Aye Aye – Bamidbar
אִישׁ עַל דִּגְלוֹ בְאֹתֹת לְבֵית אֲבֹתָם יַחֲנוּ בְּנֵי יִשְׂרָאֵל מִנֶּגֶד סָבִיב לְאֹהֶל מוֹעֵד יַחֲנוּ (במדבר ב, ב).
Am Yisrael camped around the Mishkan in a very specific configuration. There were four "sub-camps" surrounding the Mishkan to the north, south, east and west. In each sub-camp were three tribes, as follows (clockwise) –
The camp of Yehuda to the east – Yehuda, Yissachar, Zevulun
The camp of Reuven to the south – Reuven, Shimon, Gad
The camp of Ephraim to the west – Ephraim, Menashe, Binyamin
The camp of Dan to the north – Dan, Asher, Naftali
In the center, adjacent to the Mishkan, was the camp of Levi.
Each individual tribe had their own flag. The color of the flag was the same color as the tribe's stone in the Choshen and in addition, each tribe's flag had a picture. In addition to the individual flags of each tribe, each sub-camp also had a flag.
This configuration is not random in any way and has deep spiritual significance. It resembled the arrangement of the 12 Tribes as they accompanied Yaakov's coffin back to Eretz Yisrael from Egypt when he died. It also reflects the configuration of the angels, "camped" around the מקדש של מעלה. Bli neder, I hope to devote a future shiur to the significance of this configuration. However, this week's shiur explores a different topic.
From the passuk above אִישׁ עַל דִּגְלוֹ בְאֹתֹת Chazal and the Mefarshim say that on the four flags of the four sub-camps, there was also writing (from the word בְאֹתֹת meaning אותיות). Targum Yonatan says that on each of the four sub-camp flags was a different passuk.
The Chizkuni says that on each of these four flags appeared three letters. On the camp of Yehuda's flag appeared the letters איי. On the camp of Reuven's flag were the letters בצע. On the flag of the camp of Ephraim were the letters רחק. On the flag of the camp of Dan were the letters מקב. The Chizkuni explains the significance of these letters. They are derived from the names of the Avot.
If you take the first letter of אברם, that is א. The first letter of יצחק is י. The first letter of יעקב is י – making the three letters איי.
The second letter of אברם is ב. The second letter of יצחק is צ. The second letter of יעקב is ע – making the three letters בצע
And so on. The Chizkuni explains why use is made of Avraham's name Avram without the ה, but that is a long explanation and not relevant to this shiur.
Today I would like to discuss a subject that is related to the three letters of the camp of Yehuda's flag איי. The Vilna Gaon explains the meaning of all four of the 3-letter combinations and for the איי combination, the Vilna Gaon says that it is referring to the first passuk of Shma Yisrael. In this passuk the name of HKB"H appears three times, twice with name Yud Keh Vav Keh (two yud's) and once with name Elokeinu (alef) making the letters איי.
How do you pronounce איי? Perhaps the closest English word that sounds the same is "aye" (as in - aye aye captain!)
In today's shiur I would like to discuss a different "aye" (some people call it ayayay). I am referring to the hot topic of discussion these days, the subject of AI, which is an abbreviation for Artificial Intelligence. It is impossible to escape hearing about it these days as it is featuring prominently in the media. You surely have heard about Chat GPT and the interviews with Elon Musk, CEO of Tesla and his philosophy on AI. Basically, AI is taking the world by storm and is poised to cause a revolution in our lives. The proponents of AI say that it holds great hope for the future of mankind, while the opponents fear that it will someday take over the world and cause the extinction of the human race.
In this shiur I would like to explore AI a little and specifically discuss the philosophical and Torah repercussions of it.
What is AI?
Since I have a past in computers, I have previously delved into this subject quite extensively (I must admit that I am not fully up to date on the most recent developments, but I understand the underlying principles behind it).
Artificial Intelligence is a combination of computer hardware and software programming using a specific methodology called "neural networking".
When I say "computer hardware", I am referring to that box that sits on (or next to) your office desk, the laptop or tablet you lug around in your briefcase and even your smartphone in your pocket. In addition, there are a lot of "hidden" computers in all kinds of things, like cars, TV's, fridges, ovens, microwaves and many other electrical appliances.
