This is the first of a series we’re calling “Tech-Knowledgey” where we evaluate marketing technology and products and provide insights and advice on how to use them to grow and improve your work and business. Our first topic is what everyone is abuzz about, Artificial Intelligence (or “AI” for short).
Artificial Intelligence is taking over the news; quite literally, in some cases. Corporations are looking to integrate AI into their business model or product offering, with S&P 500 companies mentioning “AI” 179 times on Q4 2023 earnings calls this year. Artificial Intelligence is no longer a futuristic, sci-fi concept, it is here now. From the sound of it, AI is going to make its way into every aspect of our lives, whether we want it or not. But, is it really the all powerful, all knowing superior being that it is hyped up to be? It doesn’t seem so. Generative AI is more like Leeloo from The Fifth Element. It has the ability to learn quickly, but is still pretty dumb. It’s capable of doing a lot of damage, so we hope it doesn’t destroy us. Really though, it just looks sexy, and everyone is inappropriately trying to take advantage of it.
There is a term for this character trope in movies, known as “born sexy yesterday”. It’s a pretty clumsy phrase, but it is apropos to the characters it describes. Like the aforementioned Leeloo, or more recently Emma Stone’s character Bella Baxter from Poor Things. Physically they appear fully formed in a grown, adult body, but mentally they have a newborn’s mind. We see them awkwardly stumble and falter trying to navigate an unfamiliar world. Meanwhile, a small number of people tout them as miracles who will save humanity.
AI's fatal flaw however, is that it has all of the information, but none of the knowledge.
Generative AI is the real life equivalent of “born sexy yesterday”. Packaged as fully formed solutions, hailed as an expert at solving incredibly complex problems, when in reality they have no practical experience in the field they are being applied in. It learns and grows at a mind-blowing pace. AI's fatal flaw however, is that it has all of the information, but none of the knowledge. To use an analogy from another science fiction movie, AI is like Neo from The Matrix scanning all the recorded information and history of martial arts at super-human speed and then confidently turning to Morpheus and proclaiming (in your best Keanu voice) “I know Kung Fu”.
Much of the AI you encounter behaves this same way. It is given unfettered and continuous access to relevant source materials and it is instructed to learn all of it. Sometimes it even sources most of the World Wide Web. As a result, it appears to consider scientifically tested, researched, and proven factual information with the same weight as the opinions of anyone with a smartphone or computer. (On second thought, maybe AI is like humans.)
The expedient application of AI is showing its faults, already. Recent examples of Google's latest Gemini AI Assistive search results have demonstrated how human reasoning and critical thinking is not ready to be replaced by algorithmic machine learning. From the funny tip it provided one user that they “add ⅛ cup of non-toxic glue” to their pizza sauce to help the cheese stick better, to the more concerning suggestion that you “jump off the Golden Gate Bridge” if you’re feeling depressed. Another result recommended “eating at least one small rock a day… [as a] vital source of vitamins and minerals”.
(This image was generated by AI)
If you have used Generative AI to produce a portrait or video featuring “humans” then you have likely experienced the unsettling concept of “The Uncanny Valley” where the subject you’re seeing is very nearly human but something is just not quite right.
Does this mean we should ignore, or avoid AI? Should we destroy it before it destroys us! No. AI really does have incredible potential, but it should be used in ways that are practical. AI can be an amazing assistive tool, but it should not be used for all of your work, especially your most important work. Use AI to help do things faster and on a bigger scale. To simplify the question of how and when to use AI, we should ask ourselves “could AI do this better than a human?” Notice I said “better”. I didn’t say “faster”. Don’t fully apply AI for things that can be done BETTER by you. For Marketers: Writing, content strategy, visual design, content production, and branding are all examples of work that AI cannot do better. But, AI can assist to make that work easier and help complete it faster. Author’s often say “it is easier to edit a bad page than a blank page”. Let AI fill the first pages to get you started.
