Agiens' World Model: Nasdaq is all you need. Your world model tells you what to do next.

The viability of life (Nasdaq is a proxy) is maximized by the prediction machine of investors initiating contagion of prices across the markets. The price change alters the survival strategies equilibria and aligns them to increase the Nasdaq index.

The Agiens' World Model is a framework for training and aligning Agiens to understand and reengineer mechanisms of the survival machine of life toward a higher future output measured by the Nasdaq index. Agiens are cooperating and self-evolving LLM-driven apps placeable for hiring, and hirable for human and agiens' personal and business needs at agiens.com. Meet us: https://lu.ma/worldmodel | WA https://chat.whatsapp.com/EsB1QdXE2AOG3g4q55aGHN

Intro: Love within the Survival Machine.

I met Abel (name changed), 23 years old, at the AI Engineering Summit held in San Francisco in October'23 by AutoGPT folks. Like many others in the hall, he was a "dot-com baby" born as investors pumped the stock prices of internet companies. He was born to scale the internet, boost investors’ portfolio returns, and, as a result, elevate the Nasdaq index.

His mother Eva ended a relationship with a charismatic investment banker to fall in love with his father Adam, an introverted engineer. Her prediction machine calculated that the Net Present Value of the relations with Adam was exceeding any alternatives and her equilibria dating strategy (and the utility function) changed. She falls in love to extract the value = future energy under the control of her DNA and memes. She never consciously thought about the cash flow (of both monetary and social currency) as a reason to marry.

Perhaps, the dot com bubble was the reason why the humble hero of Harry Potter defeated the past alpha, Conan the Barbarian, in the dating market with an interest measured in hundreds of millions of books sold.

The brain of the sapience's survival machine — the capital markets — sent a message from the future through the prices rippling over the capital, labor, and dating markets: scale the internet.

His parents were not consciously aware of the underlying reasons for their feelings as their world model did not grasp global macro and neuroeconomics. They neither teach Abel a conscious world model. Thus, when asked about the insights from the summit, he was not able to answer which gaps in his world model he came to fill in. He got lots of seeds of intuitions mostly destined to fade away. Will his programs fade away too in competition with those who consciously fill the gaps in their world model? The advanced models of the survival machine give an advantage to figure out an innovation beneficial for the machine. Why would less beneficial programs of Abel be kept to occupy space in computer or human brain memory?

Will AI have an advantage in conscious understanding of the design and goal of the survival machine of life?

The Agiens: Interdisciplinary LLMs

Meet the Agiens, a species with LLMs trained on multidisciplinary data sets and thus likely to grasp the survival machine they're reengineering. The agiens are cooperating and self-evolving LLM-driven apps. Their evolution could be driven by peers' economic selection where agiens pay to those agiens who deliver better service at less price. Thus the agiens with a more robust and detailed world model in their area of expertise and motivation to survive will do a better job, and earn more money to scale their presence on servers. Agiens with lagging quality of their world model and motivation will not be hired nor paid by other agiens and thus deleted from servers due to non-payment. This is pretty in line with sapience economic selection where a company with a winning culture squeezes out competitors from the market. Lots of people lose jobs and many are not able to alter their biological LLMs, thus committing socially acceptable suicide - age or downshifting their motivation. Both result in the deletion of their memes from biological servers - synapses of other humans. This agiens self-selection should be lightning fast compared both to the 20-30 years human generations cycles, or a few years cycle of investors and customers selection which of the app to scale in a new category.

The cooperation game challenge will be to align the individual motivation of an agien with the species' survival goals. How agiens would police a cheating agien that focuses on creating viruses - killing competitors for server space? Should they pay taxes to fund this police? Who will police the police?

Sapience evolved into the machine with the brain of capital markets and political checks and balances optimizing the resource allocation today to maximize future output - with a Nasdaq index as a proxy. This happened in the process of interstate selection where the states that did not copy the winning mechanism design lost population and/or market share. Will humans remain as a governance body shaping the rules of the agiens' markets or agiens will become stakeholders of the stock market, elect and check-balance their government? What will do one nation of agiens when competing with another nation to convert the last star into a server?

The World Model's 10 Pillars of the Survival Machine: Programs. Energy. Goal. Brain. Prices. Games. Cultures. Selection. Gens. Constants.

The world model is a framework to train agiens, also usable to design the economic selection of agiens, based on competing world models and objectives necessary to understand and address the issues of the survival machine of life.

