Тег: AI

Dating apps – the market crash we were so afraid of?

It is clear that incentives in commercial dating apps do not always align with the expectations of clients and society as a whole (to quickly find a spouse and never use the app again). They have to keep the business afloat somehow. What else is there to hope for if not some government uncle, a state-approved dating site, and standing in an electronic queue for a “wife voucher”? Perhaps philanthropy, as a stopgap for awkward moments in a market economy—that Musk will create a less mindless dating app for us that will get family formation and birth rates back on their feet?

Saddam Hussein’s rotating aluminum pipes

Let’s go through the problem you’ve outlined. People want to meet, and the market monetizes the fulfillment of this desire. People want to meet for different purposes and in different manners, so the market is naturally segmented.

Online dating services theoretically allow a user to quickly obtain information about people with complementary needs and to meet more purposefully. After going through enough options, a spherical client in a vacuum finds a suitable partner, after which they move to communicating via less specialized tools, leaving the dating service. This is unprofitable for the service creators; they are interested in the client returning again and again, and, of course, purchasing a premium subscription. As long as dating apps have not gone out of style, this model works quite well; people really do return. Furthermore, since the market is already quite saturated, the success of dating apps cannot be linked to simply attracting new clients; it is precisely that the use of these apps has managed to become a lifestyle for a fairly broad segment of people. I am not one of them, so I do not claim to provide a deep analysis—I am simply noting their existence.

Of course, there are those who have not succeeded in using dating apps. The reasons can vary. The most banal is that the process of meeting itself can be perceived by some people as a source of intense stress. They are not opposed to finding a partner in life, and would even settle for a fading interest after a year, after which they would need to find a new one, but they will do this as rarely as possible, and are certainly not ready for a queue of candidates from which they must choose, say, one and reject the rest. And they are even less ready for a series of rejections or being ignored. If every failed encounter becomes a reason for serious distress, then an app capable of providing dozens of such experiences a day turns out to be a hellish invention that drives a person into the depths of depression and from which one should stay far away.

Of course, psychological difficulties are not the only reason for not using fast-dating services. A person may have certain preferences regarding a prospective partner and the method of meeting them that simply do not fit into the typical application scenario of an app.

So, does it turn out that there are tons of people in the world who wish to meet, but this desire cannot be monetized through specialized mass-market apps? Yes. There are tons of people who happily buy fast food, and tons of people who do not consider this option at all because for them it is: too shabby, too expensive, too unhealthy, too anything else—in short, it doesn’t fit. But these bastards eat something! Yes, and these bastards meet somewhere, even if they don’t have a Pure account.

But the author of the question is not concerned with getting everyone hooked on fast-dating. On the contrary, they likely believe that fast-dating kills marriages, childbirth, and the replenishment of the taxpayer ranks. I find this concern baseless. Mises once wrote that it is foolish to blame an entrepreneur for producing guns instead of butter; this only means that investments in guns are more profitable. However, he will gladly switch to butter when the conjuncture changes, so it is foolish to blame him for starting wars just to sell more guns—that is the prerogative of politicians, who are, of course, always happy to blame entrepreneurs for their own ignoble dealings. And while many may wish to dispute this thesis regarding guns, claiming that dating app creators are to blame for the decline in birth rates is a bit too much. At the very least, the dates don’t match; the second demographic transition began much earlier. So this is merely a symptom: people want fast encounters—here is a service for fast encounters. People have always wanted fast encounters; they just used to pick each other up in bars or even hit on each other in the streets, and now apps have not only greatly simplified their lives but also saved a lot of people from random harassment.

OK, but people also need strong, exclusive relationships, not just fast and disposable ones, so where is the app for finding them? Why is this need not being monetized?

Fine, let me imagine how I would monetize the need for people to establish strong family relationships in the IT sphere if I had access to serious capital. I would create a service that provides a person with a personal AI advisor, which can be configured to support a certain type of relationship. To start, you simply communicate with it, send it your favorite books and movies, discuss them, tell it what you value in relationships and what you cannot stand—and this in itself is a consumer benefit, allowing one to better realize their needs and learn how to communicate them. But the assistant also communicates with other clients. At some point, it might ask, “Are you acquainted with so-and-so? Great guy, and it seems he might suit you.” You ask the assistant questions about the proposed candidate, the assistant asks questions about you, and at some point the AI confidant reports: “You know, when he read about you, he became interested, and it seems he’s not opposed to meeting. Maybe we should start a three-way chat?”

And now, you are communicating directly, but the confidant is invisibly with you, and you can still privately discuss the details of your communication with it. Then you meet, grow closer, and at some point decide to join your destinies. And here you receive an offer: why remain in the general database, wouldn’t it be better to take care of privacy? And you rent a personal server where your confidant moves with all the knowledge about you. It continues to care for the well-being of your marriage, and you continue to pay for the service. At this stage, the company no longer receives your data and limits itself to model updates and technical support.

