AI Literacy: The value of education and the paradox of university programmes
- Balázs Kis
- Jul 2
- 7 min read
Updated: Jul 3
This piece was sparked by a panel at the Lisbon Language Summit in May, where a panelist called out universities for doing too little to help language professionals meet today’s changing requirements. I began to wonder whether universities, or any education system, would ever be able to live up to expectations like this – and that it might be unfair to lay such expectations on higher education. Let me explain why I think this is so, at first in general, then in the context of AI literacy.
The trap of linear careers and learning
Most careers don’t follow a linear path and don’t stick to one single profession or skill set. In other words, we need to “upskill” or “reskill” a few times during our professional life.
Many people end up in a profession that is not directly related to their main or last university degree. [1] For example, at the software company I work for, the product management team includes a Latin teacher and two history PhDs. However, regardless of their original degrees, they also happen to be very competent at designing language software.
Still, paradoxically, we expect universities to furnish students with a complete and up-to-date skill set that makes them “employable” right after they graduate. We say this as students or parents; we ask for this as employers; and even in groups like the European Masters of Translation (EMT), we measure the “employability” criteria of university programmes by this. To be fair towards the EMT programme, their framework also emphasizes lifelong learning and regular updating of skills. [2]
Although universities are expected to release “employable” graduates, it is difficult to anticipate what skills will be required in a specific working environment. This is a tall order for the university, the employer (or client), and for the graduate, too. Sometimes it turns out that there is an unexpected, random skill that you didn’t pick up at university but that proves to be crucial in your current job.
Think: what is the skill you most frequently use in your current profession? What is the skill you have been able to use throughout your career consistently? Where does this skill come from?
Let me give you a very personal example: I am good at understanding communication, even in languages that I do not speak. I trace this back to the Latin classes at high school where we were taught to use almost forensic methods to decode meaning and then translate the text. This is a byproduct of a class that was supposed to improve our general literacy, and not the result of any graduate or postgraduate education I later had.
The role of universities
You can expect universities to do one thing: prepare you to adapt to new circumstances. To make a hard left and pick up a new profession when you need to. To pursue the knowledge you need. This is more important than the actual knowledge that you acquire at university and that will become outdated in less than a decade. Some of it might remain relevant, but that is an accident rather than careful planning.
What are the specific ways a university can prepare you to adapt? Well, it can teach you to learn. Ironically, I think the best way to learn to learn is to fail an exam or two, and fail early during your education. This is how you recognize you need to adapt (by preparing differently, for example). This works only if you have a safe space to fail, that is, you can try again. Most universities are like this.
A university can also teach you to think critically. This is also possible in a safe space, where you are encouraged to ask question and challenge your instructors without judgment and without repercussions. In my view, the entire education system, from kindergarten to PhD, should work this way. However, in most cultures, only universities come close to being an environment conducive to questioning and challenging authority.
Education systems (and we no longer talk about higher education only) work well if they accomplish two things: equip students with immutable bits of knowledge; and prepare students to be able to adapt. Everything else is the cherry on top, though obviously, teaching up-to-date skills will make it easier to get employed at first. Specific technical skills will also help students to evolve and adapt to the changes in the techniques learned.
However, knowledge associated with a specific university degree or career path is ephemeral. It will keep fresh for 3-5 years at best, at which time you need to reskill. That’s when the value of your university education shows: how quickly and efficiently can you pick up the new skills?
This does not mean that we should learn or study less. If anything, we should learn more. Part of the point is that you never know which skill will prove useful in one situation or another. Let me just refer back to the personal comment I made in the previous section... about my high-school Latin class.
The impact of AI on education
What does all this mean in the context of “AI is here to stay”? It means that AI will be part of our personal and professional lives, pretty much regardless of the profession or chapter of life we are in.
To cope with a world with AI in it, we need to be AI-literate. But what does AI literacy encompass? This is a debate we have yet to have, but the objective is something like
a set of skills and learned habits that help an individual control an AI-based system (however carefully the AI is hidden), and recognize whether or not the AI works to their benefit. This needs to start from an early age so that minors are able to use AI to their true benefit and do not become manipulated by AI systems that are potentially designed to condition and exploit them. [3] [4] [5] [6]
A truly open society will want – and enable – its citizens to learn to control and oversee AI systems, so that they benefit everyone, and providers are held accountable. [7] Less open societies will want – and enable – AI systems to control and track their citizens, so that the benefits stay within a narrow group in power and meaningful accountability is evaded. The narrative that regulation inevitably stifles innovation is patently untrue; with genuine intention, efficient innovation is possible even when proper accountability is in place. [8] [9]
How traditional education is critical to AI literacy
So, at this time, we don’t know much about what “AI literacy” encompasses, but we do know that critical thinking and language skills are essential, since modern AI is built on language. The better we understand language, the more we understand the behaviour of AI, and the better equipped we are to control it. Learning foreign languages helps us understand language in general. You don’t need to study a lot of theoretical linguistics to access that wealth of knowledge.
Strangely enough, to better control AI, we need more of what we’d think is “traditional education”, or general literacy: reading, writing, conversing, in multiple languages. We can't learn to think critically if we don't possess all of the above. All this together is the skill set you need to be able to judge the output of an AI, and it is the first step towards being able to fact-check its responses.
Another field where education needs to change is mathematics. In the parts where I come from, primary school math is abstract numbers, sets, shapes, fractions. It does not come connected with the part of reality where they are relevant. Mathematics taught in school must help kids understand money, traffic, phones... and AI. I don't know what that math education would look like but we need a paradigm shift here.
And when it comes to a university degree... Today's AI is so much intertwined with language that it should not be possible to achieve a degree in the humanities without learning a healthy dose of math that allows graduates to understand, even on the surface, how these systems work.
Reconciling the humanities and STEM
This also means that the boundaries between the “humanities” and “STEM” must become less rigid; maybe they should vanish altogether. For the longest time, I’d been watching how difficult it was to introduce technology into humanities classes: it always required a champion with special zeal. I have seen translators, teachers, historians, and other humanities scholars choose the profession to avoid mathematics. Like I pointed out above, , this should not be possible. Especially as I learned in the mid-2010s that this was probably an artificial cultural boundary, at least according to [10].
AI and social media have been transforming society and upending existing social structures. Together, they have the potential to dismantle meaningful human connections altogether, and erase all chances of democracy and open societies. If we want to prevent this, and we might be in the last hour to do that, we must start with renewing education.
Summary
To remain in control of AI and use it to their benefit, humans must learn AI literacy. This is a set of skills that helps us know when to use and when not to use AI; how to judge and correct its output; and how to adapt to its changes.
This requires critical thinking and insight into how AI operates. Critical thinking requires a high level of literacy, preferably in multiple languages. An intellectual must possess all of this, and that includes the technical prowess to grasp how an AI model works. For this, the rigid boundary between humanities and science must vanish.
Education must transform to accommodate this, starting with university programs. But this change should permeate into secondary and primary education, too. Only this will allow future generations to keep their freedom and their open minds. And the work must start yesterday.
Sources
[4] AI Literacy - AI and Teaching and Learning - Research Guides at University of Southern California
[9] Regulation that enables innovation (Deloitte)
