where they are bored and need a change. They are very aware of their value but no longer want to compete, or prove anything, and their ambition is to teach others how to achieve success. Importantly, profit does not matter to them, but rather appreciation and remaining in memory. These are diamonds. They are rare and dedicated to the cause. By becoming a coach or mentor, they consistently change someone's life. Of course, the person whose life is being transformed must put a lot of energy into it. The effect is worth it. You're probably wondering how this works in technology. Well, the model here is open source. Created by experts, enthusiasts, and passionate individuals - I could list even more. There emerges a parallel to commercial (well-funded) technologies, less colorful and usually less user-friendly but much better technology. And as long as there is no desire to commercialize it, it remains the purest form of breakthrough thinking, available to those who want to give something of themselves and make changes not as New Year's resolutions, but out of pure need.

Two weeks ago, I conducted a survey on LinkedIn asking where the biggest benefit of generative AI is seen. Of course, the survey is not representative, but I think it well reflects the current market trend. 67% of responses indicate that generative AI will be used to increase work efficiency (it should make the work I do easier), 19% and 13% respectively for making my life easier and developing my skills. These responses do not surprise me and confirm the research conducted by the Organisation for Economic Co-operation and Development (OECD) and the International Labour Organization (ILO). Looking at history, particularly the turn of the 19th and 20th centuries, over 70% of patents were issued for work-easing solutions. Wait a moment. Isn't the greatest fear of introducing AI or other modern technologies the loss of jobs? Of course, it is. It was, is, and will be because progress forces us to adapt and learn entirely new things useful in new conditions. This can be called development, competence change, or simply adaptation. A few days ago, I had a discussion with one of my friends about why AI is a hot topic now. We pointed to solutions that have been working for at least 10 years, making our lives easier or our work more pleasant. And it's all based on AI. So why is there so much talk about generative AI now? Do companies need to find another topic for development (with the failure of METAVERSE, although I think in a few months it will be loud again when you can buy a new Apple VisionPro), or perhaps because technology company employees are being laid off en masse (though we are talking about single percentages and according to star theory those more administrative than substantive).In my opinion, about which I have already written in several previous articles, three conditions have been met. Cheap, quick results, and most importantly, easy to use. This is what generative AI is today. There is one more interesting (though not surprising) element. Dozens of organizations and startups are springing up, creating their own often open-source solutions. Most of them are trying to address some specific productivity enhancement problem or a scalable use case. One can certainly compare the patent fever with what is happening today. However, there is one significant difference that we do not notice at first glance. Boredom. After the crazy signing up of users for OpenAI subscriptions, the number of users in the last month has fallen by over 10%. And it will fall. Those who saw the potential for REPEATABLE improvement, easing their work, will use it. Those who wanted to try but didn't change their habits will stop using or change the platform to see something new. Just like 20th-century inventions, some are with us today, while others have been forgotten. Those who stayed and are part of everyday life either changed our habits or cemented them. The same thing happened with the Google search engine or social networks. They're like bottle caps or straws for a frappuccino. We don't think, we just use – automatically. Generative AI and most of the use cases currently being tested will tend to hide (white label), weave between currently used solutions, and converge with other technologies. They will be like syrup for a drink or ketchup for fast food. Then we will stop asking ourselves whether AI will destroy us, replace us at work, and whether it will have a negative impact on our lives. Today we don't ask these questions when we think about the Internet. And what's more, we don't blame the Internet (as a technology) for pathologies or negative effects. We blame users for using it badly. And I wish the same for generative AI, that we could blame people for misuse, not the technology itself.

This is the point at which the growth rate significantly slows down, and the value achieved by a given variable (e.g., production, sales, revenues) remains relatively stable over a certain period. This is the time when fear arises among those managing the company - on the one hand, fear of lost opportunities, and on the other hand, fear that a competitor will suddenly emerge and push the current achievements into oblivion. DISRUPTION recently wrote about paranoiacs and leaders stuck in boiling water. This time I want to focus on technology-driven business models. One such model that emerged with the fourth industrial revolution is the Freemium model. It's a particular model, which, when applied, can bring substantial revenue but must be strongly rooted in fast and autonomous technology. What is the phenomenon of this business model about? Traditionally, in the so-called classic world, there is a Premium business model. No one needs an explanation of its principle. We encounter it every time we buy a first-class ticket, obtain VIP status, or receive a Gold credit card. For an additional fee, we get better service, a better product, or better treatment. Creating a premium product is based on the principle of limited availability, prestige, and personalization. The profitability of this business model is directly proportional to the value the customer receives (apparent or real) and, most importantly, to the creation of a habit of its use. Essentially, thanks to the right marketing, a service or product can be rooted in this model. But what does it look like in the Freemium model? The assumption here is as follows. The customer receives a full service or product. Uses it. Appreciates the value it gets. The service or product becomes part of their "daily routine". In the second phase of its use, when the customer is convinced that it brings value, disruptors appear - at first slowly, and then more and more frequently. In the case of mobile games or music services, after a while, ads appear that take up to 30 seconds after a game stage or 5 songs, constantly serving the same advertisements. For financial services, the need to answer survey questions, or for logistics services, having to dig through stacks of flyers and advertising folders. 

