Wellbeing by Aleksander Poniewierski Consulting

LLMs_ The Zoltar of the 21st Century

The esoteric market (divination, horoscopes, and prophecies) worldwide is worth hundreds of billions of dollars. It is equally popular in Poland. Every day, hundreds of thousands of Poles turn to divination and prophecies. Websites analyzing dreams (dream dictionaries) are visited by tens of thousands of people daily. One could say that this is a harmless form of entertainment, were it not for the fact that hundreds of people make significant decisions about their future based on divination or prophecies. A very popular form of entertainment in the early 20th century at amusement parks and traveling circuses was the ZOLTAR machine. A wizard in a turban, who for a few cents, accompanied by Persian rhythms, would print a ticket with a life prophecy. As we know, such prophecies can always be tailored to reality, so everyone can find something for themselves. It's similar to the fortune cookies at a Chinese restaurant or the horoscopes from gazeta.pl. Many studies have been conducted to explain why people not only believe in such prophecies but also remember them for a long time. The mechanical Zoltar was attributed with reasoning abilities and empathy towards the customer, while operating 24/7/365. A few months ago, the Apple team published a paper titled "GSM-Symbolic: Understanding the Limitations of Mathematical Reasoning in Large Language Models," which clearly states that LLM models do not have reasoning abilities, but only statistical mechanisms for calculating probabilities. It seems that everyone should be aware of this, but that is not the case. LLM models are like the Zoltar of the 21st century. A few days ago, with the appearance of the SORA model, the internet was flooded with videos showing generated characters that, despite being very realistic, broke all physical laws. Skiers going uphill, gymnasts with three pairs of legs performing mid-air somersaults as if gravity had no effect on them. Laughter, mockery, and widespread criticism. Crowds of former LLM model supporters, as a prelude to the destruction of humanity and the omnipresence of artificial superintelligence, received a cold shower. In the Gospel of John (20:24–29), Doubting Thomas had to see in order to believe. In our AI world, "Blessed are those who have not seen and yet have believed" has now become a foundation for mockery, as many have been persuaded that their professional and life future depends on a quick and unreflective plunge into the world of Zoltar. Artificial intelligence (machine learning models) has powerful applications in industry, transport, and science. This is indisputable. Wherever real-time (or near real-time) decision-making gives a competitive edge, or initiating the correct—optimal—action is valuable in itself, AI is thriving. These areas are developing at an incredible pace. The effects of AI are not visible, but their consequences are. They bring benefits. Any Zoltar-like applications, not only extremely costly (OpenAI fears it won't be able to offer users access to the latest models because they lack the computational power), but also with results that, outside of entertainment and the ability to stimulate endorphins, do not bring business benefits. In other words, they are useless. It's fun to generate an image (especially if I don't have talent and can suddenly make something I've always dreamed of) or write a marketing strategy. But when you calculate the costs versus the benefits, there’s no way it can be justified. And here's a reflection: Zoltar was not supposed to predict the future. It was supposed to be a tool for making money off emotionally unstable people or those who treated such divinations as a form of entertainment. The machine paid for itself in one month, which is why it was a hit in its time. LLMs don't pay for themselves so quickly, so they are gradually becoming a burden on their environment. Undoubtedly, they raised awareness of advanced analytical technologies and sparked the imagination of millions of people. They also provided easy money for hundreds of thousands. Zoltars faded into obscurity with the advent of slot machines. After all, in the end, people count their expenses and expect to win. A picture from Tikal in Guatemala. Monumental structures that cost the lives of thousands of inhabitants. The immense costs of construction and divine power over the masses. These buildings also served scientific and research purposes (astrology), but that was a marginal application.  This image is dedicated to all those who understand what it means to carry the pyramids. To those to whom they bring eternal glory and strength, and to those who must pay for it.

