top of page
Writer's pictureIgor Alcantara

Does Autism Make You Better at "Talking" to AI?


As someone on the autistic spectrum, I've always felt a certain kinship with technology and artificial intelligence. There's something about the structured, logical nature of machines and algorithms that just clicks with the way my brain is wired. I know my initials (Alcantara, Igor) are AI, but that does not explain at all how I feel about this. I find myself wondering - is there a connection between the unique thought processes of those on the autism spectrum and our apparent affinity for interacting with AI?


It's an intriguing question, and one that I believe deserves further exploration. After all, we know that autism is characterized by a different way of perceiving and processing information. Autistic individuals often excel at pattern recognition, attention to detail, and logical reasoning - all skills that would seemingly serve us well in the realm of artificial intelligence.


Take the Moravec Paradox, for instance. This principle states that tasks that are easy for humans, like basic motor skills and social interaction, are actually incredibly difficult for machines to replicate. Conversely, the things that are challenging for humans, like complex mathematical calculations or chess strategy, are relatively simple for AI to master.


Now think about how this might apply to the autistic mind. We often struggle with the very things that come naturally to neurotypical people - navigating social cues, reading facial expressions, and engaging in fluid conversation. But when it comes to the more "machine-like" aspects of cognition - systematic thinking, attention to detail, and logical reasoning - we tend to excel. Could it be that this inverse relationship between human and artificial intelligence is in some way mirrored in the cognitive profiles of autistic individuals? Perhaps our brains are wired in a way that makes us better equipped to "speak the language" of AI, to understand its underlying logic and communicate with it more effectively.


When interacting with AI like ChatGPT and I don't get the answer I like and as I keep asking it keeps getting worse, I can easily identify when the AI is starting to "hyperfocus" on one or two specific details, forgetting all about the main picture. Identifying that makes me work with the AI to bring some other non-related topic, distance itself from the hyperfocus to then try to analyze the same problem with different lenses. It is like I am outside my body analyzing how my brain works.


I even tried the funny exercise of getting some of my old articles, things like from 1990's or early 2000's and upload them to a few AI Detect tools. You know, those websites people use to identify if a text was written by an AI or by a Human? Yeah, a couple of those. Most of my old texts were flagged by AI Generated content. However, I wrote them years before we had Transformers or LLMs. Does it mean I think like a robot? Am I C3-PO? Possibly not. Maybe may brain is simply wired in a way that is different than most people.


Of course, this is purely speculative on my part. As a person of science, I know a bold statement like this requires strong evidence, and I certainly don't have any hard data to back up these claims. But I can't help but wonder if there might be some truth to it. After all, many autistic people report feeling a deep sense of comfort and connection when interacting with technology, as if it's a language they innately understand.


If (and there is a big If here) there is indeed a correlation between autism and AI proficiency, I believe it could have profound implications for the future. As artificial intelligence becomes increasingly integrated into our daily lives, the ability to effectively collaborate with these systems will only become more crucial. And if autistic individuals possess a natural advantage in this regard, it could open up new avenues for us to contribute to the rapidly evolving field of AI development and implementation.


Moreover, understanding these potential connections could also inform how we approach the training and education of both AI systems and the humans who will be working alongside them. Perhaps certain teaching methods or learning environments that cater to the unique cognitive styles of autistic individuals could also prove beneficial for AI training. And conversely, insights gleaned from AI research could shed light on how best to nurture and leverage the strengths of autistic minds in the workforce of tomorrow.


These are just some of the intriguing possibilities that arise when we consider the potential intersection of autism and artificial intelligence. Of course, much more research and discussion would be needed to substantiate any of these ideas. But I can't help but feel that there's something there, some hidden connection waiting to be uncovered.


As an autistic individual, I've always been fascinated by technology and the inner workings of machines. And now, as I ponder this notion of a potential affinity between my own neurological makeup and the logic of artificial intelligence, I can't help but feel a sense of excitement and possibility. Who knows what we might discover if we start to explore this idea more deeply?


So, I pose the question again: Does autism make you better at "talking" to AI? It's a provocative idea, to be sure. There is possibly no correlation whatsoever. But perhaps, just perhaps, there's more truth to it than one might expect. This is also an invite for the researchers out there to collaborate and help us understand this better (if you want to help, fill out this draft questionnaire and share with others), and, moreover, to other people in the spectrum like me: do you feel the same or am I just the weird exception I always felt to be?



49 views0 comments

Comments


bottom of page