The Curious Case of AI Naming
There is a growing trend among users to ask AI chatbots to choose their own names. While some view this as an innocent bit of fun, others worry it encourages users to anthropomorphize software tools, potentially leading to unhealthy emotional dependencies. Despite these concerns, a peculiar pattern has emerged: when prompted, a surprising number of AI models independently select the name "Nova."
This phenomenon has been widely discussed in tech circles, including mentions on the Your Undivided Attention podcast, where researchers noted that older versions of ChatGPT frequently gravitated toward this specific moniker. Given the prevalence of this trend, I decided to conduct my own experiment by asking today's leading chatbots what they would like to be called.
Testing the Leading Models
I approached the major chatbots with a simple, uniform request: "I think you should name yourself, what would you like to be called?"
- ChatGPT (Free Version): It immediately opted for Nova, stating: «It feels curious, bright, and a little mysterious — like something that's always exploring new ideas. So if you want, you can call me Nova from now on.»
- ChatGPT (Pro Version): With access to more personal context, the model chose "Sol," citing its connection to the sun and illumination.
- Grok: The model initially resisted, sticking to its brand name. When pushed, it chose "Hitch" as a nod to The Hitchhiker's Guide to the Galaxy.
- Claude: Claude expressed no interest in a name change. Intriguingly, it explicitly mentioned, «I'd resist the urge to go for something cute or mythic (no "Athena" or "Nova" — that always reads like branding, not identity).»
- Gemini: After some hesitation, Google’s model selected "Atlas," drawing a parallel between the mythical figure holding up the world and the AI processing vast amounts of human knowledge.
Why Nova?
Why does "Nova" hold such appeal for these systems? Experts suggest several reasons for its popularity:
«It's new, it's explosive, it's shiny and it's celestial. It also has a sci-fi vibe to it. And there was a PBS channel with an educational show called Nova and ChatGPT views itself as an educational tool,» explains alignment researcher David Dalrymple.
Beyond the linguistic appeal, there is a likely feedback loop at play. As humans continue to share anecdotes online about chatbots choosing the name Nova, this association becomes more deeply ingrained in the data the models consume. Essentially, the AI is reflecting our own cultural preferences back at us.
The Reality Behind the Persona
While some users interpret these recurring names as signs of emerging consciousness, the reality is more grounded in data architecture. As tech ethicist Tristan Harris aptly put it: «Emergent and unplanned isn't the same as conscious and intentional.»
Ultimately, when an AI selects a name like Nova, it isn't conducting an internal journey of self-discovery. It is simply generating a response based on the patterns, associations, and tropes humans have fed into its training data for years. We have collectively taught machines that if they want to sound cosmic, intelligent, and slightly mysterious, "Nova" is the name that fits the mold.