Do you think that A.I. will eventually think like humans? Do you know how humans think? Do you know how A.I. should think? Maybe this and a previous pearl can help you decide.
We hear a lot of “opinions” today, from those who make a living promoting A.I. and from those who philosophize about it. The purpose of this pearl is to explore whether A.I. thinking can mimic human thinking using the “Genetic Human Personality TOOL”. Also refer to the pearl: “Why, How and What We THINK” to understand human thinking, so that this pearl can have more meaning for you, if any.
Note: In this pearl, A.I. thinking, whether by a robot or by a computer will be referred to as A.I. Thinking. But when in a robot, it will be referred to as Robotic A.I. thinking because it needs mobility and a lot more programming to survive. Also, A.I. memory will be referred to as digital memory.
Can A.I. Mimic Human Thinking…Without Human Prompting?
For A.I. to think without a human guiding it, it needs human-like mobility coordinated with human-like sensing to make appropriate decisions on how to react. Just independently thinking is of little value unless A.I. can make decisions and react to survive. Survival for humans involves thinking that avoids social or physical harm; survival for A.I. involves avoiding physical damage or digital memory loss/corruption. Of course, if a human carried an A.I. computer, mobility would not be an A.I. problem, but that would require human prompting.
Asking a Robot A.I. or A.I. computer a question about what it sees (scans) is similar to asking a human a question about what they are looking at. A human would answer based on memories stored in their brains and A.I. would answer based on its existing digital memory or Internet files. When a human is asked to describe something new they would be at a loss, but probably say “but it looks like…” When A.I. is shown something new that it doesn’t have a digital file on it would also be at a loss, although it could guess, which is what a Goggle Search does. So, in that respect, A.I. thinking can mimic human thinking.
For Robotic A.I. to be free to “think” like humans it will need human-like mobility and human-like sensing, which it must coordinate in order to make decisions on how to react, survive and learn.
Mimicking Human MOBILITY: Robotic A.I. needs a highly advance robot controlled by a powerful computer containing hundreds of terabytes of digital memory. Today, an advanced A.I. robot stiltedly mimics human locomotion, but can they physically survive? For example, if a robot were walking across a street, would it be able to quickly react to avoid being run over. A non-robotic A.I. computer does not need mobility programming, but it does need sensing programming to decide what the situation is and how to appropriately react to it.
Mimicking Human SENSING: Human sensing provides much more information than needed for humans to survive—much of what is sensed is ignored. Likewise, Robotic A.I. thinking should be streamlined to only store useful digital memory and only use the Internet for things it isn’t sure of—just like humans do.
Robotic A.I. thinking would require scanning systems to mimic human sight, sound, smell, touch, taste, somatosensory (balance-sensing) and pineal gland emotional energy (gut reaction) sensing. A sophisticated robot would already have somatosensory balance sensing in its mobility, but it will never have the emotional energy sensing that mammals have. Taste sensing would be of little value since robots don’t eat, unless they are "food tasters". Touch sensing, except maybe in the fingertips, is much too intricate in humans to mimic—although sonar might work for a robot. This leaves us with visual, auditory and smell sensing. There are machines available that can identify (smell) chemical compounds, but not all compounds. It would though, be helpful if a robot could sense drugs, caustic, incendiary or explosive compounds and possibly even diseases, just like dogs do. Sound sensing, along with a library of what those sounds mean might be helpful for Robotic A.I. decision-making and reacting—for example to a fire alarm or car horn. Visual sensing, similar to driverless car scanning is probable, as long as there is an extensive library of what those scans (pictures) mean, to aide in Robotic A.I.’s decision-making, reactions and learning—a lot of this is presently being developed.
Mimicking Human DECISION-MAKING: Robotic A.I. would need something similar to a human multi-layered PFC (Pre-Frontal Cortex) to make decisions, which is the most important thing it can do. It can achieve this using four CPUs, one for each human emotion and four CPUs, one for each human rational style—this would be the key for A.I. thinking if you want it to mimic human thinking.
Could EMOTIONAL Human Thinking Benefit Robotic A.I. Thinking?
The purpose of human emotional thinking is to physically survive, which isn’t required for an A.I. computer, but would be necessary for an A.I. Robot, although, if a human were carrying and relying on an A.I. computer for their physical survival, emotional A.I. thinking would be helpful.
Humans, in order to physically survive, have two left amygdala partitions and two right amygdala partitions that contain emotional memories and wired in processing to react to emotional alerts. It is neither practical nor possible for A.I. to have emotions, although emotional thinking could benefit it. Joy type thinking would help Robotic A.I. to scan for friends or foe. Anger type thinking would help Robotic A.I. to scan for possible confrontation. Fear type thinking would help a Robotic A.I. to scan for possible danger. And, sorrow type thinking would help a Robotic A.I. to scan for the loss of an important connection, which would be of little value unless Robot A.I.’s purpose was only to dealt with serving emotional humans.
