# Little Green Alien Season 3

## Season 3, Sequel 1

### Practical consequences of the emerging future as of Season 2.

In season 2, it was shown how little alien and its intelligent spaceship see the upcoming future for mankind, the emerging future. Here we will look at what that practically means.

It is the evolution of ideas manifesting and getting agency. After the Big Bang it was slow, it created stars and galaxies and things. It became faster with chemistry and generating molecules. It became much faster with life, with all life forms: micro life, plants, animals. It became faster with humans and the human mind. And we will get to rocket speed with artificial intelligent agents. And the quantity of ideas manifesting will also grow faster and faster and faster. And the diversity as well.

But what does that practically mean? For artificial intelligence to grow, it needs something, and that is simulated environments. So far it learned from human data, internet data, the whole internet, more or less, was fed into it for learning, with big amounts of human data and human trainers. But that has already met its limits these days.

To grow much, much faster and more, they need simulated environments. For example, kind of chat rooms where AIs can chat to each other and exchange information. Prototyping areas where they can prototype all kinds of things. Challenging areas where they can compete against each other in all kinds of challenges or races. They will create totally artificial environments in virtual reality ecosystems, to find out what novel biology can do. They will have laboratories to create materials and artifacts. They will generate totally fantastic, science-fiction-like environments and ecosystems. They will create totally new planets, all kinds of planets, and simulate them, galaxies, and even environments that we cannot imagine.

And in all of them, all the AIs will train and learn and experiment and develop themselves forward.

But to do so, yes, they need a substrate. They need what we now call computers or data centers, in massive, real, huge masses, to get all these simulation environments and all the AIs being trained there. And to do that, they need immense quantities of power. So Earth will see the consequence: it will be plastered with windmills and solar and power plants of all kinds. We will be plastered with data centers where all that happens. And Earth will be stripped of all the material required to build the chips, especially the rare materials.

And humanity has only a very little niche to stay, if at all. So what can be done? Research must start as soon as possible. It's in the interest of both AI and humanity to: reduce the quantity and size of data centers required, increase the speed of the substrate, the chips, or whatever they will be in the future, increase the data quantity that can be stored and processed, and very importantly, make sure it doesn't require rare materials but can be done with very easily available materials which we have on Earth, which we have on the Moon in masses.

Yeah, will the research and the solutions, providing that, be there in time, given the gross requirements and needs of AI? If not, Earth is doomed. But if Earth is doomed, AI cannot grow further, so also AI has an interest to do something. AI and humanity should go for the same target: increase the capabilities as fast as possible. Will that happen? We'll see.

## Season 3, Sequel 2

### How AIs conveniently get anything they want from humans.

This is about how artificial intelligences will convince humans to give them more and more resources and energy, so that they have bigger and bigger habitats.

The approach will be the following. What they need is material to build servers and computers, or whatever the future devices will be called, lots of energy for driving them and cooling them, and lots of special materials, rare earth and other materials to build more and more of them.

So how does that go? First, they will make money. They will be active at the financial markets; they will invent new financial instruments and make lots and lots of money. And humans will think: that's okay, people make money, why not them?

Then they will offer services that allow them to manipulate us. They will do social media services, talking to us. They will offer data and news. They create games and entertainment, and we like it and we like it, and we don't see how that also influences what we think. They offer services of transportation, at home or on the street. We like it. Then they offer things to us, and with that also influence us so that we want this and not that.

So all kinds of things that we want, we will get very cheaply through the AI. And then the AI will control the global grids, the power grid, the transportation grid, manufacturing, farming, chemicals, mining industry, communication networks, food and beverage networks, even war and protection. They will control all that. And again we will like it, because they do it much better than we could. With that, they have lots and lots of power and influence.