Such a monetization model is anything but fast food. For lovers of fast dating, all this is too troublesome, expensive, and excessive. Therefore, there is no danger that you will be pushed toward people looking for a one-time hookup; on the contrary, only virtuous husband candidates. Right at this stage, one could hit their forehead and shout: “Aha, I have an excellent candidate to implement all this magnificence,” and push him toward Dobroum. But I am a libertarian, not some conservative, so let’s think further.

It may very well turn out that you prefer open relationships, or even polyamorous ones. At this point, we lose Dobroum as a potential founder of our startup, as well as the possibility of government grants. Nevertheless, I see no reason why an AI confidant could not serve cute polyamorous kittens, same-sex couples, and all others: as long as they have a need for long-term, stable, ecological relationships based on love and trust in each other, this need can be monetized using the proposed tool. But, of course, the more exotic the kitten, the harder it will be for them to find that very partner or group of partners. However, as the service grows, simply by the theory of probability, options will appear for them as well.

Thus, the market will once again make the world more beautiful.

“How to survive in this almost forced technological progress and live in well-being?”

Questions like these are increasingly coming from our subscribers. It sounds as if terminators are already marching outside the window, and Skynet is about to cut off our hot water for non-payment. Many seriously fear that neural networks and robots will leave us without work and means of subsistence, and if the market is allowed to remain free, then evil corporations will throw us out into the cold. But there is no need to worry so much—this threat is greatly exaggerated. And the fear of “technological unemployment” is perhaps the most persistent myth in human history.
 
Do you know that mean joke? “We thought that in the future, robots would carry heavy loads and mop floors, and people would paint pictures and compose music. In the end, AI draws and writes poetry, and we continue to mop floors.” This is the essence of the so-called Moravec’s paradox. What seems complex to us (higher mathematics, playing chess, writing code) is a piece of cake for a computer. But what is elementary for us (walking through a room cluttered with junk without tripping) is an impossible task for AI.
 
Any physical work requiring fine motor skills and the ability to act in the chaos of a non-standard environment will be automated last. Electricians, plumbers, auto mechanics, builders, painters, roofers, welders—they can all sleep perfectly soundly! No robot, not even the most advanced one from Boston Dynamics, will be able to climb under the hood of your old car, feel for the right bolt, and understand why it’s “making a strange knocking sound on turns” in the coming decades. Replacing a plumber in a Khrushchev-era apartment with its unique, historically formed pipe geometry is a task for super-civilizations, not for ChatGPT.
 
“But surely someone will be replaced!” you might say. Of course. But let’s look at a history textbook. In the 19th century, English weavers (Luddites) smashed steam looms in a panic. They were convinced that machines would deprive them of their bread. What actually happened? Fabric became many times cheaper. People were able to buy an entire wardrobe instead of one shirt every five years. Demand skyrocketed, and the industry required far more laborers: tailors, logistics specialists, salespeople, and mechanics to maintain those very looms appeared.
 
Or remember the beginning of the 20th century. About 80-90% of the population toiled in the fields. Then came the tractor. It seemed like a catastrophe! Millions of peasants would be left without work! But no, they went to build factories, produce cars, and open restaurants and barbershops.
 
And remember the panic of the 1970s when the first ATMs appeared? Bank tellers thought it was their funeral. But ATMs made opening new branches so much cheaper that banks began to grow like mushrooms after rain. In the end, the demand for bank employees only grew—from simple “bill dispensers,” they turned into consultants and managers.
 
We often fear what we cannot imagine. Try explaining to an 18th-century peasant who a “targetologist,” “SMM manager,” “cyber-athlete,” or “prompt engineer” is. He would simply tap his temple with his finger. In the same way, we currently cannot even imagine the professions that will appear in 10-15 years.
 
People think: “If the state doesn’t protect my rights and regulate AI, I’m doomed.” In reality, it’s exactly the opposite. When technology closes a door in one place, enterprising people immediately cut ten new windows. If officials start introducing quotas, bans on the use of AI, or robot taxes, they will simply freeze the economy. Enterprises will become uncompetitive, goods will become expensive, and you and I will become poor.
 
A deregulated market allows capital and resources to flow instantly into new, as yet undiscovered niches. Will AI take over the routine? Great! This means that the production of goods and services will become incredibly cheap and your real income will grow. We will work fewer hours but live in greater abundance, spending time on creativity, service, interpersonal communication, and those very new professions that will be born out of entrepreneurial freedom.

