Giving a name to AI: cautionary tales and advice for brands

“Every single company, whether a candy bar manufacturer or a software company, seemingly has to show that it is doing something to leverage AI,” says Jonathan Bell, founder and CEO of Want Branding. “And that often requires some kind of adjacent brand, which, of course, then needs a name.”

Several brands, as you may have noticed, have simply taken to adding ‘AI’ (or ‘.AI’) to the ends of their names. Think Stability AI, Spot AI, Mistral AI, Shield AI, People.ai, Otter.ai, Arize AI, Crowd AI, Toggle AI and so on. And, of course, there’s OpenAI, the company that has become something of a flagship for the entire wave of AI innovation that’s currently underway following its hugely successful launch of ChatGPT in late 2022 and that has probably helped to establish the ‘AI‘ suffix as the name du jour for up-and-coming brands looking to make a name for themselves in the industry.

Adding ‘AI‘ to the end of a brand or product name “is an easy but often perhaps a cheap way of doing it without much thought,” says Bell.

A parallel can be drawn between this naming phenomenon and a similar one that followed in the wake of the dawn of the internet in the late 90s when scores of new brands with ‘.com‘ at the ends of their names began to emerge. In those early days of the world wide web, it made practical sense for companies to make unambiguously clear that they were technologically savvy enough to have an online presence. (Remember, this was back when ‘online‘ was itself a new, hip word.)

Over the slow process of many years, however, the internet became so deeply embedded into most of our day-to-day lives, into the very fabric of popular culture and commerce, that it became more or less superfluous to add ‘.com‘ to the end of a brand name. Most people these days automatically assume that any given brand – unless it‘s incompetent beyond belief or run by a group of Luddites – has a website and probably some degree of social media presence.

The ‘.com‘ naming trend, in other words, began as a worthwhile marketing tactic, but “at a certain point that was eroded and it became meaningless,” says David Placek, founder and CEO of Lexicon Branding. There are still, of course, some brands (Hotels.com, for example) that have chosen to use their domain names as their official names, but such a strategy is far less common today than it was when the internet had the shiny-new-toy factor.

AI could follow a similar trajectory of cultural adoption as that of the internet: today, it’s all anyone can talk about; tomorrow, it’s basically taken for granted. Just as people today assume that brands today have an online presence – even when they don’t have ‘.com‘ in their names – we could soon reach a point as a society in which AI is so ubiquitous, so deeply integrated into our devices and our modes of working and communicating with one another, that adding ‘AI‘ to a brand or product name becomes passé. Placek says he’s “absolutely positive” that we’ll cross that threshold sometime within the next two years, after which point “everybody will assume that there’s something AI-related” built into most brands and products.

Given that forecast, adding ‘AI‘ to the end of a name “can be a disservice for building brand strength over time, because [the market] becomes crowded,” says marketing agency Tenet Partners CEO Hampton Bridwell. “There are a lot of names with a similar sound or styling and that creates a situation where you don’t have differentiation or memorability within the name.”

Anthropomorphic names and the sad tale of Clippy

There have, of course, been other naming trends that have recently emerged around AI. For example, many AI-centered products have been given human-sounding names, apparently in an effort to make the underlying technology – which could potentially come across as a bit threatening to a culture that’s been weaned on films like 2001: A Space Odyssey and The Matrix – feel a bit less alien and intimidating.

Consider IBM’s Watson, an AI model originally designed to answer questions that gained global fame when it won Jeopardy! in 2011. There are also more recent examples, including Siri (Apple), Alexa (Amazon) and Einstein (Salesforce).

As the journalist Charles Duhigg points out in a recent article in The New Yorker, Microsoft (which became a leader in the burgeoning AI industry following its recent multi-billion-dollar investments in OpenAI) has had to learn the hard way about the risks involved with trying to anthropomorphize AI. In 1996, the company introduced Clippy, a smiling virtual assistant with big eyes and a paperclip for a body, who could answer simple user questions on Microsoft Office platforms. The character became widely loathed by users. The Smithsonian called Clippy “one of the worst software design blunders in the annals of computing,” as Duhigg quotes in his article. Microsoft killed Clippy off in 2001.

The company once again tried its hand at anthropomorphizing algorithms in 2016 with the launch of Tay, an AI-powered chatbot whose conversational style reflected that of a typical teenage girl. Tay rather quickly descended into a fit of hate speech and was deactivated less than 24 hours after its launch.

Apparently wiser after the Clippy and Tay debacles, Microsoft is now naming its AI products in a manner that suggests utility and even a touch of fallibility. Copilot, the name of the company’s recently launched suite of AI-powered productivity tools, insinuates something that can be reasonably relied upon to provide a measure of assistance, not something into which one should invest one’s whole trust.

The curious case of ChatGPT

Perhaps the biggest irony in the realm of AI names is the fact that ChatGPT, the product that, more than any other, catalyzed the burgeoning AI Revolution, has such a widely disliked name.

For one thing, says Bridwell, the word ‘chat‘ in a brand name “is pretty limiting – it really doesn’t embody what the whole thing is about in terms of [how] it delivers value. It’s a terrible name. Over time, [OpenAI] should really think about rebranding it.”

Even OpenAI CEO Sam Altman agrees that it’s not an ideal name. During a recent podcast hosted by comedian Trevor Noah, Altman said that ChatGPT is “a horrible name, but it may be too ubiquitous to ever change.”

ChatGPT’s suboptimal name could stem in part from the fact that the OpenAI team that built it did not initially have high hopes for its prospects as an uber popular app. It was referred to internally as a “low-key research preview” in the period leading up to its launch and it was intended as a means through which the public could begin to interact with OpenAI’s GPT large language model more broadly so that the company could then collect feedback and fine-tune the technology accordingly.

Many within the OpenAI team were surprised when ChatGPT attracted its first million users in just five days, becoming the fastest-growing app in history.

Advice for marketers

According to Want Branding’s Jonathan Bell, brands that are looking to promote their use of AI through an optimized name should take their time. “It needs to be well thought-out,” he says. “It shouldn’t be something that’s done casually over a quick meeting, where you just simply add ‘AI’ to [the name]. Companies need to think about: What are they specifically doing? Can they deploy AI in a way that is really effective, or is this something that’s been done that could come across as bandwagon-jumping?”

Placek, who’s prone to referencing cognitive science and linguistics when discussing the psychology of brand- and product-naming, highlights the importance of sound symbolism – that is, the associations between particular sounds and the concepts that they evoke in the mind of the hearer. “You don’t want something too soft and you don’t want something too clever,” he says. “[You want something that’s] a little bit on the more serious side that [suggests] intelligence … sound symbolism should play a role in selecting and developing your names.”

When prompted to describe the qualities of a great name for an AI brand or product in fewer than 10 words, ChatGPT wrote: “Memorable, clear, unique, relevant, easy to pronounce, globally appealing, scalable.”

For more on the latest happenings in AI, web3 and other cutting-edge technologies, sign up for The Emerging Tech Briefing newsletter.