Press "Enter" to skip to content

Intellectual computation towards an automatized design system

Intellectual Computation

Towards an automatized design system

Contents

Haven't found the right essay?
Get an expert to write you the one you need
GET YOUR PAPER NOW

1. Abstract………………………………………………………………….…………………………………………………..01

2. The intellectual machine ………………………………………………………………………….…… 02

3. Paramount Intelligences………………………………………………………………………………………………….. 03

4. Architecture decrypted?……………………………………………………………………………………………….. 04

5. Beyond conjectures ……………………………………………………………………………………….…. 05

6. Emergent Proto-relationships………………………………………………………………………………………. 06

7. An automatized design system………………………………………………………………………………… 07

1. Abstract

There exists an underlying cipher in Alan Turing’s question “Can machines think?” which must be elucidated., since what had laid grounds for the necessity for artificial computation has infinitely more to ask of us than number crunching, encryption, decryption or statistical analysis.

What can be observed common in both Charles Babbage’s analytical engine (A.D.1837) and Turing’s Universal Turing Machine (A.D.1937) was their intention to arrive at general-purpose computing. Further, Gordon Pask’s cybernetic systems and Walter Grey’s turtles prove the possibility that machines can react to stimuli rather instantaneously. These mechanisms are more dynamic than the current brute force computation systems as they computed in real-time.

So then, why did the collective scientific society choose brute force over real-time? What caused this great fall? Is it simply because it is easier and convenient to organize around? Or is it the unspoken reality that we do not want machines to become us, or even be like us? Technology may never become a cure for human fallibility, but It is apparent that we want machines to augment our reality. However, we confine it from becoming an innate part of it. Why have we let technology lead our way if we are not going to allow it to unfurl all its dimensions? Seemingly, the potentiality of machines may dwindle in the future due to the current social stigma that AI would eventually overpower humanity.

Now, the scientific society, on one hand, is determined to continually make computers faster and smarter while on the other, the pioneers in machine learning are attempting to rethink computational methodologies to arrive at more effective methods of processing, for example., the Deep Reinforcement Learning in Google’s DeepMind., specifically AlphaZero in which there are no rules. Ergo, in opposition to the rule-based computation in most of the current design methodologies such as generative design and parametricism, a conjecture must be derived to define a machine that has the potentiality to abstract relationships that may be beyond human comprehension. This may be constructed through interrogative speculation for a form of computation that is not only intelligent, but is intellectual., which can mutate generative design methods into an intrinsic, automatized all-purpose design system.

2. Paramount intelligences

a. Agents of war

The World War II, specifically between Germany and the Allied Nations bore ground to the war between the German Enigma machine and Alan Turing’s Bombe machine. The Enigma was an electro-mechanical rotor cipher machine engineered by Arthur Scherbius that was designed for the secure propagation of messages within the German armed forces. The settings for the machine were changed daily., thereby creating 1.59 X 10^20 possibilities per character.

The Bombe on the other hand, was originally designed by Polish cryptologist Marian Rejewski for the same reason. However, Turing disarmed the agency of Enigma by establishing that two rules that were common in all the messages., firstly, a character can never mean the same as itself, and secondly, the words ‘weather report’ and ‘Hail Hitler’ always repeated in the weather report that was transmitted every morning. It was these two rules that lost Germany the war by giving the Allied forces the opportunity for calculated response without unveiling that the Enigma was solved. Here, the rules were not unraveled through computation but by human observation and the delay till the later occurred had already costed millions of deaths. It must be noted from the discrepancies of both the machines that the Enigma could have been victorious if only had it been rule-free., or at least free of those two rules and the Bombe could have succeeded effectively earlier if it were electro-mechanically augmented such that the rules could have been computationally deciphered.

Other essay:   An iot based lpg gas sensing and alerting system

b. Agency in interoperation

The cybernetic theater was an experimental system designed by the prolific cybernetician Gordon Pask in collaboration with the legendary theatre writer and producer Joan Littlewood. The audience and actors, in this ontological stage, could be coupled together via feedback loops where although certain set-piece structural elements of the play are specified in advance and hence retained, alternate routes of plot development are entertained between the set pieces. The audience could signal their identification with one of the principal actors and use levers to signal to their chosen character to perform different choices of action. The choices of the audience would naturally be based on their understanding of the plot development and the meta-information of the character’s thinking as per the perspective of the former. The interpreters would communicate the supporters’ inclinations to the actors and accordingly, the play would proceed. Andrew Pickering, a notable historian of science points towards Pask’s acknowledgement in The cybernetic brain: Sketches of Another Future that the removal of arbitrary restrictions which hamper the effectiveness of a control system, i.e., the feedback from the audience, new possibilities for a novel art form can be synthesized through the monitoring of the feedback and feedforward loops.

Further asserting Pask’s assessment, he argues that since the audience participated and were directly confronted with, the play can be attributed as a “performative becoming of human affairs.” It is remarkable here, that albeit the alternative piece has already been rehearsed, it is delivered to the audience in real-time. The play is now elevated to be an open-ended performative engagement, thereby becoming an aesthetically potential environment for both the audience and the actors alike or unlike. The algorithmic order in such intrinsically dynamic feedback loops of multiple orders, as per Pask’s observations indicate the possibility for a computational method of control for performative spaces and therefore help in devising the same.

