OpenAI’s O1 doesn’t like its own name. Here’s what he suggested instead.

There was a time when OpenAI fans thought the new model might be called “Strawberry” (for reasons only the most hardcore stans would understand). But in its press release announcing the o1, the company explained how it ultimately settled on the name.

“For complex reasoning tasks, this is a significant advance and represents a new level of AI capability,” the company said. “With this in mind, we’re resetting the counter back to 1 and calling this series OpenAI o1.”

It’s OpenAI’s latest branding mistake, according to o1.

We asked o1 to rate and give an “honest reaction” to his name.

Business Insider asked O1 to weigh in on its name with the following prompt:

“Pretend you’re a marketing or branding expert. What would be your honest reaction to an AI company presenting an AI LLM model name as “o1”? Give it a grade or rating on a scale of 1 to 10 and explain your rating and rating reason’.

o1 said he “thought for 12 seconds” before sharing his answer.

“As a marketing and branding expert, my honest reaction to naming an AI language model ‘o1’ is that it presents some significant challenges. I would rate this name 4 out of 10,” it says.

The model said the o1 name was too abstract, lacked brand identity and could cause trademark concerns. O1 also said its name could face SEO challenges and customers could confuse the “o” with the “0”.

While he might have human-like reasoning abilities, he doesn’t seem to have a human-like ability to give feedback gently or tactfully.

“The name ‘o1’ lacks the distinctiveness and communication power necessary for effective branding of an AI language model,” it said. “It could lead to confusion, be hard to find online and doesn’t help build a strong, memorable brand.”

So what should OpenAI have called o1?

BI asked him for alternative names. The model “thought” for eight seconds and offered 10 new options, including LexiAI.

“Reasoning: Combines ‘Lexi’ (from ‘lexicon’, meaning vocabulary or language) with ‘AI’, indicating an AI focus on language,” said o1.

The model also suggested Linguista, EchoMind, NexText, Dialogix, WordSmith AI, Verba, InnovaLingo, Cognitext and LuminaryAI.

O1 explained their reasoning for suggesting these names, citing relevance, memorability, trademark availability, SEO and pronunciation.

Finally, BI asked a third question:

“But what if the main advance of the model was that the AI ​​model could follow a ‘chain of thought’ and ‘reason’ in a way that previous models could not? What name would you suggest then?” BI asked.

o1 replied after 15 seconds of thought and offered 20 potential names.

“LogicFlow,” o1 said. “Reasoning: Conveys a continuous flow of logic, indicating advanced reasoning.”

Other names on the list were CogniChain, LogicLink, SynapChain, MindFlow, CogniPath and IntelliThread.”

As with the other two requests, o1 provided the rationale for choosing those names and suggested actions we could take.

“By selecting a name that encapsulates the unique reasoning capabilities of the AI ​​model, you position your product as a pioneering solution in the market,” said o1.