The High Price of Premium Access

Guess what… 10 times the price In the 1950s and 1960s, consumerism reached its peak worldwide. The wealthier societies in Europe after World War II, as well as those in the USA, where new pop culture was emerging that would set global trends in fashion and daily life, significantly changed the form of communication. With the rise of television and daily newspapers, product marketing took on a new level. Many excellent experts emerged with the goal of reaching as large a group of customers as possible and convincing them that they should buy. Not just that they should buy, but that it would be shameful not to have it.One of the most famous advertising magicians of the time was David Ogilvy. A controversial figure, often criticized. However, no one can deny that he changed our world forever. He was the one who said, "The consumer isn't a moron, she's your wife." Numerous strategies developed by him and his team still form the basic approach to campaigns today. It is a canon. In this forest of innovations developed in New York, London, or Tokyo, there is one principle that is extremely useful and effective for business leaders. This principle is leaving the customer (or employee) with a certain, but highly controlled, space for interpretation. Ogilvy said that advertising must make the customer think.According to this principle, a brand should not provide the solution to the customer on a silver platter. Instead, it should prepare the message very carefully so that its interpretation by the customers is clear, and often obvious. In advertising, this technique is realized through a strong slogan and an image or picture associated with the product. "Just Do It" and the Nike logo are understandable worldwide. Of course, you could write, "Buy Nike shoes because they're the best for running, basketball, etc.," but that would insult the intelligence of the audience. "Red Bull Gives You Wings," "I'm Lovin' It" for McDonald's, or many others, are direct but not entirely so.Many companies have mastered this method, motivating their customers effectively. We are used to it. But before our eyes, an evolution of the message is happening, particularly in the realm of modern technology. What does the new message look like? The world, with our consent, is splitting into two main groups. The first group is the more prestigious, wealthy, and demanding one, served by specialists and dedicated experts (a group with time and money), while the second group is focused on cost, and the service can be performed by a machine. An example is today’s banking, helpline services, or even medical advice. You handle your matter because it’s typical. Marketing has worked this out with a simple message: "Don't waste time in line, do it yourself for free."We can observe this trend with knowledge-based products. Companies that launched their products in the typical Silicon Valley way (build mass, then sculpt it) give their services for free or in a freemium business model. They face a huge challenge in starting to make a profit. It’s even said that once you’ve given something for free, you’ll never make users pay. And this is where the genius of Ogilvy and Rosser Reeves, known for the "Unique Selling Proposition" (USP) concept, comes in. The idea is that after a period of use and getting familiar with the product, its full price must be ridiculously high. Not just expensive. Extremely expensive. Less than 24 hours ago, OpenAI applied this method exactly. The paid subscription won’t cost $20. It will cost $200. And you know what? People will pay because they will realize they are joining the first group, and the second group, the ones who don’t pay, won’t have access to the full experience. That’s how it’s done.

Black Friday

50 kilometers outside Warsaw, no one knows what the future of AI will look like. It's astonishing. There is a complete lack of respect and understanding of what is important. Of course, it's a bit ironic. Politicians, when asked about the price of bread or butter, don't know the answer. People in smaller towns or villages have no idea what ChatGPT is. Only a very small, almost microscopic part of society is aware of the decline in token prices or the time needed to train a model. The topic of new technologies is very important. Just as important as the economy, demographic issues, or national security. But for the average global citizen, these topics are secondary or even tertiary. A few years ago, I visited the R&D department of one of the largest universities in the United States. A group of young innovators there invented a shoe sole made using 3D technology. After a few years, these soles became the standard for most Adidas sneakers. Some people like them, others don’t care for them. Talking to these guys, I felt they were doing something extraordinary. They were changing the world and our everyday lives. Today, I don’t think about it at all. Only a vague memory of that meeting comes to mind whenever I put them on. What we discuss in very isolated expert, business, or political groups has little to do with the everyday problems of citizens around the world. It's important to realize this. A 10% drop in the price of gold or palladium on the Kuala Lumpur stock exchange is just as relevant to a villager near Poznań as the price of potatoes in Somalia or cattle in Ethiopia. There are about 300 million motorcycles in the world. Every Asian resort is teeming with them, not to mention India, where they are a primary mode of transport. But few people know that there are about 100 million AK-47 rifles (Kalashnikovs) worldwide, with over 50% in the hands of citizens, not the military. The estimated number of rounds of ammunition for these weapons is 21 billion, meaning there are nearly three rounds for every person on earth. 95% of companies worldwide are or will be users of artificial intelligence. Only 5% of companies create it to some extent. We buy, use, and don’t worry about the valuation of the companies producing it. Society, just like individual countries, has long been divided into givers and receivers in many different dimensions. Mass-market technology has been with us for 250 years. Consumer goods entered the market 100 years ago. In the average household in the Western world, fewer than 5% of items are older than 50 years, and that’s only because they have collector’s or sentimental value. In less developed economies, over 30% of items are older than 50 years. In the poorest regions of the world, this figure reaches nearly 50%. The consumerist society is bombarded with new products, trends, and is constantly encouraged to make more purchases. The problem of fast fashion, popularized with business models like Zalando, is being discussed more loudly. Ten years ago, Black Friday encouraged people to embark on long trips to shopping malls. Now we see an exponential rise in deliveries to our doorsteps. The question is whether modern technology has taken our jobs or if we simply no longer have shop assistants but delivery drivers. Habits, needs, and motivations are changing. But remember, we are only talking about 30% of the world’s population.

