1

Artificial Intelligence in Alien

Alien: Can an android reason, problem solve and be conscious?

Tyson Baker

578072

April 1, 2003

Psychology 253.3

Litke & Moorthy

Introduction

Artificial intelligence has been a topic of science fiction, with many examples ranging from advanced organisms of unknown origin to man-made robots and androids. In the current paper, the android known as "Ash" from Alien (Twentieth Century Fox, 1979) will be analyzed and compared to humans in terms of reasoning, problem solving, and consciousness. Ash is superior to humans in terms of reasoning; he is not swayed by emotions nor does he appear to rely heavily on heuristics. Ash demonstrates his ability to problem solve throughout the movie by continuously finding ways to transport the alien life form back to Earth for analysis, usually without arousing the suspicion of the fellow crew, who also are not aware he is an android. Ash appears conscious in his interactions with his environment; he shows ability to plan and execute ideas to alter his environment to suit his needs. Based on the behavior of Ash, it is therefore concluded that Ash is a being that is aware of itself and environment, and shows sufficient reasoning and problem solving abilities that are not only equal to that of humans, but possibly beyond the abilities of humans.

Source Summary

In the movie Alien (Twentieth Century Fox, 1979) a seven-person crew aboard a transportation spaceship awaken from their long sleep to find themselves in another star system. They expected to be near Earth, but their ship encountered a repeating message that may have been from intelligent life and automatically began seeking out the source of the message. The crew land on a planet and send three members out to investigate the source. The away team found another spaceship with eggs in it. One organism, known as a face-hugger, attached itself to a member of the crew. Ash allowed the crew member and the face-hugger into the ship against the orders of his commanding officer and the quarantine code. Ash spends much of his time studying the face-hugger and grows to admire it, with the imperative that it must be taken back to Earth for further examination. Eventually the face-hugger releases the crew member and dies; however, the face-hugger implanted a fetus into the crew member's torso. The fetus pushes its way out of the crew member's abdomen, killing him, and disappears into the ventilation duct. The crew unsuccessfully attempts to kill the alien, and eventually Ripley, the lead character, finds out that Ash is planning on getting the alien back to Earth; even if it kills the rest of the crew. Ash is then found to be an android and is demolished; eventually every crew member except Ripley is killed by the alien. Ripley manages to eventually kill the alien.

Comparison of Cognitive Functioning

Reasoning

There are two main forms of human reasoning: rational reasoning and irrational reasoning. Rational reasoning is the form that is based on valid inferences guided by systematic, dispassionate, sensible logic. Irrational reasoning does not distinguish between valid and invalid logic, nor does it set aside emotion, but rather bases its conclusions on heuristics, which are often contradictory and inaccurate. One might think that irrational reasoning would be not only useless, but also harmful, however, it is currently postulated that the current form of human reasoning is adaptive from a Darwinian perspective. Furthermore, however, it could also be postulated that we are still evolving and refining our natural reasoning abilities away from irrational reasoning towards rational reasoning.

In an article by Rafee Ebrahim (2001) the topic of fuzzy logic is explored in great detail with reference to computer systems. Fuzzy logic is a simplified version of true logic, not in its validity or scope, but rather in the type of data used. The common understanding of computers is that they use true logic; valid numbers and valid equations to reach valid conclusions. Using fuzzy logic, the computer can use general (even categorical) information and valid logical equations to reach sound and complete conclusions. A sound conclusion is a conclusion that is guaranteed to be correct. A complete conclusion is all correct outcomes. Thus, the fuzzy logic explained by Ebrahim gives only correct answers and all correct possible answers to problems of logic.