The customer is sent a clear message: you have this service because someone is paying for it. Someone is paying and wants you to give a moment of your time in return. But behind this "theater" hides full, often highly advanced technology that measures behavior, tests response levels, and adjusts the "offer" or rather the price for usage to the individual tolerance levels of the customer. This advanced collection of behavioral data and "pain thresholds," as it's defined in the industry, is controlled by artificial intelligence algorithms with a very extensive cognitive apparatus. And no one is surprised by this; most users are to some degree aware that they are the product, but the benefits outweigh the cost (here we could engage in a longer discussion about whether this is indeed the case). Companies using this business model are aware that without the right technology, they can't infer and optimize their margin. However, a new era is coming for the Freemium business model, thanks to Generative AI. And it happens on two levels. The first is the obvious use of tools. Free ones give slower results and leave the user to themselves. Without hints and examples, additionally paid ones lead them by the hand, offer additional unavailable functions, and various templates, examples, and operation effects. The second level is the offer of hyper-personalization and creating one's unique style. Through dialogue (yes, I emphasize this function of generative AI), a trial-and-error profile is built in which the results will be generated in the future. A digital twin of the way of operation and the created effect. Everyone has the opportunity to create their unique style in the vein of Picasso, Byron, or Chopin. But what's most important, once paid for, the operation model not only ties us to a specific platform but also makes it impossible to leave. There is no portability of the "profile"; one can't afford to break the rules, even if they become unfavorable. Leaving means not only losing followers as in social networks or reach but most importantly the established and grounded pattern. We are entering a world where, thanks to generative AI, the Freemium business model will be the most common and widely used model. Many knowledge-based or deeply regulated market firms will have to drastically change their business model to survive and earn. Tax advisory, accounting, or even software writing will be doomed to implement this business model, supported by sophisticated technology and advanced data monetization models because no one will want to pay for services anymore.

Along with the increasing standardization of services and centralization of customer service, the need for efficiency growth arises. The beginning of the 21st century for many is the blossoming of IVR (Interactive Voice Response) systems and „Self-Service” services. And not only in telecommunications companies' call centers but also in banks or administration. And here comes the question of whether this is a new business model. So no. This is not a new business model. Even if we were looking at it from the side of companies offering this service, it cannot be called that. In this case, we can confidently talk about robotization or automation. Here a specific function emerges to us, namely the paralleling of processes. So far, the maximum number of operations was limited to the number of people who performed it. Thanks to automation, it is possible to scale (increase) the number of handled matters without increasing the number of staff, which in consequence leads to increased efficiency. What about the business model? It is defined as a way of earning money on a standardized process where all activities from order to service delivery are carried out automatically - by computer systems. A very nice example showing the self-service business model is book printing (e.g., Amazon KDP) or photo album printing. We go to the website, we teach ourselves how to use the service (tutorials or YouTube videos), then we prepare the design, add content, finish the process, pay, and wait for delivery. In the background, the print is created automatically, sorted to the appropriate delivery places, the payment is completed automatically. What's more, we can only find out about the status of our order from a fully automatic system. You probably see similar examples every day. They are all connected by one common element that, at first glance imperceptible. Namely, the elimination of intermediaries. In most cases, the self-service model is implemented in a simplified B-B-C convention to just B-C. Ordering an Apple computer or even configuring and ordering cars is done this way today. There is another particular characteristic of this model. The entire responsibility for the order (its correctness) is shifted onto the customer. You cannot say that the lady or gentleman in the showroom advised me this or that. This is very typical. The self-service business model, however, has a fundamental weakness that we experience daily. 