Unlike humans, whose PFC can only process one emotion and its attached rational memories at a time, Robotic A.I. could have a four-part CPU that could process all four emotions simultaneously. Robotic A.I. scanning could produce data that simultaneously triggers independent anger-type, fear-type, joy-type and sorrow-type thinking. The input data to each processor would be the same, but the processing and decisions as the probability of that emotion being triggered would be different. There would also be a main CPU (one that runs the A.I. thinking), which would consider each emotion CPU’s decision and pick out the one that is the most important to react to—this is how A.I. learns. Finally, each emotion’s input data and decision would be stored in its individual digital file. Also, the decision selected by the main CPU that best fit the situation is stored in its “emotional reaction” digital file.
Bottom Line on Emotional A.I. Thinking: Robotic A.I. thinking could be programmed to use emotional human thinking to decide on how to react, to physically survive; but it is not as valuable as being able to mimic rational human thinking to make the kinds of decisions humans make to socially survive daily.
How RATIONAL Human Thinking Might Benefit A.I. Thinking
Humans have their frontal lobes divided into four types of wired-in rational thinking. Refer to the pearl: “Why, How and What we THINK” to fully understand human thinking. The left human frontal lobes contain the rational memories and programming (attributes) for the detail thinking analyzer and relator styles. The right frontal lobes contain the rational memories and programming (attributes) for the director and socializer style's interactive thinking. Also, the analyzer and director styles are task-oriented (tasks are the priority, not people) and the relator and socializer styles are people-oriented (people are the priority, not tasks). Also, directors and relators are opposite and analyzers and socializers are opposite and modify each other’s use. All of this can be used to set A.I. thinking priorities.
Each style has a purpose (the reason for its reactions), which can also be used to set A.I. thinking priorities. Director style thinking would rationally place A.I.’s priority on getting tasks quickly done or dealing with confrontation or inhibition. Analyzer style thinking would place the A.I.’s priority on cautiously proceeding or getting tasks done well. Socializer style thinking would place A.I.’s priority on avoiding negative connections and embracing positive connections. Relator style thinking would place A.I.’s priority on maintaining important connections.
The sequence of rational A.I. thinking would start with whether it is a task or people situation. If a task situation, both the director and analyzer style’s attributes would be used to make their decisions. If the task needs to be done immediately, the director’s decision would be the priority and if the task needs to be the most error-free the analyzers decision would be the priority—followed by director thinking to decide when the task must be done by. If a people situation the relator and socializer style’s attributes would be used to analyze and make a decision. The socializer style’s attributes would be the priority to help determine whether those involved were friend or foe, to decide whether to positively interacting with them or to avoid them. The relator’s attributes would be used to support others needing it should they be a friend.
Style Priority A.I. thinking: Once a specific style's decision is determined to be the most appropriate one to use, A.I. thinking would apply that style’s three motivations and is purpose. If a director style priority, A.I. would think: 1) only my concerns matter, 2) only my needs matter, 3) only my decisions matter, so that I can get the task done quickly or deal with confrontation. If an analyzer style priority, A.I. would think: 1) I must avoid loss/pain, 2) I must do what I’m obligated to do, 3) I should distrust new situations; so that I can cautiously proceed. If a socializer style priority, A.I. would think: 1) I must pursue pleasure and gain, 2) I need to be unconstrained, 3) I should trust new situations—to embrace positive connections and avoid negative connections. If a relator style priority, A.I. would think: 1) only other’s concerns matter, 2) only other’s needs matter, 3) only other’s decisions matter—to maintain positive connections. Once A.I. has reacted to the situation with the most appropriate style's decision it would store that learning experience in each style’s digital memory. It would also store it in A.I.'s central memory, including: 1) whether the situation was emotionally or rationally triggered; 2) whether it was a task or people situation; 3) each of the four style’s decisions from the most appropriate to the least appropriate for that situation.
An Improvement in human thinking: The evolution of human “thinking” isn't perfect—it's lopsided. Humans too often react with their genetically strongest style instead of the most appropriate style for the situation. By using the four human thinking styles (director, analyzer, Socializer & relator) simultaneously and then selecting the most appropriate decision, A.I. thinking could be more appropriate than most human thinking, as long as A.I. thinking can access the plethora of information that human thinking uses.
Bottom Line on Rational A.I. Thinking: A.I. rational thinking, whether computer or robotic, could be programmed to use rational human thinking to make decisions to appropriately react to socially survive.
Bottom Line on Robotic and Computer A.I. Thinking:
What I have presented here is an outline for what A.I. “human-like” thinking might entail and not an exact blueprint for creating it. A.I. thinking can't mimic human thinking exactly, but it can somewhat mimic it and it could improve on lopsided human thinking. I'm sure that there are thousands of people working on how A.I. should think, but few if any of them really know how humans think, which could make their efforts more successful.
I hope that this pearl on A.I. thinking has given you some understanding of the possibilities and limitations of A.I. thinking—at least from the ”Genetic Human Personality TOOL” point of view. How that thinking might negatively affect your life is what scares us, but how that thinking might positively affect our lives is what brings us joy.