Post No. 24: Robots and US

Now that we have an informed amateur’s idea about AI (see Posts 21 - 23), it might be worthwhile to explore which elements of AI are similar to us and which are dissimilar. To begin, let’s consider how we humans developed our unique view of the world. Each of us has similar yet unique views of the world around us. We each started as a single cell resulting from the meetup of a male sperm and female egg. How did our complex perceptions of ourselves and the world around us emerge from such a tiny beginning? Are there any parallels with how an AI Robot develops?

The fetus develops by a process of cell division/multiplication and cell differentiation. By the third trimester our cells have differentiated into the full set of human organs. What is most notable for our subject at hand is that a fetus’ sense of touch, hearing, taste and smell have all matured prior to birth. A fetus’ sense of taste can detect the foods and their smells that its mother eats. It can detect touch sensations from perioceptor cells throughout its body. It can hear sounds outside the womb, including its mother’s voice, music, etc. However, when a baby is born its vision is very limited. It can only see a few feet in front of it and can detect moving objects at that distance. Despite having most of its senses operational, the baby has yet to form a conceptual awareness of the world it now comes into contact with. The coding for how this will happen is embedded in the DNA of every cell of the baby’s body. Coding occurs for both humans and robots. Let’s explore some basics of this coding in humans and how it is similar yet different in robots.

DNA, the code for growth and development of life forms, is found in the nucleus of every cell of our bodies. The process of how it eventually results in creation of our human bodies is complex, but the basics are somewhat easier to understand. Within the nucleus, messenger RNA (mRNA) reads the essential coding embedded in DNA and transports it to the cytoplasm in the cell. Cytoplasm is essentially the shop floor of the cell for manufacturing the proteins needed to build and regulate our bodies. The coding is embedded in a sequence of nucleotides which make up both DNA and RNA. A ribosome reads mRNA nucleotides and in turns provides the code for specific amino acids. The final step occurs when transfer RNA (tRNA) reads this code and assembles the amino acids into a protein, the building blocks of our bodies. Of interest, the amino acids, needed to build proteins, enter the cells through the blood stream (or through the placenta for the developing fetus). All the necessary ingredients for amino acids in the bloodstream are created by the food we eat. Amino acids, needed for the cellular manufacture of proteins, pass from the bloodstream through the porous cell membranes of our bodies. Hence the importance of good nutrition. The proteins manufactured by the cell create the structure of all our organs (skin, bone, heart, lungs, etc.), and other essential bodily functions such as antibodies (for defense), hormones and others. The coding for cell division and differentiation is also encoded in DNA. All the main actors in this manufacturing process from DNA to RNA to final protein manufacture are composed of molecules from five basic elements: hydrogen, oxygen, carbon, nitrogen and phosphorus. Thus the code for all aspects of the human body are made up of, and regulated by, molecular combinations of these elements. How are AI robots different?

Computer coding is not based on molecules of the five elements mentioned above. Rather, all computer coding is based on elementary particles, such as electrons or photons. Manipulation of these elementary particles are what makes possible the coding of ones and zeros (bits and bytes) at such incredible speeds. Considering the above, the physical basis for human coding (molecules of five basic elements) is rather cumbersome compared to computer coding based on incredibly smaller elementary particles. Could this coding advantage mean that robots will be superior to us? Before we attempt a response to this question, there is further “coding” that happens for humans that is qualitatively different from other life forms.

All animals have similar coding described above in their cell development and growth. But humans, with their large, specialized brains, have the additional ability to create abstract concepts of the world around them. These concepts are similar to the algorithms that computer scientists create for AI. These computer algorithms provide purpose and direction for an AI. In a very similar manner, humans are born into a society that provides basic algorithms for purpose and direction. Other animals, consider a dog for example, do not have these higher level algorithms. A dog born in Saudi Arabia behaves similarly to a dog born in Ireland. But a female human, born in Saudi Arabia, will have a significantly different algorithm than a female born in Ireland today. The AI also learns from experience, not unlike how humans learn from experience. For example, females in Saudi Arabia have the potential to change their algorithms based on their life experiences. We humans can change the algorithms that provide purpose and direction in our lives. The coding of our human algorithms begin from the moment we are born into our particular society and our particular time. Someone born in 1948 in the U.S. has likely developed different algorithms than someone born in 2010. The ancient Greeks were one of the earliest humans to realize that we have the ability to change our view of the world (our algorithms). Sometimes these changes have been productive and sometimes not. Communism and Fascism were widely held algorithms that did not fare too well with experience. It seems that the algorithms for us humans are always undergoing change, some small and some large.

It is fortunate that we humans control the algorithms for AI. The last post (Post No. 23) described how it is unlikely that we would create algorithms that lead to robots taking over. But it is theoretically possible for this to happen. Let’s keep an eye out for that rogue, crazed computer scientist who wants to create an AI monster!

As an aside, now that we know a little more about the coding inside animal cells, it might be interesting to look at how the Moderna and Pfizer mRNA Covid vaccines work. Once the genome of the Covid vaccine became known (the one thing China did right), it was relatively easy to develop a messenger RNA (mRNA) code that could manufacture the spike protein of the Covid virus. The immune system then recognizes the manufactured spike protein as foreign and produces natural antibodies to neutralize the Covid virus. Although the code itself is easy to develop, it took scientists over a decade prior to the Covid epidemic to ensure this technology worked effectively with the body’s natural antibody defenses.

Humans and robots are similar in that the “manufacture” of both is directed by coding and we both learn from experience. Robots learn from massive amounts of data fed to it by humans as well as from limited human supervision and feedback. They get direction and purpose from human developed algorithms. We get direction and purpose from the societies in which we are born and from our lived experiences.

I have found that exploring this whole subject of artificial intelligence and robots has shed more light on who we are as humans. I’m quite certain that AI will not only create wonderful advances for our human lives, but also give us pause to better appreciate the nature of being human.