You can't contact anyone (I mean a person) to dispel doubts or ask about benefits. We are limited only to the content that was prepared by the supplier. Well, at most, we can look on the Internet on discussion forums for opinions and advice. But that's why it's cheaper and faster. That's why this model is called Self Service. Over the past few years, the self-service business model is becoming increasingly popular. But along with its popularity, customers are expecting its modifications. The feature that is the impossibility of interactive ordering (i.e., advice) and hence the discomfort associated with the uncertainty that what I order will not be what I wanted, makes companies only test it for more complicated services and often withdraw from it or introduce semi-automatic solutions. Generative AI, in particular, advanced solutions supporting decision-making processes, are expected here like nowhere else. The possibility of consultation while maintaining the impression of a conversation with a person increases the comfort of order execution but most importantly, significantly the efficiency of sales (conversion) and cross-selling. The lack of a sense of security and comfort when placing an order or making a complaint, which was the cost of this model and savings for the customer, can be almost completely restored. Additionally, it becomes possible to use services such as an autonomous concierge making decisions for us in routine service and product orders. For this reason, we will soon witness the emergence of a new business model (autonomous) Self Service. Do you see what I'm trying to say? Yes, today's marketing is supposed to encourage us to buy products or services. With this new business model, companies will have to develop marketing that will have to convince our personal AI adviser to purchase them. It's interesting whether they will convince him with Instagram influencers' ads or BOGO-type promotions.

Thanks to realistic animation and similarity to everyday physical laws (movement or color sets), our subconscious recognized them as true. They have become the standard to which we adjust other creations. If you think about it, pop culture is full of patterns considered true but actually coming from a film studio. Thanks to computer animations of these "created images," we can quickly see galaxy collisions, the Big Bang, or jump to superluminal. Visualization of the structure of an atom or blood cells in our blood or the appearance of viruses and bacteria surrounds us from every side. Thanks to computer games, the imagination of graphic designers, and perfect execution, our minds have built up the course of a hydrogen bomb explosion, a 360-degree view of a missile's flight path, or the course of knight sword fights in the 12th century. But do we realize that most of these images are completely untrue? They are as false as images of angels or saints painted on church walls or depictions of the devil and God from medieval books. So how is synthetic media different today? And why do we talk more and more loudly about this issue as a problem? We are dealing with two issues, or rather, problems:

1. Creation is carried out using artificial intelligence based on a real, existing object,
2. The created object can realistically react to changing environmental conditions; it's interactive.

Twenty or thirty years ago, only multi-million dollar film productions could afford to create synthetic dinosaurs or terminators. Today, anyone can create at least the same quality characters (or objects). However, the purpose of such a generation becomes problematic. Historically for entertainment or education. Deliberate deception, as we have established, served to soothe imaginations. Today, besides entertainment, there is a strong desire to increase communication efficiency (a naturally looking bot will perform better than a low-quality cartoon character), but also the temptation for illegal or even criminal actions. Synthetic media used for this purpose can cause a lot of confusion today. Current legal procedures, directing audio and video materials to forensic laboratories, operate within a few to several dozen cases per month. The number of advanced media manipulation cases is still not high. What will happen, however, when these cases will number in the thousands or tens of thousands daily thanks to AI? We will probably use the same technique as the criminals for manipulation, but a revolution in this area will be necessary. However, I am thinking about another aspect of this revolution. To what extent will we be able to believe in what we see and what is presented to us? Today, I look at photos from my youth, and I know they reflect reality; I know these events took place. Our brain works so that it adjusts memories to the currently presented facts. So it can often add a story that didn't take place. You could say it plays a cognitive trick on us. If we start living in a world of synthetic media, and I'm not talking about images of dinosaurs, terminators, or nuclear explosions, but changes in how we looked (filters on social media) or what we did (fabricated photos or clips from exotic island vacations), not only will we deceive ourselves and our loved ones today, but we will believe in the future that it was true a fact. The problem of modification through AI may be one of the biggest challenges we face for another reason; there is social acceptance and consent to this modification. We do not see abuse in it. What's more, "beautifying" reality is already an irreversible process. In the coming years, we will accept further steps in falsifying reality, in the name of innocent fun, entertainment, or jokes. Just as we give up our privacy for free mail, social media accounts, or participation in a quiz, we will agree to hand over biometric data for the price of permanent makeup, a concert film where we weren't, or a postcard from a 5-star hotel vacation. We know that today our sold privacy has turned against us, but imagine how sold biometry can turn our lives into a nightmare.

