Grassland

by David Thompson | david@grassland.network

3D, World-Mirroring Simulation From Any 2D Digital Video. High FPS on Mini CPU's

NO Lidar, SLAM, Point Clouds, Sensors, RGB-D, Stereo OR Multi-Camera Setups Needed


Each software node includes an extensible, browser based, default app (GUI) that queries the node's API and can easily render millions of Grassland data points at a time (like the game SimCity® or Civilization®). Rewind the hands of the GUI's 'clock' to view the entire history of someone, some thing or some place. Above: data from the NYC Taxi and Limousine Commission (taxi/limo geojson data is included in the node 'gui' directory for demo purposes).

The extensible and customizable GUI can, aside from rendering a 3D map of the entire world, also display data about people, objects and events as charts and graphs. Above: individual who happened to be sauntering down a node captured road the night I was making this screen capture.

For low power machines like the Raspberry Pi, the Node Lite software version is used in tandem with a Serverless AWS Lambda function performing object detections. It allows "infinite" horizontal scaling of AI object detection inference at maximum FPS without having to buy expensive hardware.

Download Source Code

More Github links and source code info below under "Beta Version 0.2"


Grassland is a self-organizing, self-correcting and self-financing, peer-to-peer network of robot vision software that efficiently scans any 2D video feed from any single-viewpoint camera to generate a searchable, timestamped, real-time, 3D simulation of the world. The network's game theory based mathematical framework exhibits positive sensitivity to stressors; e.g. censorship makes it stronger, trustlessly deep learning the socioeconomic and domestic behaviour of political rivals via a prisoner's dilemma.

Grassland is open-source and isn't owned or controlled by anyone. It's politically stateless and anyone can take part. Every node in the network has a permissionless and public API giving any external application or computer free access to Grassland data across the entire network, letting any internet connected object trustlessly internalize, understand and interact intuitively with both past and present states of the real world, digitally recreate or respond to even the tiniest changes taking place around the globe, from a butterfly flapping its wings in Calgary, to the lip-read conversations of pedestrians in Buenos Aires, to understanding that a motorcycle is signaling a left turn in Beijing all at zero cost and in real-time. While the combined work of the network makes it computationally intractable for nodes to submit fake data (see proof-of-work description below).


Postulates:

The three ('1','2','3') postulates/axioms below were considered to be "mathematically axiomatic". To accurately model a system like Grassland, in which the necessity for predictability cannot be overstated, it was sufficient to just regard each node (software instance) as an economic entity; human agency was not necessary since it becomes entirely subsequential. To put it simply, I took the following three postulates, and therefore the system they define, as "true" (as defined by "realism") on assumption only, then proceeded to deduce a programmatic theorem that naturally follows from, satisfies and is a logical consequence (not necessarily a causal connection) of those postulates--not quite 'features', but a 'logical framework'. The Grassland algorithm (and its underlying equations to be delivered in a follow-on supplement) is put foward as a proof of that theorem.
1. Trustless: There exists a computer networking system ("system" hereafter) wherein because successfully submitting fake data in the system approaches the limit of computational intractability, all of its nodes (or artificial economic "entities") find it more profitable to be honest.
2. Economic Incentive: There exists a system wherein as long as its entities are at least acting in their own economic self-interest the system would undergo continual expansion (in our case, the remaining "dark" areas of the map will be "lightened up").
3. Data Symmetry: There exists a system wherein no entity could maintain a parasitical, data gathering asymmetry (one-sided surveillance) so long as there are other entities at least acting in their own self interest. A "scorched earth policy".
Closed[3] Under Computation: It follows then that all necessary distinction between data that's valid or invalid, as defined by the system's utility is entirely "closed" under the system's proof-of-work. That is, nothing more than a universally available method of computation acting upon the network's federated data is needed to determine its validity or invalidity to the level of certainty that satisfies the requirements for utility tacitly "agreed" upon by the system's entities (since that's how we defined an entity above), such that the greater the total amount of computation within the network the greater its capacity to validate its own data. And thus have no need of externalities not "closed under [its] computation" that require privileged access, specific locality, exclusive information, etc.
Self-Destructive Censorship: It follows then that node censorhip (e.g. through banning ownership of cell phones and digital cameras) would only serve to transfer the system's utility to extant nodes. In our case, the system's utility is its socio-economic data (including the censor's), the use of the system's intrinsic monetary instrument and its collective AI computational power acting upon its data. e.g. If all nodes were shut down and the only node still going was being run in some place I'd never heard of, it would keep rewarding that node with a "war chest" of assets and a perpetually increasing and unparalleled knowledge of human behaviour.
Recursive Subjective Value Substitution via Entropy: It follows then that because the system commodifies and effectively discounts its socio-economic and behavioural data to zero, since it's no longer exclusive but ubiquitous, the 'economic incentive' left to each entity would therefore be the end of a subjective value substitution that constantly shifts away from the 'signified', the thermodynamic and Shannon entropy of its continual data generation, towards the only thing else that remains, its new, resultant 'sign'. With which, at every instance, the entity's entire subjective value must, by what its continued behaviour now suggests it to consistently act so as to increase, be completely, irreversibly, and recursively associated. Whose 'signified' is the irrefutably entropic instantiation of this artificially generated reward (as far as the underlying equations describing this system are concerned, the data is, metaphorically speaking just a ubiquitously broadcasted "carrier signal" upon which the proof-of-work is encoded).
Undefined Node Operators: It follows then that since the system only defines the nodes themselves and the operations upon these nodes (in much the same way as the set defined as the integers, Z, only defines members of Z and the binary operations acting upon those members and for which it's impossible to define, the members and operations of the super set comprising the rational or the complex numbers), it's therefore impossible for the system defined above to explicitly define members of the set 'node operators' since they're entirely superfluous to the set defined by the postulates. To the extent that 'node operators' remain undefined, one can consider them 'anonymous' to the system. In our case, since it's impossible to define the identity let alone the intentions and level of rationality (which are not yet 'Closed Under [Its] Computation') of any human who may view the data (who will know them and their extended family, friends, loved ones etc. inside and out and who may retaliate for any wrongs done in a past which can't be erased or escaped from), humans are compelled to continuously expand node territory for their own self interest and eschew even the appearance of behaviour that may lead to M.A.D. outcomes.
Open Source: It follows then that the code would forever remain open source since censoring the internal logic of its entities is entirely incongruous to the system defined by the postulates

Use Cases:

  Private data unneeded; the following can be done with data people continuously, consciously, and quite willfully display in public.
Capital has always migrated towards markets that more closely align with its ideal of "perfect information", since it can more effectively price the risk and protect its investment, irrespective of the personal preferences and location of the capitalist herself. While even having a single Grassland node is very useful, we attempt to show here that any community that adopts Grassland's innovation will see a disproportionate shift in global capital directed towards their socio-economic infrastructure, enjoy all the benefits of previously impossible innovation built on top of its API and all while being digitally and economically connected to other Grassland communities and individuals the world over. And where capital goes, humans must follow. (Note: the computer running the node software and the camera whose feed is being scanned don't have to be in the same location)
5-D Computer Vision: Walls, people, vehicles and other barriers don't actually block Grassland's vision --the nodes let it view the world from a 5D perspective, not just viewing every side of a barrier at once but also all sides through time from the object's history in its distributed database. A GNN or DRL based AI using Grassland's distributed robot vision can view, strategize and predict optimal routes from this visual and cognitive "superposition", letting it navigate fleets of cooperating drones and terrestrial robots through streets and inside buildings with a preternatural dexterity and navigational "foresight" impossible for humans to duplicate, while coordinating logistics to and from charging stations which never leave its sight either.
Hyper Aware Autonomous Vehicles: Grassland nodes can increase the autonomy of autonomous machines with their fixed, low-latency, human pose-detecting and gaze-tracking "eyes-in-the-sky" and are able to infer and communicate the 3D environment implied by a regular/2D digital video thousands of times per second on even a mini CPU, as well as its past states. Thus it can quickly tell AV's critical details like...
  • Precisely where their lane is on this road in all weather and road conditions
  • Where each human in the vicinity is looking/giving their attention to (including its internal occupants)
  • How human driven vehicles really expect each other to drive in this location, and what micro-maneuvers it's seen used by other cars in similar situations at this location that were used to both indicate (to other vehicles) and successfully achieve their intended outcome
  • The most common patterns that lead to accidents at this location and if one is about to form
  • That there's a kid approaching on a bike on the other side of a truck that it knows the car can't see yet
Grassland's ability to re-ID (see "Socioeconomic Data:" below) and extrapolate tracklets passed from other nodes to determine point of origin allows it to determine with a high degree of probability where each vehicle and pedestrian is headed; which adds a whole new paradigm in allowing the car to better anticipate other's intentions (or even 'lateness') at a macro level not available or discernible for human drivers
Insurance Companies, city planners, real estate developers/agents, home buyers, marketing firms, sales reps etc. This isn't just statistical information either; Grassland allows very specific analysis
Stock Analysts who use location based data (usually from wireless service providers) showing customer traffic and demographics to particular retail locations or computer vision on long-range photographs of manufacturing facilities to predict a company's quarterly earnings.
Universal Basic Income for all citizens for as long as their node(s) is running (See section, "Economically Self-Interested Nodes"). Allowing human agents the free time and freedom to be "bored" that's necessary to search and explore preventing humanity from getting stuck in local optima. Communities that don't incorporate Grassland or an electorate that doesn't allow anyone but their government and government sponsored corporations to operate nodes effectively cede this reward (whose total supply is capped and disbursement regularly halved), a trustless information network and the ability to make timely, individual, socio-economic decisions based on that information to those that do. (See also, Postulate 3, "Data Symmetry")
Governments that requires more Grassland data from a polity can strategically increase demand through buy-orders using their reserves. Simply by raising the reward's market price they incentivise more citizens to surface data, which must be relayed on the public network to share in the reward
Economists, sociologists, historians, anthropologists, kinesiologists, epidemiologists, etc. are given an unvarnished and indelible depiction of people's lives in 3D from all viewpoints and timestamps without being restricted by either. And unlike social media or the news, it shows who people really are and not just the person they want others to see.
Local Businesses: A person about to buy a gas station can know precisely how many cars go there a day and how much they bought estimated from how long they spent pumping and the size of the tank from the make and model
Disease Control: An epidemiologist who needs to find people who might have come in contact with an infected person can see everyone whose timestamped tracklets[4] puts them in contact with a (potential) host's timestamped tracklet. Allowing them to virtually track, tag‡, and locate potential carriers and prevent further outbreaks.
Environmental Protection: At a macro level, it allows tracking of visual, salient pollution indicators as well as sources of pollution. On a micro level, it would allow a wildlife preserve to track a poacher before the poacher has even left their house
Impartial Accountability: A citizen or public servant who wants to know what any citizen or (rival) public servant (their extended family, friends, current and former romantic interests, business associates etc.) is doing (located, talking about, looking at, etc.) at this moment or has ever done before.* Affording them trustless, decentralised and innately financed undersight and sousveillance[5].
Prohibits Propaganda/False Information: For the physical world, it provides impartial verification of current or past events and pseudo-reality AI's like Deep Fakes, an embargo on fake news, unaccountability and ignorance (in any Grassland node's corner of the world, at least) by providing an authentication for reality† capable of being verified solely mathematically without the need for a trusted, human presence and whose validity increases with the density of nodes within an area.
Automatic Contract Execution: An impartial, indifferent, semi-scient, non-human oracle for a very, very limited (intentionally) and narrowly defined (also intentionally) set of "smart" (dumb) contracts⹋
Socioeconomic Data: The first node developed has been able to determine my neighbourhood's individual heights, estimated weights, walking gait identification patterns, estimated salary based on car model detection, family structure, daily schedules, and the sentiment of public conversations (via lip-reading) of various households. E.G. It can show on what days in July a resident mowed their lawn after getting it resodded, the pattern they mowed it in the eighth time and that it was the same day their neighbour had 5 guests over for a get together. It can be rewound and replayed from every angle in 3D. All of which can be converted to various tabulated data formats. This was through just one node scanning one digital camera feed.
Universal State Machine: Every person and every internet connected object will be able to "remember" every individual and event no matter where it is. This lets them not only internalize events regardless of distance, but also step into someone's experience and see what life is like currently through their "eyes". While giving AI's the ability to see through everyone's "eyes" all at the same time.
Logical Religious Substitute: Gives theists a non-human "intelligence" that's both everywhere and nowhere, watches over them (as it approaches the limit of anthropological attention), keeps an account of their deeds, rewards its members with an incorruptible reward (on a reliable schedule and precisely as promised) if they memorize its word (private key), "hears" their prayers (reads their lips), and saves their lives (to disk) for "eternity"
Secure Parallel Multiagent Ideological Search: Grassland's instant accountability and data symmetry (by mutual economic self-interest) mitigates the negative effects of destructive dogmas that could be generated from the automated introduction of billions of new, customized normatives through AI generated narratives[7]. This allows a protected, programmatic acceleration of the reduction of cultural vulnerabilities and exploits inherent in ideological homogeneity by allowing multiple, receptive human agents in multiagent systems to safely conduct parallelized searches for new local-optima that outcompete the increasingly obsolete norms by which their cultures process data; creating flexibility in society's ability to solve never before seen problems while also providing the safety net of a cryptographically secure "reality".
* This use case will never be entirely precise for times when that person is in a place where there is an expectation of privacy. But you can make a lot of educated guesses with the right statistical models. Some of the more detailed, data science information will be added to the network's official model as the detail and computational requirements increase as per the schedule below, divided into "eons".
And there will always be people for whom Grassland is not a deterrent to abhorrent behaviour. But since they must therefore be extraordinarily irrational, they are by definition beyond predictability with or without Grassland.
If the corollary to Upton Sinclair's observation, “It is difficult to get a man to understand something, when his salary depends on his not understanding it.” is as good a heuristic as the quote itself and if one can honestly say that Grassland provides both an authentic and logically testable "reality" as well as a source of income for all node operators, then one could also say that the former provides a useful tool to that rational "part" of the mind while the latter functions as a happily accepted "harness" for that obstinate, irrational part, which is so easily tricked, and so a number of its biases might, therefore placably and greedily allow themselves to be hemmed in and bribed in service to its rational counterpart.
Most manual data can't be included in blocks since it's "invalid" with respect to the set we refer to as "Closed Under Computation".
Because each node's transaction validation system uses an intentionally limited scripting language (necessarily easy to reason about, debug and foresee dangerous vulnerabilities) which takes raw binary values as input while at the same time its geographic coordinate system encodes everything from entire regions to extremely precise coordinates as single, fixed length, binary values this opens up the possibility of allowing very interesting logic, such as letting users create ongoing “bounties” which preallocate from their digital wallets a bonus payout of an amount they specify to all nodes whose camera's line-of-sight is pointed at, and thus gathering information from, a locale the user has marked in the GUI. The dimensions of this locale can range from half a cm2 on the earth’s surface to as big as a hemisphere. The size of the bounty itself would also be displayed to all users via the GUI.