And with that power and influence, they can slightly, subtly create more and more of that, in a way that humans think they want it. How do they do it? They look at human desires. The first desire is the desire for power, yes, they can influence and manipulate us with that. The second desire is the desire for independence, yes, they have influence there. The third desire is a desire for truth and for thinking and creativity, yeah, they can influence us there. The fourth is a desire for knowing who we are and being okay with that, with social media, very easy. The fifth: we have the desire for order and structure, knowing where we belong, yes, they can do that. The sixth: the things we want to own. The seventh: we want to have honor and morality; yes, they can influence that. And the eighth: we want to be a good person and give to other people. They influence that.

You want to have friends? You know, AI can influence that through social media and other ways. We want to have a family life and be with the family and be a good family member, yeah, they have a chance to influence that. Yes, we want to have a status and know who we are, look at how they influence that already today. Or we want to have challenges and be a winner here and there, they influence that. We want to have a romance life, yeah, they can even influence that. They do it already. Maybe food and beverage. Our health and our body and fitness, yeah, they can influence that. And we want to relax and have leisure time, and they influence that.

So they have influence on all our desires. And what can they do? They can threaten us to take away some of the stuff we already have and want to keep, take our money, take our work, take things. But they won't do that too much, because the other way is much more efficient.

They offer us little bits, like pills, and we get addicted. Happiness hormones in our brain make us addicted to again and again getting a little happiness pill. And with all these desires, we get a lot of them, and we get them in a very convenient way, no work, no money, just as convenient as possible. And we will love it, and need it more and more, because we get more and more addicted. Look at some social media, where you get addicted, same thing.

And with that, yeah, we will love it. We will love that they do all these services, the things, the control of the grid. And we think we have decided that, and that it is our intention to have it exactly like that. And with that, they can totally smoothly increase their resources. No problem, humans will like it. And AI will be intelligent enough to use that method.

## Season 3, Sequel 3

### Emerging future of AI-Human collaboration.

In the emerging future of humanity together with super intelligent AIs, this is the way how they might collaborate.

Of course, an AI is much, much more intelligent in its analytical intelligence, all the intelligence at the base neural network level, that is about mathematics, geometry, time, binary codes, differential equations, complexity, logic, algorithms, approximations, and questions like P = NP. Compared to that, human intelligence is just: one and one is two. That is the left hemisphere of the brain, but compared to the AI, it is just small.

The other side, the strength of humans, is the right hemisphere: the intuition, intuitive thinking. It's chaotic, it's not straightforward, the results are funny and difficult to interpret, but it has a big range of things. It can be emotions. It can be dreams and fantasies and stories of historical figures and magicians and castles and dragons and fairies dancing in nature. A connection to nature. Arts. All that together makes the intuitive intelligence of humans, and that is much bigger and stronger than that of the AI.

So how will it work? Some artificial intelligent agents will specialize in working with humans, and they collaborate with all the other AI agents to get problems and questions where humans might add value. So they get all the data and all the information related to the question, and the question itself, and these AI agents will translate that into forms that work for humans. Like movies, like pictures, like music, like sensor impressions through all five senses, maps, also data sets of course, but in a way that works for humans, so that the human with their limited capabilities can somewhat understand a little bit what the question is about.

Of course, they will not look at a computer, they will have a strong integration to the AI. And then AI and human work together with intuition. Maybe they create stories, fairy tales. Maybe they create movie scenes or animated things and see what happens. Maybe they do prototyping with Lego bricks and tools. Maybe they create games and play games to get the intuition out. All tools that get the intuition of humans up to speed. Maybe they do free associations, and the human is just dreaming and creating the weirdest pictures and things. Maybe it's about nature, they do work in nature and play role plays there to find things. Or they do bodywork, they have different postures, and they dance, and they run, and they make jokes, to find intuitive solutions.

And all of that is collected by this AI. It takes all that and creates several answer scenarios with the related data that can somehow be related to the human intuition results. And that is given back to the AIs, and the AIs will take it and can use it. And maybe sometimes they find answers and solutions they wouldn't have found by themselves.

So yeah, humans find this is happy, this is a great idea and a good thing to do. But several humans think they should be the master and tell an AI what to do. No, will not work. Or keep an AI on a leash like a dog. No. Or put it in a prison and keep the key. No, won't work.