Ash uses rational reasoning throughout the movie, from persuading crew members to agree to investigate the message beacon, to ensuring the away team is not interrupted when they have a chance of encountering the alien life form. In the first example, a fellow crew member does not wish to land on the planet unless he gets a pay increase. Ash recognizes that the way to persuade this particular crew member is through monetary means, so Ash reminds the crew member that if he does not agree to land on the planet, it is in his contract to forfeit all of his pay. Needless to say, the crew member was more than happy to land on the planet after that reminder. In the second example Ripley wants to follow after the away team to inform them that the message beacon was not an S.O.S., but some sort of warning. Ash believes that Ripley may bring the away team back before they encounter the alien life form, so he makes Ripley believe that she will not reach the away team before they find the beacon, thus Ripley does not leave and Ash's chances of examining the alien are increased. Ash also shows his ability to reason using probability far surpassing that of the human crew; shortly before his demolition, Ash confesses to Ripley his assessment of the crew's likelihood of survival against the alien: "I can't lie to you about your chances, but you have my sympathies."

Even though Ash is an android, he does not sacrifice his knowledge of irrational reasoning for his surplus of rational reasoning. He understands how to manipulate his fellow crew members and not be manipulated himself, along with many more hard-logic based ideas such as probability.

Problem Solving

Very closely related to, and often overlapping with, reasoning, is problem solving. Problem solving often uses reasoning and insight in order to obtain a goal which is seemingly inaccessible. There are many types of strategies used for problem solving: the largest split can be shown between heuristic and algorithm strategies. Heuristic strategies are often based on intuition and require little mental effort, but do not always yield a correct answer, but are used about 90% of the time by humans. Examples of heuristic strategies are the hill climbing heuristic and the backward search heuristic. The hill climbing heuristic uses the idea that any step that seemingly gets you closer to your goal is a step in the right direction. The backward search heuristic can be useful when there are few known goals, and many possible paths to them; it mentally starts at the goal, and then works towards the current state. Algorithmic strategies guarantee a goal state, but are slow and require much mental effort. Examples are the exhaustive search and the means-end heuristic. Although the means-end heuristic is debatably a heuristic (even shown by the name), the function of it is more similar to that of an algorithm; it allows the temporary moving away from the goal state in order to find the most efficient way to reach the goal. An exhaustive search is the most inefficient and examines all possibilities.

In an experiment by Pratihar, Deb, & Ghosh (1999), a fuzzy logic and genetic algorithm hybrid were made for a self-navigating robot to reach its goal destination when confronted by three or eight moving obstacles. Essentially, the fuzzy logic was based on rules made by the genetic algorithm. The fuzzy logic was to move a certain direction when a moving obstacle is on a collision course with the robot under certain circumstances. The genetic algorithm is a set of rules governing when to move and what direction to move based on internal simulations and the comparison of these simulations in their effectiveness of reaching the goal state. These rules are then programmed into the robot for the fuzzy logic to use to govern the actions of the robot. Using these methods, the robot was able to predict the optimal, or near-optimal, route to travel to the goal, in some cases the planning was so far in advance, there were no rules necessary incase the moving obstacles were 'very close' by fuzzy logic standards. It is possible that the results of this study may indicate another twist in artificial intelligence; that of thinking ahead when there is time, and then acting on the rules thought of. In other words, the artificial intelligence may be able to create scenarios mentally and identify the most efficient rules for dealing with such scenarios, so that when time is critical, they do not have to compute such rules or act on 'reflex'.

Ash is essentially a problem solver in the movie. His goal is to bring the alien back to be examined further, and all other considerations are secondary. As mentioned earlier, Ripley finds out Ash's intentions and shortly thereafter his identity as a robot is discovered. Ash has a clear objective, but no clear means of accomplishing this objective, therefore, Ash must postulate a plan, and possibly many plans to achieve his goal. Furthermore, his ability to keep his plans a secret show incredible social problem solving in his ability to manipulate his crew members. Many times Ash makes subtle statements and gestures which show his influence; often times using reasoning to influence his crew members to see things as he does.

Although Ash's problem solving skills seem very impressive, possibly even superhuman, his actual ability to accomplish his goal was insufficient. Possibly had Ash not malfunctioned, he may have been able to complete his objective.