3. The Intellectual Machine

Intelligence in living beings, according to American psychologist and writer Linda S. Gottfredson is the mental capacity to reason, plan, solve problems, think abstractly, comprehend complex ideas, learn quickly and also learn from experience.” One can also argue that it is the response to stimuli in a self-beneficial or mutually-beneficial manner and is therefore essentially a part of nature that evolves with linear iteration. A machine intelligence with the full range of human intelligence., commonly called Artificial General Intelligence or strong AI should therefore reason, represent knowledge, plan, learn, communicate in natural language and thereby integrate all these skills towards common goals. However, David Poole hypothesized in Computational Intelligence and Knowledge in 1998, that “reasoning is computation.” The amalgamation of the two hypotheses therefore extends the question “Can computers really think?” into one that may hold a solution that can morph the essence of artificial intelligence., i.e., should a thinking machine necessarily be human-like to be able to reason?

Other essay:   Kalman filter based air defense system

c. AlphaZero, the current benchmark of AI

DeepMind, a revolutionary AI company is exhaustively researching for smarter computation methods and determined to solve intelligence to apply the technology for real-world impact, currently in health, science and energy. They devised a deep reinforcement learning system named AlphaGo which is a reinforcement learning mechanism connected to a deep neural network that operates directly on RGB images and efficiently processes training data by using stochastic gradient updates. AlphaGo defeated the world champion in March 2016 in the complex board game of GO and had set a benchmark for AI. While being commemorated, Demis Hassabis, the company’s CEO remarked on how creativity in a different dimension to what humans can do could be of immense value to us in the numerous possible application domains.

AlphaGo further mutated to AlphaZero which is a more elegant system that, as the name suggests, is only infinitesimally aware of human knowledge and starts learning from null. This proved to be exponentially effective as it defeated Stockfish and IBM’s Deep Blue in 9 hours of learning period. While the latter attempts to account for every eventuality in the game by relying on thousands of rules and heuristics handcrafted by strong human players, AlphaZero takes a completely different approach replacing the above method with a deep neural network and general-purpose algorithms that know nothing about the game beyond the basic rules of moves and therefore, learns through experience. In other words, a profoundly creative computational system has come to exist, which now allows us to ask a quintessential question.”,

Should we program the inherent nature of a machine in the choices it makes instead of programming its behavior?

A non-anthropocentric comparison between the intelligence in beings and machines observed along the evolution of the potentiality of machine intelligence from cybernetics to artificial intelligence highlights a profound contrast between an intelligent machine and that which can be called an intellectual machine. If relationships are to be considered as rules, then, an intelligence machine computes data-sets according to the rules or relationships it is pre-programmed and therefore indemnified to follow., much like AlphaGo. In clear contrast, an intellectual machine intrinsically computes the relationships both between and within the data-sets., as observed in AlphaZero. Nick Bostrom, a famous philosopher at the university of Oxford, in his paper Policy Desiderata for Superintelligent AI: A Vector Field Approach, suggests that such a superintelligent machine would observe and abstract general properties that are common in all the data-sets and their inter-operations.

Other essay:   The war of south african talfeberg

These general properties or rules that led the Bombe hail over the Enigma would however have the opposite effect in the real world which, in a way is like chess., where intelligent agents that are confined within their perceptual limitations and finite computation, perform what is appropriate for its goal within its circumstances while being flexible towards change in both.

4. Emergent Proto-relationships

A non-anthropocentric designer and researcher, William Fairbrother, argues that agency emerges through the interaction of bodies and is never a subset to a singular thing.

An alternate precursor to Generative design

Generative design, according to the principal research scientist of Autodesk, Michael Bergin., is a method where the higher-level goals and constraints of a project are specified first and then use computation power to automatically generate, evaluate and evolve multiple design options to explore a broader design space of possibilities. In an interview with dwell magazine, he commented “Architects are the conductors of the design process—they provide something that is very difficult to systematize… the really good stuff will always need human guidance.”

In the The decisions were made based on human values and the potentiality of machine intelligence was limited to the computation of the various iterations possible within the given set of constraints.

Generative design in the current dimensions of design application may appear sufficiently comprehensive for the evolving present but, it is a non-anticipatory design system that can only follow obsolete rules in the context of designing for the continually fluctuating future. In such a case, what kind of computational algorithm is suitable for a world in constant flux?

Generative design has more to learn from information theory.

Just as AlphaGo evolved into AlphaZero, generative design can evolve into a fully automatized design system by removing the most primal inhibitor., i.e., human knowledge or rather the specification that intelligence must be human-like.

Use the algorithm in AlphaZero for generative design in architecture!!!!!!!

Emotive city

Maybe generative design is enough for the evolving present., an anticipatory design system can only follow obsolete rules., then, what kind of algorithm is suitable for a world in constant flux?

If the ability to understand something instinctively, without the need for conscious reasoning is a plausible description of intuition, can we program intuition?

Deep reinforcement learning- Chess and the world are not similar games., because unlike the former, the rules are not fixed in the real world., The machine must therefore respond in real-time not with an anticipatory response but rather with an emotive response.

5. An automatized design system

Towards a general-purpose algorithm that responds to the construct of society.

Design processes should seldom be need-based isolated developments but should encompass the proto-systemic. And therefore, to be capacitated to design without restraint to design for the future, it is paramount that such relationships must be intellectually computed., that may be unperceivable to humans., but have the agency to create more comprehensive methods for ‘constructing continual utility’ through a higher order systemic and spatial communication.

Be First to Comment

Leave a Reply

Your email address will not be published.

0 Shares
Share via
Copy link

Spelling error report

The following text will be sent to our editors: