Prototype: github.com/cirosantilli/Urho3D-cheat

Prior art research: github.com/cirosantilli/awesome-reinforcement-learning-games

Less good discrete prototype: github.com/cirosantilli/rl-game-2d-grid YouTube demo: Video 1. "Top Down 2D Continuous Game with Urho3D C++ SDL and Box2D for Reinforcement learning by Ciro Santilli (2018)".

The goal of this project is to reach artificial general intelligence.

A few initiatives have created reasonable sets of robotics-like games for the purposes of AI development, most notably: OpenAI and DeepMind.

However, all projects so far have only created sets of unrelated games, or worse: focused on closed games designed for humans!

What is really needed is to create a single cohesive game world, designed specifically for this purpose, and with a very large number of game mechanics.

Notably, by "game mechanic" is meant "a magic aspect of the game world, which cannot be explained by object's location and inertia alone" in order to test the the missing link between continuous and discrete AI.

Much in the spirit of gvgai, we have to do the following loop:

The question then becomes: do we have enough computational power to simulation a game worlds that is analogous enough to the real world, so that our AI algorithms will also apply to the real world?

To reduce computation requirements, it is better to focus on a 2D world at first. Such world with the right mechanics can break any AI, while still being faster to simulate than a 3D world.

The initial prototype uses the Urho3D open source game engine, and that is a reasonable project, but a raw Simple DirectMedia Layer + Box2D + OpenGL solution from scratch would be faster to develop for this use case, since Urho3D has a lot of human-gaming features that are not needed, and because 2019 Urho3D lead developers disagree with the China censored keyword attack.

Simulations such as these can be viewed as a form of synthetic data generation procedure, where the goal is to use computer worlds to reduce the costs of experiments and to improve reproducibility.

Ciro has always had a feeling that AI research in the 2020's is too unambitious. How many teams are actually aiming for AGI? When he then read Superintelligence by Nick Bostrom (2014) it said the same. AGI research has become a taboo in the early 21st century.

Related projects:

Related ideas:

Bibliograpy:

A keygen, short for "key generator," is a type of software used to generate product keys or license keys for software applications. These keys are often required for activating legitimate software or games, allowing users to access full features and functionalities. Keygens are typically associated with software piracy, as they can be used to bypass the legal purchase process of software. Using keygens is illegal and unethical, as they infringe on copyright laws and the rights of software developers.

TODO vs COCO dataset.

As of v7:

The images and annotations are both under CC BY, with Google as the copyright holder.

It good to think about how Euclid's postulates look like in the real projective plane:

Unlike the real projective line which is homotopic to the circle, the real projective plane is not homotopic to the sphere.

The topological difference bewteen the sphere and the real projective space is that for the sphere all those points in the x-y circle are identified to a single point.

One more generalized argument of this is the classification of closed surfaces, in which the real projective plane is a sphere with a hole cut and one Möbius strip glued in.

Ciro Santilli's preferred visualization of the real projective plane is a small variant of the standard "lines through origin in ${\mathbb{R}}^3$".

Take a open half sphere e.g. a sphere but only the points with $z>0$.

Each point in the half sphere identifies a unique line through the origin.

Then, the only lines missing are the lines in the x-y plane itself.

For those sphere points in the circle on the x-y plane, you should think of them as magic poins that are identified with the corresponding antipodal point, also on the x-y, but on the other side of the origin. So basically you you can teleport from one of those to the other side, and you are still in the same point.

Ciro likes this model because then all the magic is confined just to the $z=0$ part of the model, and everything else looks exactly like the sphere.

It is useful to contrast this with the sphere itself. In the sphere, all points in the circle $z=0$ are the same point. But this is not the case for the projective plane. You cannot instantly go to any other point on the $z=0$ by just moving a little bit, you have to walk around that circle.

Robert d'Escourt Atkinson (1900–1960) was a prominent American psychologist known for his work in the field of educational psychology and his contributions to the development of several psychological theories and models. He is perhaps most recognized for his role in developing the "Atkinson-Shiffrin model" of memory, along with Richard Shiffrin.

As of my last update in October 2021, there is no widely recognized or notable person, concept, or term known as "Todd Brun." It's possible that it could refer to a specific individual, a character in a story, or a term that has gained relevance after that date.

Werner Brandt refers to a former prominent German businessman known for his role in the corporate world, particularly in technology and automotive sectors. Notably, he served as the Chief Financial Officer (CFO) of the multinational company SAP SE from 2008 until 2014. Brandt was involved in several strategic developments and financial operations during his tenure at SAP, which is recognized as one of the largest software companies globally, specializing in enterprise software.

Zhong Lin Wang is a prominent scientist and researcher known for his work in the areas of nanotechnology, materials science, and energy harvesting. He has made significant contributions to the development of nanogenerators and triboelectric devices, which convert mechanical energy into electrical energy using the triboelectric effect. His research has applications in various fields, including renewable energy, flexible electronics, and sensors.

"Comet Man" is a superhero character from American comic books, primarily published by the now-defunct publisher Atlas Comics in the 1970s. Created by writer Roy Thomas and artist Frank Brunner, Comet Man first appeared in "Comet Man #1" in 1976. The character, whose real name is John Ray, gains superhuman abilities after exposure to a comet's radiation. These powers often include flight, energy manipulation, and enhanced strength.

Functional analysis is a branch of mathematical analysis that deals with function spaces and the study of linear operators acting on these spaces. It is a subfield of both mathematics and applied mathematics and is particularly important in areas such as differential equations, quantum mechanics, and optimization.