If you don't stop using AI I'm going to lose it

Around a month ago now, I got an email from Doha Debates inviting me (and anyone else who signed up at their booth during the Career Fair) to sign up for one of their upcoming debates. The question posed to the would-be debaters was as follows:

 “Superintelligence, Is humanity ready for the intelligence explosion?”

Putting aside that ‘superintelligence’ and the ‘intelligence explosion’ are both entirely meaningless and demagogue-like phrases, the debate topic (which should have been phrased, “Is humanity ready for AI?”) is topical enough, I suppose. Nowadays every company and their mothers are pushing ‘AI features’ onto their products to feed the ever-expanding LLM bubble. An AI chatbot here, some AI auto-fill there - just integrate some ChatGPT into the website and call it quits. It doesn’t help that we as a generation are all-too-willing to accept our new AI overlords, either - especially here at GU-Q. Even now, on December 1st as I write this, GU-Q hosted an event titled “AI, Skills, and the Future of (NO) Work,” playing into the insane narrative spouted by brainless Silicon Valley CEOs that AI is here to free us from our labor.

If I had a penny for every time a fellow student casually mentioned their use of AI to summarise a reading or to correct their essays, I could pay off the next two years of my tuition right this instant. It’s astounding how, as soon as ChatGPT became a mainstream and accessible tool, students on this campus latched onto it and will not let go. ChatGPT, Claude and their ilk have, apparently, become integral to the student workflow. Rather than skimming articles, looking for important points in paragraphs to focus on, students have turned to what is effectively glorified autocomplete to tell them what they need to know. These are skills that we, as a student body in an institute for foreign service, should have. How could you possibly not know how to read and formulate your own opinion? What do you mean you have to let AI tell you what to do? ‘Of course’ you AI’d your essay? Be serious with yourself for two seconds and realize how insane that sounds. Yet, people have somehow forgotten these elementary skills. So much so that my fellow cohorts are continuously surprised when I tell them I have not and will not use AI in any capacity. I will skim all my readings and I will write all my own words and if that somehow isn’t enough then I will reap what I sow and I will suffer for it. At least it was sowed by my own hands.

“But it’s more efficient!” I hear you cry out. “It lets you get more done quicker! I brainstorm with ChatGPT! I let it edit my work!” To that I say that you are robbing yourself of the opportunity to learn those skills for yourself. Letting an AI guide your thoughts and writing prevents you from having your own voice in your writing. You prevent yourself from forming the foundational skills necessary to carry yourself in this world. You cannot present your own ideas in your own way if you rely on AI to do it for you. And there is an addictive quality to AI as well. Since it is so easy to use and so convenient for what we do, it becomes easier and easier to rely on it instead of ourselves. It’s a spiral that starts with summarising an article and ends up with you asking AI to do practically everything for you. AI is such an easy fallback plan that it becomes hard not to use it once you’ve started.

And hey! - I know. I sound like a pretentious nerd or one of your professors. To that I say: tell me something I don’t know. More importantly, in this age you have to be pretentious about writing. To hold on to your creative agency, you need to grab on to the dorky, nerdy idea that the very act of writing even an essay is a form of self-expression. Hold on to it tight. AI tools like ChatGPT are attempts to govern the ungovernable, and we cannot let it continue to proliferate as it has. Excuse my flowery language, however we must remain lawless in this respect.

Don’t think I’ve forgotten about AI multimedia because I have a bone to pick here as well. Instead of commissioning an artist or editor for their skill in an age where capable artists constantly go starving, many GU-Q event flyers have resorted to using AI-generated images. Since this semester began, I’ve seen an AI-generated Jack the Bulldog repeatedly displayed on every screen in this campus. It’s an absolute travesty that GU-Q would rather resort to using AI than commission artists to create new Jack the Bulldog images in what I would assume is a cost-cutting measure. If the AI attack on writing is too pretentious an idea for you, at least agree that the AI attack on art is as plain as day an attack on creativity. This extends to AI images, videos and music - all of these lack originality, creativity and personality. They are replicants, created by a machine pulling strands from its library of stolen artwork and Frankenstiening them into something resembling a finished product. 

An AI cannot ‘think’. An LLM’s neural network is programmed to output the next most likely set of words in a string. There is no underlying understanding of your prompt, just an amount of tokens in a string to process. An AI image generator averages out pixels from a large library of images to create the most likely image. Everything about AI is replication of other, pre-existing things. You can argue all you want but at its core AI is a tool for replication - a bad one, at that.

What frustrates me the most about this is the general ambivalence our campus has taken over this. Because AI is a useful tool to many of us, we can brush over all its flaws and dangers. We claim to be revolutionaries speaking out against the world order yet willingly and freely give our personal data to OpenAI - a company on equal terms with Meta’s cartoonish evil and incompetence when it comes to handling user data. OpenAI has had massive data leaks and constantly sells personal user data to private entities. We care about the environment and saving the planet from a climate crisis but not enough to spare the thousands of liters of water used to cool AI server labs. We fight back against the imperial hegemon until they give us a new, ‘helpful’ toy to play with. Then, we play right into their hands. Maybe it’s a lack of awareness or maybe it’s our desire for convenience that trumps all else but we are ignoring the wider and deeply dangerous implications of AI use. OpenAI has your personal data. Every single one of your ‘conversations’ with ChatGPT is on their servers. Do you really want them to have that?

If the proliferation of AI means anything, it's that we need to re-examine our obsession with finished products rather than effort. This goes for the institution of education as a whole but also applies to us as a population. We need to raise our heads and recognise that a strong effort is just as important as a good outcome. Essays and presentations and images are not ‘products’ for us to use - they are things we craft. 

Look, I get it. Once you start, it is not easy to stop. Using AI is an easy way to get things done and I understand the impulse to use AI once you have started regularly using it is a very hard one to get over. It makes sense - if something can make your life easier, why not use it whenever you can? I understand. Here’s my advice: delete ChatGPT off your phone. Log out of your OpenAI account. Use Google. Skip over the Gemini AI overview and look at actual results from real websites. Even Wikipedia is more rigorously fact-checked than AI results. If you’re tech-savvy enough, switch to a different browser entirely. I’ve been using DuckDuckGo on Firefox both to protect my personal data and so that I can toggle off the AI summary. There are ways to get away from AI. It’s up to you to do it.