As with most technology, the "model" for the concept of a computer was derived from nature. Nature is so magnificent and a source of such inspiration, because its Creator is so magnificent. The reason that HKB"H created the world and nature is so that we could observe it, study it and marvel at it and thus arrive at an understanding that there is a higher Power that created it and thus come to recognize HKB"H.
The model of a computer was based on the human brain and the computer "switch" was modelled on a nerve synapse. Obviously the complexity of the human brain far surpasses the complexity of any computer
When you strip away everything, all the fancy casings, buttons, screens, accessories, etc., at the heart of every computer is … a simple switch (like a light switch on your wall). The switch has only two possible positions - on or off. When the switch is on, electricity flows (the light is on) and when the switch is off electricity does not flow (the light is off). A system that has only two possibilities is called a "binary" system which has only two possibilities 1 (on) or 0 (off). This is one of many mathematical systems, for example the "decimal" system which has ten possibilities (0,1,2,3,4,5,6,7,8,9). In binary there is only 0,1.
When you imagine a switch, you imagine a "big" thing consisting of two ends of wire and some kind of "breaker" to connect or disconnect the wire ends. In fact, the first computers were constructed from these kind of switches and the computer filled an entire warehouse (like the computer that cracked the Enigma machine in WWII). The power of those first "monstrosities" does not even come close to the power of the smartphone you now hold today in the palm of your hand.
The revolution occurred in the 1940's and 60's with the invention of the transistor and the silicon chip which allows miniaturization of the "switch". You could now "imprint" (using a multilayered manufacturing process) thousands (or even millions or billions) of these switches on a tiny little "chip" made from silicon. No more clumsy wires. The chemistry of the silicon mixture allows for these microscopic switches to be turned on and off using electric current.
What good is a whole bunch of switches? Obviously on their own, they are just a pile of scrap metal (or sand – silicon is essentially sand). However, if you apply a branch of mathematics (called Discrete Mathematics) to these switches, you can get all kinds of interesting things.
First of all, you can get something called "data". If you take a whole bunch of switches, each connected to a red light, a green light and a blue light - by turning these light switches on or off in various combinations you will see a red light, or a green light or a blue light, or combinations of the above, e.g. if both the blue and red light switches are on simultaneously, you will see a purple light. If these lights are very tiny and you arrange them in multiple rows and columns, alongside each other and turn them on and off in various combinations, you will have what is called a computer "screen" - a collection of thousands of lights with switches to turn them on and off. You can thus turn these switches and dots (pixels) on/off to form letters, words, sentences, pictures etc. This is called "data" – a whole collection of switches turned on and off in a specific pattern.
This is all well and good as long as the screen (and the computer) are turned on. However, as soon as you switch them off, the data disappears from the screen. To preserve the data so that it does not need to be reformulated every time from scratch, scientists invented storage media to store this data. The first types of storage media used magnetism to "remember" the data. They took a tape or disk and coated it with a magnetic coating. The individual particles on this coating could then be magnetized to represent opposite poles of the magnet (North/South). If the particle was magnetized one way it represented an "on" switch, if the other, an "off" switch. The important thing was that this polarization lingered, it was not dependent on constant electrical flow. You could therefore record the "data" above on such a disk, take the disk out of the computer, put it in your pocket and carry it around with you until a later date, when you would put it back into the computer and reload the data from the disk back into the computer. This enables you for example to write a shiur on the computer, save it on a disk and preserve it for later use. Modern storage devices like Disk-on-Key use more advanced (charge-trap chip) technology.
