It took two and a half years for Airbnb, ten months for Facebook, and 2.5 months, for Instagram to sweep a million users.

Then there was ChatGPT.

Launched in November 2022, this technology took just five days, propelling generative AI and AI in general to the forefront.

And then the recurring question of how AI will take away all our jobs reared its head, as is the case with almost all intriguing tech innovations that changes the game.

The no-brainer short answer to the above question is: No, Nope, Nein, Nyet.

AI (and for the context of this article, ChatGPT and other generative AIs) can augment human credentials by automating specific tasks and freeing up time for more complex and creative work.

AI and marketing

The language model offers powerful capabilities for generating human-like text. 

This is a piece of old news. 

Marketers and the technology teams supporting marketing have used AI for years. It is no secret that machine-learning algorithms govern Meta and Google advertising. 

Also Read: The rise of Social+ 2.0: How in-app communities and AI are reshaping the consumer tech landscape

Even Amazon has used artificial intelligence to design personalised experiences for some time now. 

In general, however, personalisation will continue to be an area where AI can provide actual value. E.g., it can assist significantly in deciding the type of messaging and content to put in front of users. 

However, for several reasons, marketers should think twice before using it to develop strategies for embracing diversity and representation in branding.

The risk of perpetuating biases and stereotypes

Although, OpenAI, the company behind the first GPT and its subsequent versions (we are on GPT-4, in case you have missed the memo), added guardrails to help ChatGPT evade challenging answers from users asking the chatbot to, for example, take a dig or say a slur or commit crimes.

One primary concern with using generative AI models such as ChatGPT for developing inclusive branding is that the language model is trained on large datasets of human language; it can reproduce patterns of discrimination and prejudice in our society.

Woke-washing and lack of authenticity

Practices (or collaterals, campaigns, content insinuating ) in business that provide the appearance of social consciousness without any substance are labelled as woke-washing. 

Such content fails to engage authentically with issues of diversity and representation. 

Because the language model generates texts based on statistical patterns in an existing language, it may not capture the nuances and subtleties of diverse perspectives and experiences.

This results in the risk of creating content that lacks genuineness and fails to resonate with diverse audiences. 

Complex and multifaceted ethical considerations

Different generative AI types exist, such as text-to-text or text-to-essay, text-to-art or text-to-image, text-to-audio, text-to-video, etc.

ChatGPT and it’s most recent version, GPT-4, is an extensive computational pattern-matching software and data modelling apparatus, which raises critical ethical considerations when used to develop inclusive branding. 

Also Read: ChatGPT becomes the helper or killer to all occupations in Vietnam

Such tools can generate large volumes of content quickly and cheaply to the extent that the algorithm may inadvertently disclose someone’s personal information.

There is a risk that marketers may prioritise efficiency over ethical considerations, such as the need to obtain consent from diverse communities and ensure that their perspectives are accurately represented. 

The caveat

AI is powerful. It will continue revolutionising the enterprise landscape, impacting all streams, from marketing to human resources. 

However, its effectiveness relies on human intelligence. 

As a marketer, the insights provided by AI are valuable if they are fuelled by data and understanding and used to evolve marketing plans, improve communications, and drive positive change for both customers and the business. 

And, although ChatGPT and its ilk (Smartwriter.ai, Phrasee, Jasper.ai) may be entering into an eternal category of functionality, getting leveraged for varied use cases, the OpenAI admits that ChatGPT-4 still struggles with bias; it could even deliver hate speech. 

Of course, the tech still gets things wrong, as people will always cheerfully point out. It is far from perfect and most likely be a perpetual work in progress. 

But then again, so are humans. 

—

This article was first published on March 27, 2023

The post Generative AI and inclusive branding: Are we there yet? appeared first on e27.