AI At Work: Generative AI and Commercial Writing
Generative AI has only been widely available for a short time, meaning that rank-and-file employees globally are still finding creative new ways in which it can support commercial writing workflows. From drafting outlines, to scaling research, to creating video scripts, knowing how to make strategic use of generative AI to aid content creation can give your marketing team the edge in a fast-changing industry.
In this piece, we’ll examine a range of use cases for generative AI, with special attention to commercial writing.
Core Use Cases
As defined by IBM, “Generative AI is a class of machine learning technology that learns to generate new data from training data.” In addition to mimicking training data, generative AI can also sometimes identify conversational context. However, current apps aren’t always reliable in this regard.
Where generative AI really shines is in meeting the need for speed. Apps using generative AI can help commercial writers and content creators quickly generate and manipulate text, audio, visuals, and more.
Here are some common use cases:
Ideation & Brainstorming
When building initial momentum for a project, an automated brainstorming partner can go along way. Generative AI can help writing staff overcome the blank page syndrome in the ideation stage by generating sample drafts on almost any topic. While the automated drafts themselves vary in quality, they can serve as a helpful springboard as your team hashes out a more focused draft
Preliminary Research
While any “facts” coming from a generative AI app need to be independently fact checked, AI can be very useful in pointing writers towards sources, suggesting lines of inquiry, and summarizing complex documents in simpler language. This can save staff a substantial amount of time, cutting down on the hours needed to dig up sources and skim through documents.
Outlines
The vast amount of unstructured information collected during research can be run through generative AI apps to create a sample outline within seconds. While your writing team will need to adjust this outline, perhaps heavily, to meet the needs of the specific audience and client, it can be an effective way to get the ball rolling.
Copywriting…sometimes
Generative AI can simplify copywriting in areas where copyright protection is not a concern. For instance, it can help formulate emails and job advertisements. It can also produce compelling messages and emails for private communication within the organization.
Chatbots
Features like chatbots can be tailored to convey information internally with employees and externally with the clients. Companies like ServiceNow already use automated chatbots to manage these repetitive tasks that demanded human attention before. While these have their limits, they can be a great first line of response for troubleshooting and other inquiries.
Design
Generative AI isn’t just for writing: it also helps with aspects of visual design, such as generating logos, posters, or even short videos. As with writing, the point isn’t for AI to do the job by itself: it is to speed up workflows by cutting down on time spent brainstorming, giving you a range of options which you can improve upon.
While offering significant assistance and time-saving benefits, it's important to recognize Generative AI’s limitations. Since it trains on human data, it is also susceptible to human fallibility. While generative AI can draw on scholarly information across various fields, it lacks the social and cultural contexts of everyday human experience to know how to best present this information.
Because of this, high-quality human oversight and editing are still indispensable. This makes human jobs more advanced and complex, since now they must review and identify content that may be inaccurate and incoherent, adding a premium to human effort.
Four Interesting Case Studies
While ChatGPT was a watershed moment in public awareness of generative AI, businesses have been anticipating this for some time. As early as 2017, five years before ChatGPT arrived on the scene, a Deloitte survey of 250 early AI adopters showed clear awareness of upcoming impact. 83% of respondents believed their companies had already received at least moderate benefits, and 76% believed that within the next 3 years, AI would "substantially change" their companies.
“In a 2Q2024 survey of 1982 business leaders and IT workers whose companies utilize AI, Deloitte found that 75% of respondents expect generative AI to impact their talent strategies within the next two years. Redesign (48%) and reskilling/upskilling (47%) were the most expected areas of impact.”
Job Descriptions
Using generative AI comes with risks, of course. In many regions, text and images generated by AI cannot be copyrighted, for instance. And there’s always a need for an editor who ensures that the final product is fit for purpose. However, for certain special use cases, generative AI can radically cut down on the time needed for clerical work.
CMC CEO Richard Pattinson has found just such a time-saving use of Generative AI. “Consider a document, such as a job description, where you already know the gist of what you want to say and need to make draft for internal circulation ”, explains Pattinson. “In such instances, generative AI can do the heavy lifting for you, potentially saving several hours of work”. He points out that the time saved is then available for honing the output and attending to other tasks. “As always, the trick is in supplying the right detail in the prompt — think of it as briefing your latest and greatest assistant.”
Scaling Research
Determining how to adapt a pitch to a specific sector can be tricky. So how about when you’re marketing a product to multiple different sectors at once? The research tasks quickly stack, and the amount of background reading required to get the gist of a particular market can easily lead to a deadline crunch.
According to Richard Pattinson, generative AI can help. “AI tools are great for rapidly scaling up editorial research”, explains Pattinson. This is particularly true when you have a topic “which is relatively unfamiliar to you but otherwise well-researched.” With careful prompting, generative AI can not only provide quick foundational explanations but also suggest interesting angles that deepen your inquiry.
Curating Material
Generative AI can do a lot more than just generate text. When given access to a quality data set, AI can also be used to curate, access, and overview large quantities of material. Just ask Morgan Stanley, who will apply generative AI to their Morgan Stanley Wealth Management (MSWM) products.
The firm is creating an “internal facing service” which helps MSWM financial advisors use OpenAI’s GPT-4 to draw quick insights from large quantities of intellectual capital. However they too are aware of the risks of over-automation. ”We believe trust-based relationships and human advice will always be valued by clients, and the financial advisor and their teams will remain the center of our wealth management universe”, explained MSWM co-president Andy Saperstein.
Automating Design
Moving beyond writing, generative AI can also unlock new possibilities for visual design. It can be used to generate initial mockups from text descriptions, or to simplify the modification of existing images. It also allows whole new sorts of branding strategies, as Nutella has recently shown.
The Nutella Unica project, a collaboration between Nutella manufacturer Ferrero and ad agency Ogilvy & Mather Italia, created seven million unique packaging designs for Nutella jars. They accomplished this by using a database of dozens of patterns and colors. These millions of jars, each with unique collector codes and designs, were reportedly sold out within a month.
Ideas For Applying Generative AI
Quality writing still requires a human in the driver’s seat to attend to accuracy and ensure output meets client needs. However, generative AI can speed up many aspects of content creation. Here are some specific applications:
Press Releases: Generative AI can draft press releases by collating official information meant for release through news and media outlets. It can summarize key points of information that need to be conveyed to the audience at large.
Initial Proofreading: When it comes to style, generative AI can serve as an initial check. It can analyze and correct documents for grammatical, logical and structural inconsistencies, helping to ensure professional quality.
Initial Drafting: Generative AI’s lack of appreciation for context hampers its ability to make creative drafts. However, seeing something done wrong can be oddly effective at revealing how to do it right. Generative AI can generate multiple drafts, which can then be rearranged and added to as needed.
Localized Content: Generative AI is highly efficient in translating content for localized markets. It can translate written content and create subtitles for video in different languages. The latest AI applications can translate content on the go in real time as you speak.
Interactive content: With its ability to quickly process and also personalize content to suit the user behavior, generative AI can enhance interactive content creation.
Part of the Toolkit
“The AI technology does what it does best, sifting through and processing copious amounts of data to recommend certain offerings or actions, and humans do what they do best, exercising their intuition and judgment to make a recommendation or select the best fit from a set of choices.”
The transformative potential of generative AI is obvious. With its ability to quickly generate drafts and mockups, it allows content creators to save a significant amount of time. Although human oversight remains essential to ensure accuracy and context and despite its limitations, generative AI is sure to make significant contributions across many sectors, not least content creation and business writing. Clearly, it will form an essential part of a content creator’s toolkit moving forward.