Obi-Wan vs Bill Ban

Obi-Wan vs Bill Ban Australia is considering a ban on social media use for children under the age of 16. Interestingly, the ban would apply even if parents consented to their children using such platforms. Biometric testing for this system is expected to begin later this year. On the one hand, it’s good that the issue is being acknowledged and systemic attempts are being made to address it. On the other hand, it’s easy to imagine workarounds, from VPNs to registering devices in other countries. What puzzles me, though, is why, in a world increasingly captivated by AI, we’re not exploring effective methods of using this tool to cleanse social networks of harmful content. Instead, we choose the seemingly simplest but least effective route. Why aren’t platform operators obligated to verify content and counteract harmful phenomena? Public television and radio stations are required to control their content. The responsibility for socially responsible business practices falls squarely on them. So why do we accept that social platforms are not accountable for the content they host? If they can recommend content with surgical precision through behavioral engines, why can’t they counteract harmful material? Ah, yes, there’s freedom of speech and access to adult content. The business must go on, of course. It reminds me of Tolkien’s Gandalf or his mentor-like counterpart, Obi-Wan Kenobi. Someone would have to determine what is good and what is bad. Social reactions, like reporting harmful content, are ineffective because they require statistical confirmation. Even when such confirmation exists, platforms often ignore reports. I’ve conducted experiments, and current mechanisms are ineffective. What about AI? Could it serve something positive instead of churning out frivolous images or monotonous melodies? Absolutely. The problem is, no one seems interested. Youth under 16 years old make up over 30% of social media users. The proportion of educational and worldview-shaping content is less than 10% of the platforms' offerings. The rest consists of drama, controversies, donation-seeking (e.g., TikTokers insulting viewers with comments like, “Are you too poor to send gifts? Get lost.”), and various scams. The problem is, such content drives clicks and views. What if live streams were automatically blocked (like they currently are when someone lights up a cigarette on camera), and the creator was permanently banned? Would humanity suffer a loss? Blocking a user’s account for a single indecent word or harmful behavior could change platforms forever—perhaps even drive them to extinction. Looking at it this way, we face a question: should we prohibit children from watching drunken brawls, or should we ban online broadcasts of such events altogether? Wouldn’t it be better to sell alcohol or other substances in strictly designated and controlled locations (physical and digital) rather than create complicated access restriction systems? Why should all citizens fund control systems when the costs could be passed on to companies running such businesses? It reminds me of establishments like strip clubs or nightclubs. Nobody demands biometric proof or other complex control systems there. Clear regulations apply to business owners, with severe penalties for violations.

The Power of Suggestion

The Power of Suggestion Friday, June 14, 2024 by Aleksander Poniewierski, APConsulting We all know that expensive perfumes smell better than cheap ones. A heavy wine bottle must contain a superior drink. A well-dressed gentleman scores more points with future in-laws than one in casual attire. Even a pricey car creates the perception that its driver (often not the owner) is a wealthy person. We fall for appearances. Sociologists conduct various studies to understand and explain why we are so susceptible to suggestion. The placebo effect or Asch's conformity experiments demonstrate that, as a rule, we are very easily influenced. This tendency is often ruthlessly exploited by various fraudsters to our detriment. In nature, the suggestive colors of insects or birds are designed to mislead potential predators. In a store, they aim to persuade wealthy customers to make purchases. Things are not as they appear or as they are described. In today’s world, this applies especially—if not primarily—to technology. From everyday tools to sophisticated devices for aesthetic medicine, the art of manipulation has reached new levels. In the world of data, this manipulation is used even more relentlessly. A few days ago, Apple unveiled its new iOS 18 operating system. Hundreds of articles hailed the innovations and features that will “change our world forever.” In reality, it’s just a handful of minor improvements and gimmicks (and the intelligent assistant isn’t even available yet—for those eager to download the developer beta, don’t bother). Meanwhile, heated debates rage on about an allegedly unforgivable clause in Adobe’s new terms of service. This clause asks users to accept that the results of their work—e.g., photos or graphics—can be used to train AI algorithms. Few noticed that such a provision (albeit in a different form) has been part of the Creative Cloud license agreement for quite some time. Euphoria and outrage. Proclamations of a new era or the end of the world. Conformity driven by the faces of famous celebrities or multimillionaires makes us believe in fairy tales. Naivety or laziness? Or perhaps a tendency to rely on "authorities" without even skimming the details? Yesterday, I read an excellent analysis of the behavior and content published by a national security influencer. It was a short text but brilliantly described how easily we allow ourselves to be manipulated—or even have stereotypes or ways of thinking implanted in our minds—just because this influencer looks good, has a calm, intelligent-sounding voice, and an appealing demeanor. What they say doesn’t really matter, as most of us don’t rationally analyze facts; we simply accept them as truth. It’s like listening to songs in Spanish, Korean, or English when you don’t know the language. You have no idea what the song is about, but it sounds good. The problem arises when you use such a song as background music for a video and publish it. If you're unlucky, that combination might be, at best, laughable. The power of suggestion is a formidable weapon, and when wielded in social media, it shapes the perceptions, preferences, and choices of millions.