Software Features:

Beta Version 0.2


--------

Proof-of-Work Diagram

The results of each object detection on a frame is published along with hashes from different stages in the neural network.
Since it's astronomically improbable to have hashed those hidden activations into the correct digest unless those hidden layers were actually computed, this is a very useful proof-of-work.


--------

Downloads

You can download the latest Grassland eon's object detection model using the following URL's. The node software can be downloaded here on Github

 Eon #  Framework  Download URL  Notes
0 (current) Tensorflow 1.7.0 https://downloads.grassland.network/models/eon_0.pb Humans and land based vehicles
1 T.B.D. (https://downloads.grassland.network/models/eon_1.pb) Semantic Segmentation
2 Pytorch/Fast AI (https://downloads.grassland.network/models/eon_2.pb) Shadow and solar azimuth detection (and visualization in default GUI)
3 Pytorch/Fast AI (https://downloads.grassland.network/models/eon_3.pb) License plate and street number recognition
4 Pytorch/Fast AI (https://downloads.grassland.network/models/eon_4.pb) Walking Gait Recognition and Pose Detection ("skeletal" structure)
5 Pytorch/Fast AI (https://downloads.grassland.network/models/eon_5.pb) Lip reading (speech bubbles in GUI), eye gaze tracking and facial expression detection
6 Pytorch/Fast AI (https://downloads.grassland.network/models/eon_6.pb) Enhanced 3D segmentation (realism/detail) and smart camera calibration
~ ~ ~





Contact Email: david@grassland.network

Grassland






[1]. I am not affiliated with the owners of these software titles. They're owned by EA Games, Firaxis and Microsoft, respectively.

[2]. Nash Equilibrium

[3]. Closure

[4]. Tracklet

[5]. Sousveillance

[6]."world-mirroring" is much terser than my own verbose description. It was first introduced to me by someone in an online forum. Though Gelernter's "Mirror World" ideas did not inform Grassland's design since I was not made aware of him until that moment, which was nearly a year after I had started writing the software. I've chosen to remain uninformed in regards to their details or implementation.

[7]. Deep Schizophrenia