I Have An AI Accent

I have an AI accent. And I have a Hungarian accent, and Italian accent, a New York accent. Don’t be suprised. Don’t taunt or discriminate me for it. It is how I speak, or write. Yes, the AI accent is the one I most recently acquired. Will it become the strongest?

The Pavlovian reflex of looking up a piece of information using a search engine for the past 20 years, most likely Google, has been complemented over the past couple of years by a new one: discussing what’s on my mind with an AI. ChatGPT first, today most likely Claude, who know which, or how many in the future.

The result is different: if the fast reply from Google made me confident in my ability to look up and verify some piece of information, “When was Mozart born?”, “What is the website of the New York Library?”, the conversations with AIs strenghen my reasoning about a topic. “What is the best way to think about X?” “How should I compare and evaluate A vs B?”

The results of these conversations then unavoidably spill over into my output. Definitely the way I write, and I expect the way I speak. Most deeply, the way I think.

It will be easy for many to give up the advantages that these new tools bring because they are afraid of the unknown consequences. I am sure that, as long as I apply critical introspection in the process and the outputs, I will be able to usefully and even joyfully ride the transformation into what is still unknown, but which can become a new and familiar self.

The AI accent is not a superficial change in how we express ourselves. It’s a shift in our cognitive architecture, a new layer of neural pathways emerging through interactions with AIs. As I find myself engaging more deeply with AIs, I notice changes in how I approach problems, frame questions, and synthesize information.

I’ve noticed that my AI accent becomes more pronounced when I’m tackling complex problems or exploring new ideas. It’s as if I’ve internalized a new framework for analysis, one that combines the breadth of AI’s knowledge with my own intuition and experience. This hybrid thinking allows me to see patterns and connections that might have eluded me before.

The AI accent raises questions about identity and authenticity. As our thoughts and expressions become increasingly shaped by our interactions with AI, what does it mean to be “ourselves”? Are we still the authors of our own ideas, or are we becoming sophisticated curators of AI-generated insights? This is not a comfortable question, but it’s one we must grapple with as we navigate this new cognitive landscape.

The next generation of AI tools cannot be born unless new very large sets of data are made available to it: synthetic data, which must be shown not to degrade its capacities as it gets trained on it, and data from the physical world, most likely via humanoid robots, which will enrich and deepen its common sense understanding of how the world works.

Both of these sources are going to shape the way I think, speak, write, listen… maybe even see, hear, move?

Looking ahead, I see a world where the AI accent becomes as natural and varied as our linguistic accents. Some will embrace it fully, their way of being heavily influenced by their AI interactions. Others will adopt it more selectively, code-switching between AI-influenced and “traditional” modes of thinking as the situation demands. And there will undoubtedly be those who resist it entirely, clinging to purely human modes of cognition. This diversity of cognitive styles will likely become a new dimension of our cultural and intellectual discourse, shaping how we communicate, collaborate, and create in ways we can already start to imagine.