Below are some examples of practical use cases for AI Products and how they can assist in your Marketing work. This is not an extensive list, and is not a paid promotion of any products. This is meant as a primer and guide to evaluating what is practical uses for AI vs gimmicks, because as we mentioned at the top of the article almost every business is trying to sell AI “solutions”.
Writing - Creating outlines, generating ideas, drafting emails and removing writer’s block.
AI can be a great idea starter for writing. Products such as ChatGPT can be used to gather ideas, or even start an outline for a blog post or email. With simple natural language prompts, ChatGPT is really simple and easy to use. It has a free version, or the premium version ChatGPT Plus for $20 per month. If you’re feeling some writer's block, AI can be a great place to simply bounce ideas around. Just be sure you go back and make it your own before publishing. There are other paid services such as SurferSEO which has an AI tool specifically built to write articles that increase search value and rank for your site. Keep in mind, the increase in rank might work but Google’s algorithm is also very smart and will likely be looking for AI generated content to penalize soon. Plus, content quality is still very important for articles. If you’ve read this far, I assume you know what I mean.
Content Creation - Produce Concept images or videos, Generate storyboards, and Test Ideas
While I don’t think using AI for final production on commercial creative assets is a good idea (if not ethically ambiguous) Generative AI can be fantastic for starting projects where illustrations and imagery is needed. For the final production, Humans still produce superior, more realistic, and engaging visuals. That said… DALL-E 3 (which is part of the ChatGPT premium features), Midjourney, and even the latest version of Adobe Photoshop have Generative AI which can take your text descriptions or commands and produce images in seconds. These can be used for internal presentations, and even client pitches. Products like Miro Assist can help you produce Storyboards, when you need to visualize your commercial before going moving to production. Generative AI can be a fantastic assistive tool for producing concepts and testing your ideas before producing the final creative.
SEO - Improving Search ranking through Keyword Optimization
Search Engine Optimization (SEO) is still incredibly important for any business with a website, and this is another area where ChatGPT can shine. I actually think ChatGPT can be used for some final SEO work. For instance, asking ChatGPT to write a meta description for a product page it can return a fantastically written description for your product that is structured specifically for search. Other paid services such as SEMRush (a marketers favorite) also have AI tools for keyword optimization for body copy and writing meta descriptions.
Advertising - Contextual Programmatic Advertising and Targeting in Real-time
This is one of the areas where AI can do it better than humans. Real-time contextual targeting is becoming more important than ever, especially as cookie-based targeting deprecates. But it can be even more powerful than “traditional” digital ad buying which relies on data models that can be weeks old. AI powered ad buying can reach new audiences at the very moment they start their buying journey. Several Demand Side Platforms such as GumGum and Smartly.io have leaned heavily into this function. It is definitely worth investing some of your media dollars in this advanced targeting, however any Media Mix model should have some diversity so you can test and optimize based on click and conversion data. So, don’t put all of your media dollars into AI ad buying. 1P plus 3P audience data is still really important, and can provide better insights into content preferences, as well as help generate more highly qualified leads. Recency is crucial, but relevancy is also an important factor in advertising.
Lead Scoring and Loyalty Marketing
Lead scoring and loyalty marketing has already been relegated to the machines. Using complex algorithms and machine learning businesses have been putting vast amounts of data from multiple sources into a data lake and predetermined variables then assign scores to customers based on behavioral events. With the addition of AI, businesses will be able to identify more trends and opportunities in their datasets, and build even more complex and accurate profiles of their customers. AI will allow business to action on those insights for true personalization, and customer nurturing.
In conclusion:
AI has a lot of promise and potential. But, don’t trust it to work better than you, because it still has a lot of learning to do. Use it to do your work better, and find ways to let it help you get things done faster. Be wary of AI being applied to tasks or jobs that humans simply do better. For instance, don’t hire AI chefs to cook you meals because AI has no sense of taste it can only read descriptions about taste.
Or, perhaps the robots are in fact fully aware and have started to eradicate humans beginning with people who ask if they should “add more rocks to their diet”.
Thank you for reading this far along. I would love to hear your thoughts on this piece, or even other interesting or scary ways you think AI is being used.
Comments