Collective Intelligence Technology

How do we think about AI in the context of education?

This article is meant for students and educators. It is not another explanation about how AI works. I will not explain the science behind AI, nor will I give it some bold claim about what it will become. Instead, in reading this article, I will articulate a more useful way of defining AI. I will focus on explaining what AI is conceptually, not technically or theoretically. This definition will make it easier to determine how AI can best serve learners and teachers no matter their context. So…

The Original Information Technology

We hear the phrase “Information Technology” (IT) a lot, but it will be helpful for us to establish a working definition for this article. Quite simply, IT is technology that allows us to transfer information more quickly, more precisely, and over longer distances than would be possible without tools. This type of capability is so important to us as people that the first IT ever created has become an integral part of being human. In fact, as I recently learned at the Society for the Neurobiology of Language, this technology is a human right. I am, of course, talking about spoken (or signed) language. Spoken language enabled us to think together. It facilitated innovation, culture, and community. More than anything, it connected us to the people and ideas we cared about. Spoken language, the first IT, is such an essential part of who we are, that we have a region of our brain, defined in our genes, dedicated to its understanding and production. Language is quite literally part of our DNA.

Since then, we have made many new ITs. But there is an important pattern to recognize. New ITs enhance one aspect of information transfer in very useful ways, but they do so at the cost of other aspects. Take the second major innovation for example— written language. Written language allows us to transfer information accurately across time and distance. In writing, however, you lose valuable emotional context. The reader cannot see your face, your body language, or hear your tone. Because of this, they are left to assume this information on their own. We see the consequences of this in the common misinterpreted text message or email. Other major innovations include the telephone, video streaming, and social media. All of which improve the efficiency of communication, but eliminate some human aspect of it.

Accessing Collective Human Intelligence

With this definition in mind, we can now think about our newest form of Information Technology: AI. AI is a special case, so we will have to be careful in how we analyze it. In all other cases of IT, information is being transferred from one human to another. This is not true for AI as the response of the technology is produced by the technology itself. It is also not true to say that information is being transferred from technology to human either. The Generative AI Models that we interact with like ChatGPT or Gemini are created using human-made data— written and recorded. It is trained using the collective intellectual contributions of millions of humans. This is the key difference between AI and other forms of IT. You are not communicating with a single person, you are communicating with the collective intelligence of millions of people across different time-periods, locations, and backgrounds whose thoughts and ideas have been published on the internet. Of course, it is important to consider that this information is being filtered through the parameters set by the creators of the AI model. Further, not everyone had the opportunity, directly or indirectly, to publish their thoughts online. But understand that information is not strictly being transferred between the machine and you. Instead, information is indirectly transferred from the enormous set of humans that contributed to the AI’s training.

A Useful Definition of AI

So how does this distinction inform our definition of AI is Information Technology? We need to consider this in two parts. What improvements are made to the information transfer process? And what is the cost? The first answer is straightforward. AI vastly improves access to information. We can consider two similar forms of information technology to understand more deeply. For many years, libraries served as the comparable technology to AI. Libraries are an access point to the collective intelligence of people from far away times and far away places. Prior to the library, learning required the physical presence of an expert. With the library, that knowledge could be accessed without that limitation. Not too long ago, we invented the internet, and this improved our access to collective intelligence further. Now we can access a practically infinite library from our homes, but we still have to search for the information we want. AI once again improves our access to collective intelligence by removing this barrier entirely. You can now ask for any piece of information you want, and the technology will do its best to synthesize the collective intelligence and provide the information to you. Theoretically, access to collective intelligence could not be easier except through direct telepathic connection.

It is worth emphasizing this point. You can ask for literally any piece of information. Though the response may not exactly match the truth, it is a direct reflection of what we have figured out as humans so far. You can ask a direct question like, “How far is the Moon from the Earth?” and AI will give you a direct answer. This is an example of information you could also find in the library or look up on the internet. You could also say “Write a 2-week lesson plan using Project-Based Learning for my 8th grade class on the major contributions of the first 5 American presidents.” The question AI tries to answer in this case is not so obvious, but it looks something like this ”What would a 2-week lesson plan using Project-Based Learning for an 8th grade class on the major contributions of the first 5 American presidents look like?” This information may not yet exist in a library or on the internet, but this is not a problem for AI. It exists as potential information within the collective intelligence that it derives its answers from.

The Human Cost of AI

There is much more to say on this topic, as this method of information transfer has its own limitations to be aware of. However, I will not go into detail in this article. More relevant to this discussion is the answer to the second question: What is the cost? The answer to this question comes naturally from our answer to the last. Specifically, AI is a connection to collective intelligence. I cannot emphasize enough that this is not equivalent to collective imagination, and this is the major limitation of AI. Humans have the innate ability to imagine. We can generate new concepts. We can define new abstract spaces like geometric planes. We can create worlds inside our minds that have no relation or connection to the world outside. At the deepest level, our society is built on our collective imagination. Everything we invented as humans first existed as ideas within our imaginations, before being translated into a physical representation. The first fork was not made of stone, wood, or bone. It was made of mind. AI does not have a “mind” to do this. It may be able to synthesize thousand

This is the cost of communicating with AI. Because you are not communicating with a human, the information synthesized by AI lacks any imagination. It will always give you the most statistically “accurate” answer that it can to whatever question you ask. Through creative prompt engineering, you or the people who design the models can make the model simulate imagination (just as a writer could describe their emotions quite well), but is not imagination in its true form. With this insight, we now have a useful definition of AI. AI is a form of information transfer between you and collective human intelligence. This information technology greatly increases access to almost all information, including information that has not yet been synthesized. When you ask a question to AI, you are receiving a response that is a reflection of this creative intelligence. Importantly, this information comes at the cost of imagination, meaning you, as the human, must add this element yourself.

When you use AI, use it with its intended purpose. It is not designed to give you the right answer or a final product. Though it may end up producing these, you cannot count on it to do so. You can, however, expect it to give you a very strong starting point, from which you can improve by adding your human spark— imagination. If you want to create 15 questions for your students to test their understanding of a reading, you can ask AI to do so. It will do it much more quickly than you or anyone else could, and its answer will be statistically accurate. However, if you want the question to feel real and truly engage your students, you may need to make a few changes. With this understanding, try to push AI to its limits. Use it across as many contexts as you can and find out which work best for you. This is the only way to discover how to use it to best support your work. Just as everyone has their own unique handwriting, everyone will use AI in their own ways. And more than anything, don’t forget to imagine.