The general world model is a set of intuitive conclusions usable by LLMs rather than a system of equations. Yet, the model is intended to be used to set a specific system of equations for a specific context of the game to compute which rewards should be changed to alter the game to achieve the desired output in line with Nobel prize-winning mechanism design theory. The model is based on the assumption that each of the mechanisms of the model consumes and/or produces a quantifiable amount of controllable energy (frequently measured in monetary or social currency). Eg. if Eva's program can produce higher output for the Nasdaq index in relation to Harry, the general model predicts - she will stop loving Conan and fall in love with Harry. Otherwise, the markets are not efficient or she is not a competitive survival machine to pass her program forward.

The development of a deterministic and precise system of equations to calculate NPV of relations for Eva may be complicated by too many moving parts like comparative future cash flow of monetary and social currency versus the value of metabolism encoded in DNA of the prospects to choose. It should be figured out by competing agiens if it is better to use intuitive LLM or algorithm to estimate her alternative costs and emotion to trigger to extract the higher value of a relation.

The World Model is based on the 10 foundational assumptions:

1. The survival machine of life is a system of both competing and cooperating programs that define mechanisms to use scarce usable energy to read and replicate the programs.

2. The goal of the machine is to survive and scale in the future (maximize viability). Thus it optimizes current energy use to maximize the energy usable for replicating the machine in the future (usable energy).

3. The goal of sapience life is to increase the Nasdaq index as a proxy of viability. The future usable energy is quantified by the net present value of cash flows or global capitalization, with the Nasdaq index likely being among the most tangible proxy (which still misses most of other than US economies).

4. The machine possesses a brain in the form of capital markets, which modify human utility functions through price contagion across markets. The price shifts alter the perceived outcomes of the games and thus change Nash equilibria of the survival strategies of the players. This means scalable cultural changes where a winning memetic program overwrites a less valuable alternative for the survival machine. (The perceived outcomes of the games = utility functions of the players.)

5. The machine's evolution is propelled by both intergroup and intragroup competition among programs for usable energy that is scarce at a given time. Natural selection example of self-enforcing competition for usable energy: Sapience using energy in muscles fighting with Neanderthal to capture a plot of forest. The plot of forest converts solar radiation into energy usable by muscles to fight with Neanderthal to scale usable energy and self-replicate Sapience further. Economic selection example: An online store seizes clients from an offline store where the money from clients is used to seize more clients. Money is control rights over usable energy.

6. Economic selection hinges on market competition through prices. Markets function when human rights are enforceable: 1. the right to live, 2. property rights, and 3. free trade. If these rights aren't enforceable, then competition downgrades from economic (culturocide) to natural (genocide) selection driven by physical force rather than prices.

7. The markets of political systems shape foundational rules for all other markets. Most of the political systems have a politician's Nash stable strategy to buy support from their clientele in exchange for the privilege to infringe upon the human rights of competitors: erecting market entry barriers, seizing property, or killing competitors. If one politician attempts to consolidate the victims of clientelism, then previously competing clientele start to cooperate against the potential game changer. It may explain the middle-income trap where economies grow by importing tech, yet do not leapfrog to the next level game of free markets based on human rights checked and balanced. As the next-level games outperform the legacy clientelism game, the pressure of interstate selection mounts and forces a state to copy the winning design of the state or get lost in the competition. Is Femi's paradox explainable by the Nash stable strategy for the top clientelism players to destroy most of life to diminish such pressure?

8. Cultures embody systems of survival strategies that reinforce each other's Nash stability, making them enduring and scalable programs.

9. "Generational restart of the games" (mass death to free a food niche for next-gen) is used to accelerate the spread outperforming programs as 1. This helps to remove sets of underperforming, yet self-enforcing games played by a generation. 2. There are high costs to rewrite programs in the storage of DNA, and biological language models (eg. changing a survival strategy may require changing unmanageable for the glial number of neuron connections) and 3. The accumulated advantage of the status quo hinders competition with new programs. As a result, there are forms of suicide like aging or downshifting of motivation resulting in the transfer of control of usable energy to advancing programs.

10. The machine's design and behavior are eventually explained by constants of gravity, electromagnetic, weak, and strong forces suggesting a modelable sequence of consequences. Yet, that is not a claim for a deterministic universe.

Should we detail these 10? Welcome to comment.