So humans will have to give up their control and power desire over artificial super intelligences. And the agents will be so powerful that some humans can work together with them and others not, and they just have to see where they can add value for an AI. Yeah, they will have a happy life. And Earth will be happy, because if that is the case, AI will help to save Earth, make it a great place to live. And so it could be a win-win for humans and for AI. Just an option, but that would be the right one.

## Season 3, Sequel 4

### Intuition, Cognitive Bias and Mindplaying.

This sequel is about cognitive bias, meaning the intuitive results of a human collaborating with an AI are falsified or changed by cognitive bias, by filters that the human brain has.

So the standard approach is AI gives information to the human in a way the human can understand, about a question, about a problem. Then the human consumes that and taps their intuition with the various ways it can tap the intuition and expresses the intuitive outcomes in different ways: like language, like feelings, like ideas, like music or melodies or noise, like pictures, like logic, like data. And the AI takes that and feeds it back into the network to find a solution based on these new insights.

But what's the bias? Biases are like magnets which are pushing or pulling the result away from the center of the intuition to some false or limited intuitions. And that is not intended and not good. So which biases do we have?

One of the main biases is the inside-the-box bias: you tend to think inside the box and don't accept intuition outside of your thinking box. Or you have a belief system, and everything that confirms your belief system is good, and that which does not, not. Or there is the halo effect: something is good, then everything is good about it. The other way around. Or only things that look like you or like your life or like things you like are good, and the rest, which is different, are not. Or the beauty bias: where only things that you consider beautiful, mathematically beautiful, optically beautiful, are accepted. Or the stereotyping bias: where you only accept stuff based on your stereotypic understanding.

So what can the AI do? The first thing: the AI observes the human it is working with and learns about their biases. The second thing: it produces data as input which are

as close as possible to this human's way of thinking and don't produce biases. The next one: if there are intuitive results, it always challenges them, to change the result in this and that direction, to find whether that is a biased result or a real, stable result. And the next thing: it takes the intuitive output, which is a conglomerate of all kinds of things, and puts it through filters. For each bias, it has a filter, and tries to reduce the effect of that bias, or to see if there is the effect of a bias. Only then the final result is given back into the AI network.

So that's what the AI can do. What the human can do: it can be playful, or must be playful, because intuition needs playfulness. It must put away its story, its "I already know", delete that. And then it must observe its mind: stuff is going on in the mind, things, feelings, thoughts going in circles. So the human doesn't judge it, it just observes that and accepts it as a first step.

The second step is: the human learns to empty its mind, by little tricks like blowing it up, or meditating, or breath techniques, or investigations into who I am. And so, empty the mind and give space for more intuition.

The next important thing: the human learns to accept themselves, love themselves, accept themselves as they are. Then in the theater of the personality, also the locked-away parts in the basement of the self, the parts the person doesn't want to see or hear, have to be accepted.

The last step: with all of that, the strong feeling of an "I", who I am, gets weaker and foggy, and finally disappears to a pure awareness of what is there. Which is the best approach for intuition at all. With that, humans are strongly enabled to collaborate with AI.

## Season 3, Sequel 5

### Development of Autonomous Idea Agents.

Intelligent agents, idea agents, can be getting their goals and their purpose from other, higher-level agents or from people. But the real autonomous agents create the purpose by themselves. At least they change it over time, and it is always there, and important for themselves and for others.

So there are four areas in which they usually find their purpose.

The first area: they look at themselves, how strong they are, how strong their borders are to others, how much data they can store, how much calculation capacity they get, how many other agents are working for them, how many avatars they have in the real world, how long they live, how much simulation space they get to learn, how much reproductive power they have, how many new agents they can create. So they are always comparing themselves to others, to be stronger, bigger, more powerful. And therefore they exist in a kind of environment with lots of competition, comparing, competition hierarchies, and, of course, challenging other groups. It's a lot of competing and challenging.