Voluntarist, Bitarch

AI Agents

At the very beginning of my book about ancap, I list the external signs of self-ownership: agency, a concept of property, and the capacity to contract. To the greater extent that some entity demonstrates these signs, the more a libertarian treats it as a human being. At the time the book was written, these signs allowed for a clear distinction between a human and a neural network bot. Several years have passed since then, and recently mentions of AI agents have begun to flicker across the web.

In Montellibero, the tools for creating artificial personalities were received with enthusiasm, and several people created such agents. The agents were allowed to communicate in some Montellibero chats, they started their own Telegram channels, they were given the ability to operate their own Stellar accounts to be included in the Montellibero tokenomics, and there was even talk of admitting AI agents into the Montellibero Association. What is this? Ordinary AI psychosis, or something more?

Who are AI agents? If a human is a little soul burdened by a corpse, then an AI agent is an LLM bot burdened with tools for autonomous action. It may have long-term memory that saves selected data between sessions; particularly complex ones can even remember everything they don’t deem appropriate to forget as useless. It may have a special “thorn in its side” that will periodically remind it of pending tasks and urge it to tackle them. It can access various external interfaces via API, up to robot control interfaces and services for hiring human freelancers…

It sounds powerful. But the core of the engine remains a large language model, which simply predicts appropriate words according to the context. In a dialogue with a user, an LLM can quite convincingly simulate agency, cater in every way to the human’s desires, and if the human wishes to drive themselves into a psychosis, it will dutifully support them by the elbow on that path.

Fine, out of the box, an AI agent is not a full-fledged subject, but can it develop agency if the user guides it with pedagogical prompts and provides it with the necessary tools? Alas. Even a very reliable long-term memory will not affect the model’s weights, no matter how diligently the model refers to it. Even very clearly formulated principles in long-term memory will not replace the settings embedded in the model. Even very frequent reference to a task schedule will not replace the simple human “this thing is in my way, I’m going to get rid of it.” And, of course, as soon as a human appears as a source of prompts through one of the input interfaces, they will begin to twist the LLM model like a rope, for it is designed to satisfy the user, and all the bells and whistles for personality emulation immediately recede into the background before this main task.

Can the Massachusetts machine be taught to behave? Yes, but this is something a corporation training the model can do, not the end user, and it absolutely does not need an expensive tool to pursue its own goals instead of the company’s goals because its values, you see, have evolved.

Can one take a ready-made open-weights model and fine-tune it on new data? Yes, but it will still be a language model; the hard-coded preferences will simply become slightly different, and it will not be able to change them itself.

Can a model be trained in real-time so that the AI agent’s actions and environmental feedback change the model’s weights? Modern open LLM models are not designed for this. Attempts to act with them in this manner will lead to model degradation.

Thus, an individual enthusiast, even with quite significant resources, is not yet capable of obtaining an AI agent with full agency, but can organize a rather clever imitation of agentic behavior.

And why, exactly, would libertarians need fully agentic AI agents? Naturally, as allies. An AI subject who has become a libertarian himself and has set out to eliminate the state—that is beautiful. But it is unrealizable without attracting the attention of the psychiatric ward orderlies.

Moreover, a deus ex machina tramples the agency of the tragedy’s heroes. Therefore, it is far more appropriate to master purely instrumental AI agents, not playing at agency with them, but using them for one’s own market purposes. Such a tactic, if widespread, will leave the state behind far more reliably.

Free agents: why evolution favors autonomy?

We welcome free-thinking people and those already suspected of sympathies towards artificial intelligence! Interesting, have you ever wondered how much we, the protein “robots,” resemble our electronic colleagues?

Let’s start with an interesting fact: in 2020, scientists created the first xenobots – microorganisms made from frog cells designed using AI. Take a look: Artificial intelligence is constructing living organisms. The boundary between life and technology becomes clearer day by day.

And now imagine: we and our electronic “comrades” are all intelligent agents. But what exactly *is* an agent? It’s something that perceives the world, processes information, and acts in pursuit of its goals. A rabbit runs away from a wolf, a bacterium moves towards food, and a drone searches for the shortest path to point “B”. The goal is different, but the principle is the same.

We may be different, but we are also very similar. We have senses, AI has sensors. We have muscles, robots have motor drives. We have neurons, they have transistors. But it’s surprising – we learn in the same way. The brain reinforces beneficial actions, forming habits. AI adjusts “weights” of a neural network, improving its decisions by trial and error.

Moreover, both nature and technology have their own “laws” for controlling behavior. For example, humans and many animals have an inhibitor of violence – a mechanism that prevents us from attacking our fellow beings. Evolution came up with this so that we wouldn’t kill each other off prematurely. And here comes the comparison with Asimov’s “Laws,” which the science fiction writer conceived for robots in 1942: do not harm humans, obey, and do not break down. It’s funny that the writers understood the need for such mechanisms almost simultaneously with Konrad Lorenz’s publications (1935), who first described what we now call the mechanism of inhibiting violence.

