Artificial intelligence tends to be a very polarizing subject: on one hand are fervent believers, claiming AI will solve all the world’s problems; on the other hand, are those who view it with an almost unshakeable sort of skepticism and disgust, going so far as to claim it is an inherently demonic technology. Who is telling the truth?
As is often the case, we might find that there is truth on both sides. It is my contention that if we approach the technology prudently and with an appropriate sense of judiciousness, we will be able to extract the upside — a world of personalized learning — while minimizing the downside. In this article, let’s focus on one of the most common perils AI may bring us: linguistic homogeneity.
The Homogenizing Effect of Industrial Technology
Invariably, industrial technology — meaning technology that operates at scale, allowing us to transcend borders and build products for millions or billions of users — has a homogenizing effect of some kind. As an example, I’m a shopper at Costco, as are many of my peers with spouses and children who have not yet left the nest. When I am at a social function with these peers, a comically large proportion of the men are wearing the same exact Costco socks. This is the result of many aspects of industrial technology: Costco’s supply chains and warehousing; clothing manufacturing; roads and cars.
An example of the socks worn by myself and almost all of my male peers
In the case of socks, this is little more than a fun observation. But many of us live in cookie cutter houses that have the same layout. We have the same phone. A lot of Toyota cars as well…
The economics that drive this situation are understandable and rational, and part of the advantage of industrial technology; it allows us all to enjoy higher quality goods and services that are comparatively more affordable than options available to those who do not utilize industrial technology. However, this advantage comes at a price of cultural loss; lamentations that all big cities are the same, for instance, are increasingly frequent.
AI and Linguistic Homogeneity
Artificial intelligence may bring us linguistic homogeneity — basically, a decline in the number of words and languages used. As we rely on artificial intelligence for much of our text needs, we are essentially outsourcing our text creation to a smaller set of linguistic systems that can operate at scale. Language death is already occurring, and artificial intelligence seems likely, in my opinion, to exacerbate the process.
It is worth noting that I’ve focused this issue on linguistic homogeneity — but as AI systems develop for other disciplines such as images, videos, music, DNA analysis, and more — the fortune and perils of homogeneity will appear there as well. Music technology lets us all create studio quality music from our homes and distribute them to the world, but music of all genres has converged to similar compositional standards. It seems unlikely, in my opinion, that AI will do anything but exacerbate this trend.
How to Use AI But Prevent Excessive Homogeneity
Accordingly, parents may wish to explore how they can thwart this: time spent honoring and remembering their local, cultural traditions help, as do studying languages and cultures that have disappeared. The loss of language entails a loss of meaning, to a certain degree, and so studying languages and cultures that have died or are dying may be a worthwhile endeavor in their own right – and something that may help ward off some of the more caustic aspects of artificial intelligence.
Ultimately, much depends on ensuring we use AI consciously, rather than blindly. Many families understand that smartphones possess great utility, but if used mindlessly, can be little more than purveyor of vices and mental sickness. Ways to use AI consciously might include:
- Using AI to create a curriculum that is truly personalized for your student. For instance, at Pallas, we are working on courses covering large aspects of post-World War II history that are neglected by conventional education systems; creating these courses is much easier with artificial intelligence. What do you want to learn, or do you want your student to learn, that is easier to create with AI? What can be created that adheres to your values and inspires your student?
- Critically, avoid using AI mindlessly; avoid using it all the time, and as a replacement for your own thinking. Undoubtedly, many will fall into the trap of simply asking AI every question, and accepting its response wholesale — thus avoiding any critical thinking. This, in my opinion, is a recipe for disaster.
- Last but certainly not least, pay attention to where AI might have a homogenizing effect — and where you may find this to be especially undesirable. Do you want your students to have an appreciation for non-digital art? Or languages and traditions that are not well-represented by AI systems? Be creative in finding solutions where AI can help you, but also be mindful where it may not offer any solutions and where other tools are thus needed.