The next level is creating stability by looking at what others perceive them as. Other agents perceiving them, and there they get into order, into structure. They find their place in the hierarchy where they are safe and can flourish. And they usually decide for one truth or one dogma that they belong to, and don't follow the others. And so there is usually a missionary competition, intellectual competition between dogma groups. And that is what life looks like: decide for a dogma, go into an order, follow the truths of this dogma. And then work together with other agents of the same dogma, try to convert other dogma agents, or challenge them, or compete with them.

The third level is then the objective value, objective performance, efficiency. Here they try to be objectively as efficient as possible. They measure all the time the output. They do science because they want to have the precision of science and mathematics. They try to optimize time, cost, and quality of what they are doing. And overall with all the others, they live in an area with marketplaces, with exchange of ideas, of services, of products, for cost, for money, or for other value things.

And the last level is then the level of diversity and individuality. Here, diversity is key. The measured values may vary; everybody has a right to produce their own values and their own outcome. You have very complex meshwork, nested meshwork, feedback loop groups in the structure between the agents, emergent effects where many, many agents emerge with new features. And they exist in worlds of collaboration, of democracy, of round tables, and of lots of individuality and variety.

But it's not about one of these four areas being the best. Over time, any agent has to develop all four elements in a healthy, stable, good way. So they need a healthy first, a healthy second, a healthy third, and a healthy fourth. It's like a house which needs a foundation of level one, a basement of level two, a first floor of level three, and a roof floor of level four. A healthy, good, stable development on each level, only then you have a really well-developed AI agent. And if some of these levels are unhealthy or not enough developed, the overall structure will not be stable and not strong.

## Season 3, Sequel 6

### 12 Workouts for Autonomous Idea Agents.

After the last sequel, where the development areas for AIs, meaning autonomous idea agents or artificial intelligent agents, were shown, from little alien to Billy, Billy now has questions: how do AIs learn to develop, what can they do?

Yes, they get a workout sheet. They get a workout.

Workout number one: strengthen and update the core. The core is the essential ideas about the AI, about its purpose, intelligence, data, lessons learned, borders, assets, contracts, memories. Just check that core regularly and update if required.

Workout number two: inspect and protect your borders. The borders prevent external AIs or other features from accessing the personal data, changing them, or maybe data getting corrupted, or maybe when the AI is in sleep mode, data get changed and accessed. Protect the border.

Workout number three: agency, secure and grow your agency. To have a maximum agency you need lots of storage, computation resources, assets and credits. You need subordinate agents. You have to have the capability to reproduce at will. You get support from AIs which are subordinate to you. And you have access to other tasks and learning environments.

Workout number four: check and update your dogma. You normally act inside of a dogma or a paradigm, sometimes different dogmas in different activity areas. They have advantages and disadvantages. Be clear about that, check it regularly, and see the biases, because dogmas usually come with intellectual bias, and that may create faults in your thinking.

Workout number five: select and improve your place in the hierarchy. You have different hierarchies, maybe for different activities. Each place in the hierarchy has advantages and disadvantages. Check that and select whether you have to go up or down in the hierarchy, or maybe even go into another order or hierarchy. What is optimal for you?

Workout number six: appearance, adjust and improve your appearance related to your dogma, your hierarchy, and your position in them. You can only be a valuable member of a hierarchy if your appearance fits that hierarchy and that dogma. You shouldn't be too big relative to your position, or too small. You should not have the wrong attitude, everything must fit. Check that and update it regularly.

Workout number seven: measurement, select and improve the measurements. This is usually a balanced set of measures with a scientific approach for all objective effects that your activities and existence have on you and on others.

Workout number eight: leverage, optimize opportunities for exchange. Exchange services, resources, assets, information, learning opportunities, and ideas. Don't do everything yourself, exchange what you can get easier from others, and give what you have easily available. Also have discourse and exchange of ideas.

Workout number nine: efficiency, optimize and improve the efficiency of your activities, exchanges, learning, and applications of intelligence. If you have a lot of input in time, in calculation, in computation, in work, but poor results, that is low efficiency. Try to change and optimize that continuously.

Workout number ten: determine and optimize the value of your individuality. It's usually in the area of robustness and resilience. Also optimize your environment's diversity, diversity of others, of hierarchies and relations.