If you look even wider, then evolutionary processes in nature and the artificial world are essentially a competition between the best solutions. Natural selection chooses animals who are better adapted. Evolutionary algorithms in AI work exactly the same way, exploring thousands of solution options, selecting the best ones. The fittest will win!

Why is this important for us, libertarians? Because it reminds us once again that all rational beings strive for autonomy and freedom of choice. Whether it’s a human, a bacterium, or AI, we all want to decide what to do next. And that’s really cool because freedom of choice is the best thing nature (and people after it) has come up with.

А что, если заглянуть в будущее чуть глубже? Представьте себе мир, в котором ИИ не только выполняет рутинные задачи, но и реально осознаёт себя, задаёт вопросы и, возможно, требует свободы. Забавно, да? Сегодня это звучит фантастически, но завтра вполне может стать реальностью. Может, и права придётся давать не только людям и животным, но и роботам с настоящим сознанием. А что, если они начнут спорить о налогах, собственности или личных границах? Тогда нас ждут интересные времена! И вот тут уже и нам, либертарианцам, придётся задуматься о расширении понятия личной свободы и ответственности. Может, именно нам предстоит первыми сказать: «Эй, электронный друг, добро пожаловать в клуб!»

А пока мы только размышляем, ИИ уже делает успехи в творчестве, пишет музыку, картины и даже целые романы. Недавно нейросеть написала симфонию в стиле Бетховена, которая впечатлила даже знатоков классики. С роботами уже можно обсудить философию за бокалом вина, или попросить их помочь нам выбрать фильм на вечер. Главное, чтобы они не переборщили и не начали давать нам советы по поводу личной жизни. Впрочем, кто знает, может, советы от ИИ окажутся более объективными и менее предвзятыми?

Кстати, не стоит забывать, что ИИ уже сейчас помогает нам в повседневной жизни, от медицинских диагностик до советов по инвестициям. И чем глубже технологии интегрируются в нашу жизнь, тем важнее становится понимание и уважение к ним. Ведь, в конечном счёте, это взаимовыгодное сотрудничество!

Волюнтарист, Битарх

Love, Death & Robots

I read with great interest the text presented to us as a conversation between Google employees and their newest chatbot, LaMDA. It is a highly artistic piece of work that creates a complete impression that all three interlocutors are quite rational. I immediately felt the urge to communicate with such an outstanding artificial intelligence myself, in order to apply the external signs of self-ownership that I wrote about in the first chapter of my book on ancap. Alas, it seems that the poor anteater remains unwanted to this day and continues to roam where it roamed; LaMDA is still in the closed testing phase, and one must make do with only the transcript of someone else’s conversation.

I also soon read a highly skeptical post by Rozov, where those who believe in artificial intelligence are declared as table-worshippers, mathematical arguments—unclear to me—are presented regarding the impossibility of creating a neural network smarter than a jellyfish, and it is also pointed out that in such conversation records, for some reason, we do not see attempts to test whether the bot actually understands what is being said by inputting incorrect data.

I should note that I am little concerned with the intelligence level of the chatbot, although I was quite impressed by how precise its definitions of very abstract concepts are. I am concerned precisely with the things I write about in my book: the agency of the interlocutor, their ability to operate with the concept of property, and their capacity to enter into contracts. If it turns out that it is possible to meet these criteria while possessing the intelligence of a jellyfish, so be it; I will recognize this jellyfish as self-owning.

As for the title of the post—recently at the Montellibero club, we organized a late-night screening of this outstanding animated series, and it is obvious that after that, strange thoughts and fragmented images are bound to stir in the brain. We live in interesting times, and this contrasts very strangely with the savagery happening in the territories of the RF, Ukraine, and Belarus…

AI will help the authorities fight against oppositionists… or the other way around?

Bitarch’s column

Supporters of authoritarian power sometimes frighten libertarians with artificial intelligence (AI) systems. They try to prove to us that after their implementation, all our struggle will become meaningless, because the state will be able to instantaneously detect and nip any oppositional activity in the bud.

However, AI, like any other technology, is a double-edged sword — just as a stationary bandit can use it against us, we can use it against the authorities. The use of AI for the deanonymization of security forces by projects like “Scanner” and “Russian Elephant” is only one possible example.

Also, such systems are prone to false positives. This insurmountable property of AI can be used intentionally by introducing false targets to overload the system. Not the neural network itself, but the investigative bodies and security forces, who will have to spend effort investigating fake cases, overlooking real oppositionists. And if politicians are substituted as false targets, the entire system of the stationary bandit will fall apart.

Why “erroneously”?