Modern e-commerce largely draws from basic laws of mathematics and statistics. The primary mechanism is the algorithms' use of the normal distribution, also known as the bell curve, familiar to us from elementary school. This concept has its roots in the 18th century. Abraham de Moivre, a French mathematician, was one of the first to investigate this distribution, working on problems related to gambling. His work on probability and error theory formed the basis for later research. However, it was Johann Carl Friedrich Gauss, a German mathematician, and astronomer, who made a groundbreaking discovery concerning the normal distribution at the turn of the 18th and 19th centuries. Gauss applied the normal distribution to analyze measurement errors in astronomy. His method of least squares, which he used to fit curves, was based on the assumption of error normality. For this reason, the distribution is often called "Gaussian". The idea of applying this distribution in e-commerce is based on the assumption that most of us have very similar tastes and needs. If you are looking for specific shoes, glasses, or a bag, there is a very high probability that you are also interested in the items that people who bought or searched for these goods were searching for or buying. We are all similar. Of course, engineers writing recommendation systems are looking for more precise methods of correlation, but the basic idea is just that. Of course, digital business offers us many more advanced possibilities and can exploit the dynamic asymmetry of information (which I wrote about extensively in my book SPEED and previous articles). One of the mechanisms to maximize revenue is to recommend not only similar products but also to show those on which we have the highest margin or someone paid for their promotion. Using these techniques, it turns out that the product or service recommended by us may not be the best for us, in line with Gresham's law, which can be freely applied in this context (bad money drives out good). While the social or economic impact of using these simplifications for the purchase of consumer products is small and even acceptable, for industrial applications or in the healthcare sector it is unacceptable. With them in mind, AI-based algorithms and simulation systems are being developed. Research on simulation algorithms is an area of scientific research focusing on creating mathematical and computer models to mimic real systems and processes. Through simulation, complex systems that are difficult to study in reality can be analyzed and the behavior of these systems in various conditions can be predicted. Simulating the effects of actions or choices is now one of the main directions of AI development. This is seen as the real revolution or the biggest expected benefits of AI in the coming years. This can be briefly characterized by our example of choosing shoes or a bag. If the consumer's preferences are known and we get to know his individual characteristics (this is crucial), it is possible to assess which of the many products will be best for her or him. Not just based on the normal distribution. Adding just this step makes the results much better. In the past few years, it has been proven that with this approach (personalization), consumer trust in recommendation systems has increased significantly. Proposed goods and services better match needs, so consumer trust increases. But be careful, for this to be possible, IT systems must enter the privacy zone. They need to know or discover (calculate) the non-obvious characteristics of the consumer or user. This is where the first barrier and sometimes a fundamental problem arises. In AI-based simulation systems, we often talk about basic (statistical) recommendations and advanced (behavioral) ones. For industrial systems, we are talking about solutions dedicated to individual installations or even their configurations at specific time intervals (seasonal configurations or set to produce specific parts or components). However, that's not the end. For over 10 years, thanks to the development of AI and the Internet of Things, there has been much talk of entering the Digital Twin (DT) era. This concept is very well known in industrial, transport, or logistics systems. Now it is entering medicine and the broadly understood consumer goods market. What is it and how does it differ from the mentioned statistical or simulation systems? Digital Twin is a digital representation of a real object. Imagine that we have a plane or car equipped with hundreds or even thousands of sensors transmitting data to a central computer. On their basis, i.e., the read parameters and the existing logic of their processing, you can visualize and track in the digital world what is happening with the object in the real world. For most of us, this is a very simple picture to imagine because we have onboard computers in our cars showing brake wear or fuel level in the tank. And that's exactly how it works in the case of DT. But what is the main benefit of this solution? At the current state, we can overlay statistical algorithms and simulations and predict with very high accuracy what will happen in 2, 5, 10 seconds, or a few days or weeks. This is the basis for the concept of autonomous vehicles or predictive maintenance systems. Cyber-Physical Systems (CPS) are the great hope for industrial applications of AI. But is the future of consumer AI in any way related to this? Can we imagine a digital twin for the use of recommendation systems? Before the era of generative AI, this was very hard to imagine. By nature, humans are very difficult to measure. Most of us don't know exactly what we like or why we like or dislike something. It just is. Sometimes our feelings are influenced by the atmosphere or mood. Sometimes one thing tastes good to us in certain circumstances and not in others. For example, a „bombardino” on a ski slope tastes exquisite, but at home, it's just too sweet or even unpalatable. Certain music in a club encourages dancing, but the same music listened to in a car annoys us. The development of generative AI in this area could be revolutionary. We can expect not only recommendations for a particular product but also the context in which we should use it. Recommendations extended by the creation of context. Similarity and simulation are dependent on mood. Or maybe what awaits us is creationism, and what will be recommended to us will be based on proposed, fully created environmental conditions?One thing is certain. Thanks to personalization, recommended products and services will be more to our liking. They will be optimally adapted to needs and environmental conditions, and what we call user experience today will not be expected but predefined. It's worth being aware of that.

Good luck.