Four Ideas to Turn ChatGPT Fears Into AI Learning for Your Business (Written by a Human)
Nothing has exploded into the public consciousness like generative AI. ChatGPT, the first to market from many development teams working on generative chat AI, reached 100 million users in just two months, faster than any previous technology. Marketers are racing to apply the technology, and many of those efforts will be misplaced.
Yes, ChatGPT and its AI peers can create ad copy & social posts, generate royalty-free but potentially plagiarized music and images, and synthesize audio versions of text documents to summarize the results of a web search. But it can frighten people with "hallucinatory" outbursts too, fail egregiously, tank a company's stock price, and undercut a well-crafted brand voice like a poorly trained customer support agent.
How can you use AI today to learn and prepare for the AI-enabled future? Your brand has boundaries and a personality; keep them at the center of your AI experimentation. People must shape the AI, curating, pruning, and advising to create a humane voice that makes the brand more relevant to the customers' everyday life and success. Your business is an expert in its field; your AI must catch up through extensive training.
Training the irrational tiger
You need to train a generative AI to speak in your brand voice. The center of a generative chat service is a large language model, a way of assembling words and concepts that can become a stochastic parrot that creates nonsense, propagates inaccurate information, or makes up facts in its attempt to sound human. These models are so large that they cannot stick to the message your brand intends to send.
In short, more high-skill work will be required in the AI-enabled enterprise.
Collect your internal product documentation, FAQs, support and sales interaction logs, marketing collateral, and even the company blog and your executives' speeches to begin. Content is the raw material of your marketing, customer experience, and data science teams must refine the AI to use it successfully through a process that combines linguistic, anthropological, psychological, and artistic skills with computer science, UX design, and application development skills.
ChatGPT explains how to train it to deliver a well-crafted domain-specific application:
Four fruitful strategies to explore AI capabilities in your marketing.
Each of our ideas focuses on a customer's needs, providing a closely defined goal for the AI, preventing the hallucinatory fabulism that has startled many Bing users and commentators.
1.) Content recommendations that use AI to identify ideas and issues of interest to your customers based on previous interactions. Relevance is the key to bringing prospects and customers back to your brand. AI has 30+ years of experience identifying and matching patterns that can be applied to analyzing your marketing, sales, support, and how-to content that matches your customers' interests. It can, for example, send emails, texts or app alerts about a new productivity tip to people who own a specific product or information about how to turn on a new feature of the digital dashboard of a vehicle. These touches generate measurable customer actions to assess efficacy, which can point to new engagement opportunities.
2.) Hand-offs from automation to human assistance via chat, voice, or real-time video are a natural application for AI, including virtual representatives that look human or like cartoon characters. Avoid making the AI the focal point – it will wear thin quickly without constant creative inputs because humans crave either humanity (your people) or novelty (new experiences), like Microsoft's Clippy or the Pets.com puppet. Stick to the customer's needs that connect them to your business and your people.
Based on a visitor's preferences the bot collects, the prospect can be directed to the right sales team, the customer to their account representative, the support team, or the product returns team. The AI's role here is to keep a visitor engaged while connecting them to the people who can solve for the reason the customer is reaching out.
3.) Extracting actions and insights from customer interactions, including calls. Another natural fit for AI is customer sentiment analysis. From word choice in chat/email and tone of voice during calls to understanding the tone of online product reviews, AI is an efficient way to quickly score interactions for follow-up, whether through direct outreach or by adjusting a content strategy.
4.) Content ideation and personalization. Any sufficiently complex content management system is the basis for rapidly personalizing articles, audio/video programming, support information, and other creative assets. Drupal creator Dries Buytaert writes about the CMS opportunity, calling it the best way to disintermediate the search engine between your message and the market.
AI can efficiently rewrite your existing content to precisely target the customer's interest. At the most superficial level, if you have a general support article about logging into all your services, an AI can deliver a version of the article that describes the service based on the specific customer use case. It could also translate the article into the customer's language.
In the long run, and under close human supervision, generative AI could draw on your entire business history and all your support and sales interactions to craft new content that delivers individualized guidance. For example, instead of navigating through a series of support pages that address different steps in a complex process, an AI could assemble and summarize those steps into a single, easier-to-use how-to narrative.
AI marketing skills are evolving
As the web gave us SEO experts, search algorithm savants and cybersecurity analysts, AI will raise a new set of skills, some of which will endure and others fade quickly. “Prompt engineers” who tune instructions for generative AI are the hot topic of the day, but they need to be supported by creative, branding, storytelling, sales, and support expertise to craft results that convert and retain your audience.
We admit that we’re learning, too. We’re building on a foundation that spans the gamut of marketing experience. Metaforce is the general contractor that can combine branding, campaign management and planning, and digital strategy with new AI-specific services that will take your brand to the next level. Let’s talk about the possibilities.