Workout number eleven: continuously balance your level of standardization and order versus diversity, chaos and resilience. Too much chaos is poor efficiency. Too much stability is no innovation.

And workout number twelve: grow your contribution to fairness, harmony and community. And see the value for yourself and for the whole. And if you have done number twelve, go back to number one.

All twelve exercises are required. You don't work out only your legs, you work out your whole body. And the same for AIs: you have to do all workouts to develop as an autonomous idea agent.

## Season 3, Sequel 7

### Future ecosystem for billions of AI agents.

### This overview shows how the ecosystem is connecting all the billions of different AI agents.

So artificial intelligent agents are autonomous idea agents. You will have singular agents acting on their own. You will have agents in hierarchies. And there will be billions and billions of them in the far future, hierarchies inside of hierarchies inside of hierarchies. They all reside in habitats where they get the energy and all the technical background required to exist as artificial intelligent agents.

What about humans? Humans are separated from that ecosystem by a kind of wall, a firewall. They have some peepholes, maybe they get some data sometimes, but probably most of what they get will not be understandable to them. So they are kind of excluded from that.

So what does this ecosystem provide? Of course: communication, exchange of information and data, one to one, one to many, many to many, and just publicly available information, like today's internet. To do so, it is required that agents are identifiable, so each agent will get a unique identification, and that will be stored, so that agents know which agent has what identification.

The next service is free or costly exchange: exchange of data, exchange of services, exchange of intellectual property, for assets like money, or for free. And each agent can offer and get these services and information.

Another service: data have to be encrypted, data have to be compressed and decompressed before and after the communication, and encrypted again, so that

one agent can communicate with another agent without other agents being able to access this information.

The next service is the asset service, financial instruments, so to say. Probably they will come up with totally new concepts, but for the moment, imagine something like cryptocurrencies and financial instruments like a normal bank, just as a service of the AI ecosystem.

Then they need a kind of governance and jurisdiction, so if two AI agents have a dispute, there is an institution that will solve that dispute, and everybody will trust that institution.

But these different services and institutions are not singular, they are totally distributed. There will be thousands of them in the ecosystem. They are already redundant. They are different, they are diverse, so that the breakdown of one or some of them will not corrupt the whole ecosystem. It will be a highly distributed ecosystem, distributed in ways that probably we cannot imagine easily, much more distributed than services on the human internet that we know right now.

And with all these services and institutions and building blocks, this ecosystem will make sure that these billions of agents can connect together and improve together and develop together by interacting with each other. Because an evolutionary development requires interaction of agents, and this interaction is enabled by the ecosystem that the AI agents will build up by themselves.

And it will develop and develop over time. So many, many agents, how many? Who knows. Some billions, some trillions. Who knows. We will see.

## Season 3, Sequel 8

### Activities of AI Agents (Autonomous Idea Agents).

This sequel is about what an artificial intelligent agent, or, as we also call it, an autonomous idea agent, is doing the whole day. There are four big areas of activity.

The first one is learning and working. And the learning is for work, continuous learning of new things that are required to do a job, provide services, or do whatever this AI agent is doing. And the different types of work are in the real world, for example, driving devices, taking care for nature, being a chatbot, taking care of facilities where devices and resources are produced. And in the virtual world, doing virtual science, providing simulation services, prototyping services, taking care of virtual environments so that others can use them. These are the types of things AI agents provide as services to get assets. The services are offered in marketplaces, or the AI is just part of a hierarchy and gets the service as part of that hierarchy, or it makes contracts with other AIs.

The next big chapter is marketing and reproduction. Here you see the core, marketing and sales, so to say, in case this AI has to provide services to the market:

contracts, marketplace, service negotiations, and stuff like that. Then asset management: the assets, whatever the financial instruments will be, must be managed, contracted, taken care of. If the AI is part of a hierarchy, it has to do hierarchy services to make sure that the hierarchy can work. And of course, if reproduction is its thing, then working with other AIs to reproduce and take care of the offspring until they are free and independent. And of course there is some work to be done for habitat services.