Consciousness

Consciousness is the state of being aware. Conscious awareness is the state of being aware of the external world through percepts. Reflective consciousness is the state of being aware of the self; in actions, emotions, thoughts, and motivations. Reflective consciousness is also known as metacognition. Questions concerning the existence of our free will and even the consciousness of another person other than the self are not a matter to be explained here due to their inherent philosophical, not empirical, basis. Much of the research regarding consciousness involves sleep stages, unconscious perception, dreaming and neuroanatomy. These concepts do not apply in the current paper, so they will be mentioned but not discussed. Jacoby stated the main function of consciousness is to enable the person to have an understanding of the self and environment, so the person can use the self to alter the environment in order to accomplish a goal. This view will be discussed with the character of Ash.

Franklin and Graesser (1999) created a computer program named CMattie, which is aware of the 'external world' and has 'self consciousness.' Amazingly Cmattie is not constructed to have total self-awareness, but rather only where self-awareness would be useful, such as a conflict between unconscious processes. CMattie is built to match Baars' (1988, 1997, as cited in Franklin and Graesser) global workspace theory, which gives criteria of what 'awareness' in computers must be. CMattie is used to read and interpret informal style emails from people scheduling, altering, canceling, etc... seminars. Cmattie does much unconsciously, much like humans do not always know exactly what they are perceiving, or even thinking, CMattie does not always pay conscious attention to internal processes unless necessary. For example, humans may not pay conscious attention to another person walking along; it is processed and understood unconsciously, however, if the person were to walk through a wall, then it grabs our attention quickly. This is because of the conflict between our understanding that a wall is solid and a human is solid, so the human can not just pass through the wall. CMattie works in a similar way as humans: some actions and processes are conscious, but much are unconscious until they are in conflict, such as two seminars scheduled for the same time in the same place. CMattie has many attributes similar to humans, including working memory, a 'neural net' of association, focus, and even emotion among other things. CMattie actually gets 'anxious' as the date of a seminar announcement is about to be sent out, and feels 'fear' about losing power. CMattie even gets 'annoyed' with people who send useless emails or do not respond to schedule confirmations.

Ash is an android that is on the ship to fulfill a goal. That goal is to find new life and return it for further research. Ash uses his understanding of self and the external world (much of which has been discussed previously) to attempt to accomplish this goal. Ash behaves like he has some emotions, but not remorse. Because emotions are part of the reflective consciousness, they are important to mention. Possibly the reason Ash does not feel remorse about his actions is his priorities: "Bring back life form: priority one, All other priorities rescinded" and thus feeling remorse may interfere with his first priority. He does state a deeper understanding of the alien life form, possibly even as a metaphor for what he wishes he could be, he states "I admire its purity. A survivor...unclouded by conscience...remorse...or delusions of morality." Ash's consciousness seems linked to his operation that is his 'life.' When Ash is badly damaged, he is unconscious, but when he is repaired, he is able to interact with the environment again.

Ash shows signs of consciousness very similar, if not identical to, those of humans. He seems more aware of his self and the external world than many of the crew members are of their selves and the external world, however, his emotional range seems restricted due to his primary objective.

Conclusion

Ash shows superior reasoning and problem solving abilities as compared to humans, but debatably has only equal consciousness as humans. This was shown in his ability to control his environment in an efficient manner; demonstrating his consciousness of self and external world sufficiently enough to use reason and apply these to problem solving in an attempt to obtain his goal of inspecting the alien life form in-depth.

References

Ebrahim, R. (2001). Fuzzy logic programming. Fuzzy sets and systems, 117, 215-230.

Franklin, S., & Graesser, A. (1999). A software agent model of consciousness. Consciousness and Cognition, 8, 285-301.

Pratihar, D. K., Deb, K., & Ghosh, A. (1999). A genetic-fuzzy approach for mobile robot navigation among moving obstacles. International Journal of Approximate Reasoning, 20, 145-172.