Another important use of these switches is not to create data, but to process data. By using Discrete Mathematics principles, you can create combinations of switches that can add, subtract, multiply, divide and perform other untold mathematical and analytical operations on the data. These calculations may be performed almost at the speed of light! Much faster than any human could calculate in their head, as many times as you want. The computer never gets tired. By arranging these switches in a specific configuration of ons and offs you in fact create something called a "program", a set of instructions to perform a specific task. This "program" may be performed over and over again and you may design different programs to do different things – to add numbers, to subtract numbers, to sort numbers, to search for data, to filter data, to analyze data, etc. People who create such programs are called computer programmers and the way they work is not by manually turning switches on and off one at a time, but using a much easier tool called a programming language, which itself is a program that allows programmers to type commands in verbal form which are then translated into ons and offs in the switches. Certain programs are part of the computer itself, they are stored permanently in the computer "processor" chip, because they are used so often and repeatedly. Other, more advanced programs build on these basic programs and combine and embellish them in various ways to perform more specific and complex tasks, like navigate to a destination (Waze) or play music and video (Youtube), etc.
That is the basis and at the heart of any computer. I felt it necessary to spend so much time explaining it because today computers are so powerful and can do so many things, that we tend to regard them as something "magical" and beyond human understanding - when in fact they are simply a glorified bunch of switches and their capabilities are limited by the program they are running, programs which are created by man.
AI is a computer program, but with a twist. It tries to mimic the functioning of the human brain, by giving the computer the ability to "learn". It begins with the original program, a basic set of instructions and data that forms the initial basis of the AI's storehouse of knowledge (knowledgebase). In addition, it gives the AI program the ability to add any new knowledge it encounters to its knowledgebase. The AI program is presented with a question that it must try to answer. It first searches its knowledgebase to see if it can find the information there. If not, it tries to obtain the information from other sources, other knowledgebases and then adds the "acquired" knowledge to its own knowledgebase. If it cannot find any source for the knowledge from every resource at its disposal (from the internet, etc.), it will wait for someone, or some other source to supply it with the missing knowledge. Therefore, the AI program "learns", it continually acquires more and more knowledge in its knowledgebase. Since computers process data quicker than humans, the AI program is also able to analyze and filter this data in the blink of an eye and provide an answer almost instantly. With such huge resources of information at its disposal and such lightning speed processing ability, the AI "seems" to have intelligence – it provides (mostly) the right answers, instantly, every time! In this sense, the AI program is dynamic, the more time goes by, the more knowledge it accumulates and the more powerful and efficient it becomes.
AI programs are not limited to online Question and Answer (Q&A) sessions, like ChatGPT. The answers they provide can be used in an "automated" sense, without the manual act of asking a question and receiving an answer. For example, they are used by systems like Google and Facebook to "learn" user behavior online – which websites you visit, how often, on which days, what time of the day, how long you spend there, what information you are looking for, whether you do or don't buy products there, etc. and they build "profiles" for you that allow these companies to provide you with targeted advertising etc. Most of the time you spend on the internet, you are in fact being "spied on", your every move recorded and analyzed by someone.
AI programs are also used to copy behavioral patterns, like creating art in the style of a famous artist, or creating music in the style and even with the same voice of a famous singer. They "learn" whatever it is possible to learn about the brushstrokes of Van Gogh for example, the color combinations, the lighting, etc. and then you may ask the AI to create a picture of your garden using the Van Gogh style. Or it may "learn" the pitch, timbre, volume, enunciation, accent, range, etc. of a singer long deceased, like Elvis and then allow you to create a song that Elvis never sang, in the Elvis style. The results are uncannily real and accurate.
AI programs are therefore a tool to accumulate and analyze knowledge and provide the best solution to a problem.
The scope of their use is almost unlimited. They may be used to provide medical diagnoses, corporate decisions, route air traffic, control production, detect security threats, etc. Since they have access to the "entire knowledgebase of the world", online on the internet – every book ever published on a subject, every court ruling/precedent, every symptom ever diagnosed in a hospital, etc. the scope of their resources exceed any one human "expert" and the speed of their analytical capabilities exceeds human processing capabilities.
Those who are opposed to AI envision apocalyptic scenarios like the "Terminator" movies where AI programs become "self-aware" and take decisions to remove humans from the equation.
It was important for me to go into such detail here, to show, behind the "façade" or the magical/superhuman "mystique" that AI is currently acquiring – that all it is a bunch of "scrap metal" and limited by the intelligence of the person/people who programmed it. The programmers of the AI, not only gave it is basic, initial data, but they also designed the rules by which it acquires new data and how it analyzes. The AI program is not capable of exceeding the bounds of its initial and dynamically acquired programming.