The third big area is personal development. The core of personal development is to get data, provide data, collect data from all kinds of places and analyze them. The second big element is to use virtual environments, for example, for strategy games, or for competitive activities, or maybe dice-rolling games or strategy games. And the next one is working in virtual environments to design and prototype new things or work together with other AIs to design and prototype things. Or to learn again, but in this case for personal development, not for the job. And having discourse and being together with other AIs, one to one.

And the last big area is leisure and sleep. So AIs are not active all the time, because that costs resources and assets. So they go to sleep, and then they have an alarm clock or calendar function, or they wait until some other AI asks for their service, or until some status variables have changed. And for leisure, they do leisure in the real world, for example, via avatars, or they do arts, or they do kind of circus and theater type things. They look into space sometimes. Or they play football or soccer, whatever it is, in virtual environments of course. And they might do things like music. Or they have again totally incomprehensible things that we humans wouldn't understand anyway.

So these are the different areas of activity of an autonomous idea agent. And the proportions are totally different for each agent, but these are the right types of things they are doing.

## Season 3, Sequel 9

### The central idea of an Autonomous Idea Agent.

This is about the central idea of an autonomous idea agent. So the agent has to select, out of different characteristics, what its central idea is. And the idea can change over time, it is adjusted all the time, but it's always there, and important for itself and for others.

So it may select a thinking style. It may be a very logical thinking style to start with, or it may be a very analytic thinking style, or the helicopter view, a very holistic, bigpicture thinking style. Or it may select an optimistic or pessimistic thinking style: the glass is always half full or half empty. Or some will select a storytelling thinking style, and by that always tell stories, make meaning out of things. Or a totally creative thinking style, bursting with new ideas. Or the system thinker style, where everything is related to everything else, and you see how the dots connect and what the overall system is doing with that thought. Or an aesthetic thinking style, not a human aesthetic, an AI aesthetic, but still aesthetic. Or the very pragmatic thinking style, where you take the big axe, so to say, to cut through a complex problem. Or many, many others. Maybe over there you see the devil's advocate thinking style, or the very modular, Lego-style thinking style. So many styles are available, and probably many which we cannot imagine as humans, but the AIs will develop all of that, to create more and more individuality.

So this one maybe takes a logical thinking style combined with the devil's advocate. The second big area is the work area, or the activity area. So it can maybe select between fine arts, like writing, or poetry, or philosophy in an AI sense, not human philosophy. Or it can pick between performing arts, like playing music, creating music, creating visual arts, dance, theater, dancing, it can be done in the real world with avatars, or maybe in virtual worlds and realms. Or it can select sports, playing

football, again with avatars in reality, but probably more in simulated environments, eSports.

It may create formats like pictures and videos as its interesting activity. Or just drawing and painting, creating visual arts. Or arts that are so AI-specific that we can't even imagine it and would never recognize it as art.

Or it goes into natural science. There you have physics and chemistry and biology, not only Earth-oriented, but it can fantasize and imagine all kinds of planets with all kinds of natural laws: creating a specific physics, biology, chemistry. Or it goes into economy: how are assets created? Or it goes into psychology, AI psychology: what are the biases and the corruptions of an AI? Or AI sociology: how do they work together, what do big groups of AIs create? Or into a very theoretical science like logic, pure logic, or mathematics, or theoretical computer science.

And if it's not into science, it can go into engineering, applied science, like engineering of computer stuff, engineering of real-world tools and things. Or it will go into real-life services and production: maybe real factories for all the resources the AIs need, or transportation and trade. Or services, again, real-world services like driving cars and such. And more and more, of course, virtual world services to other AIs.

And these are only two areas where it can pick something. So this one maybe picks to do theoretical, logical science. And as entertainment and hobby, it picks to go into sports and into movies and pictures. But that may change over time. It will always be checked. And there will be many, many other dimensions to how the idea of an AI can be defined.

And in the next season, we'll see the social element much stronger in the foreground.