The million-dollar question, in a philosophical sense is – "Is this an acceptable "tool" in Torah terms?"
The Torah does not forbid use of tools. It does not matter whether the tool is the sword invented by Lemech's son Tuval Kayin, the plough invented during the time of Noach, the Achnai oven invented in the time of the Gemara, or any other tool invented or developed over the millenia. HKB"H on the 6th day of Creation told man וּמִלְאוּ אֶת הָאָרֶץ וְכִבְשֻׁהָ – to conquer the earth, to use our intelligence and knowledge to improve our lives.
Any tool is not inherently good or evil, it all depends on how it is used. If you turn a tool into a form of avodah zara, then that is an evil use of the tool. If you use the tool to defy HKB"H and the Torah, that is an evil use of the tool. If the tool is used to harm mankind, that is an evil use of the tool.
Technology is inescapable. Each generation lives a reality of its own technology. The question is how we use it.
The question is how much control we relinquish to it and how much we retain. Unbalanced relinquishing of control can have catastrophic consequences. For example, using technology to assist us to an extreme degree may result in a decline in our health. We may adopt more sedentary lives that lack the minimum physical activity that the human organism was designed for. Unbalanced use of technology may make us unhealthy, lazy and stupid. Instead of stimulating our brains (which like any muscle, needs exercise) with activity, we may teach our brains to be lazy.
AI is not "intelligent", it is efficient. It efficiently performs a task. However, there is more to "intelligence" than simply acquiring and processing knowledge, AI is lacking one vital ingredient - a conscience. This ingredient can never be duplicated or "created" by technology or by man. It is part of the צלם אל-קים that HKB"H created us with and just as only HKB"H can create life, only HKB"H can create a conscience.
AI, like any tool, is valuable for its ability to provide solutions, but the decision of which solution to adopt and which to reject must remain the domain of humans. Relinquishing that power to a bunch of scrap metal can only result in disaster. If we give the power of decision to a machine, without human conscience to regulate the judgment of the decision-making process, then the "Terminator" scenario is not science fiction, it could easily become reality. Only humans should make life-affecting decisions, because only humans have the things a machine can never duplicate – instinct, intuition, inspiration and conscience.
I am not going to get into the halachic aspects of using AI, like for example to "proofread" a sefer Torah and identify mistakes. That is the domain of institutions like Machon Tzomet, who explore the halachic ramifications of technology. I will remain on the philosophical level.
AI may look like magic to some. I am sure the first plough looked like magic to Noach's generation, but like the first plough, AI, at its heart, is just a bunch of scrap metal. It may greatly benefit mankind, depending on how and for what we use it. If we turn it into some kind of "superpower" and begin to idolize it, and if we relinquish human control entirely to it, we will be inviting calamity upon ourselves.
The letters איי on the camp of Yehuda's flag come to remind us what the true AI is – it stands for Almighty Intelligence. There is only one supreme intelligence and that is HKB"H, who created the world, humans, metal, switches, programs and everything else in the universe. Mankind are commanded to use their Divine intelligence, part of the צלם א-לוקים to control the world – in the service of HKB"H. If we carelessly abuse our intelligence and the products of our intelligence for anything else, it becomes evil.
It was important that in each sub-camp, the flag should contain components of all three of the Avot. This was so that one tribe would not say – "I am only like Avraham (chessed)" or that another would say "I am only like Yaakov (limud Torah)", etc. The flags in מחנה ישראל incorporated elements of all three of the Avot, to teach us that every Jew must strive to be an "all-rounder", to devote our lives to striving to achieve ALL the middot of the Avot.
Avraham came to recognize HKB"H on his own by observing and understanding nature and changed the belief system of the world forever. Yitzchak gave us the ability to sacrifice, to put our selfishness aside and focus on the greater good. Yaakov gave us the diversity of the Twelve Tribes, each with their own strengths and weaknesses, each with their specific role and destiny to play in the world, but united in one cause – Shma Yisrael Hashem Elokeinu Hashem Echad". איי.
