It’s been 700+ days since we first came across ChatGPT in November 2022. My first reaction? I LOVED IT! Of course, we all love things that make our lives easier. But soon after I was done playing around with it, I thought “AI is here to take our jobs!”. I’m sure many other people thought the same since we are a species that prides ourselves on our ability to think and have the ‘sixth sense’. In this article, I'll share my first-hand experiences of using AI to try to replace some of the tasks I do day-to-day as a marketing associate here at Rival, and what implications, challenges, and opportunities exist from my POV.
I remember how fascinated I was, using ChatGPT to make my life easier, and had the same thoughts as everyone else.
“It can write any poem I want it to!?”,
“How is it writing copy for 10 products with 30 variations in under a minute!?”,
“How is it doing SO much research in such a short time?”,
“This is MAGIC, I probably don’t need to work hard ever again”.
It was pretty impressive, especially from the lens of a generation that grew up without smartphones & social media. And then, fear set in. Will it take my job? Am I going to be unemployed soon? When I ask people about their thoughts on AI, I get wildly mixed answers (at least in marketing,).
Some people think that it’s going to make them millionaires. Others think that it’s a dumb technology that’s overrated. Very few say they don’t know how to use it. And most think that they will be unemployed soon.
The Current State of AI
Since 2022, AI has made great progress, especially in data analysis and content creation. It can do a good job of taking simple instructions and giving us back an actual, usable output. Naturally, this makes us wonder if AI will soon be at a stage where it can take over all the tasks that we marketers do.
Moreover, the efficiency and ability to scale tasks that AI brings to the table feels a bit threatening. Why? Because humans get tired. Humans run out of ‘creative juice’. Humans have mental blocks. Humans complain. Humans need vacation days. And, humans cost a lot of money. On the other hand, AI does not complain. AI does not have a bad day. AI does not worry about the workload. And AI can work 24/7 without charging you for overtime - which from a capitalistic perspective, feels like the best invention since the internet.
However, the reality is not so black & white. Marketing is not just about writing content, graphic design, or doing data analysis at scale. It’s about understanding human emotions and cultural nuances, areas where AI still has much to learn. At its heart, marketing is about connecting with people on a personal and emotional level, which requires empathy, creativity, and strategic thinking.
How AI is Changing Marketing
In a recent survey by Hubspot, 64% of marketing professionals said that they use AI tools in some form in their jobs, but the purpose and level of integration varied widely. Only 21% of marketers said it was extensively integrated into their daily workflows.
How are they using it right now?
- Data analysis/reporting (used by 40% of marketers).
- Research (39% of marketers).
- Content creation (38% of marketers).
Safe to say that right now, we’re at the breakout stage in the tech adoption curve. This means that sooner or later, AI will be a significant part of the way we work (whether we like it or not).
Personally, iIf you ask me, I use AI to summarize meeting notes, to get started with research, to come up with ideas, to write basic copy, or to retrieve information from long documents and websites. It is already integrated into my work life and is a significant part of how I work.
How AI Thinks & Works
To understand AI, it's helpful to first understand how ‘translating tools’ like Google Translate work. I’ve added this as part of the article to prove a point - that AI is not close to being as smart as us. You can continue reading if you want to understand how AI thinks (I know, this is the boring part where most of you will doze off). Otherwise, you can skip straight to the next section - ‘Why AI is not a threat to the marketing job market’
How Translating Tools Work
First, let’s establish that translating tools use ‘algorithms’ to convert text from one language to another.
Early translation tools used ‘rule-based methods’, which involved manually programming the language’s rules into the computer. This means that you would add rules one by one to make the computer understand & remember the meanings of different words. Modern translation tools (Google Translate) use machine learning, particularly neural networks, which can learn & understand language patterns from vast amounts of translated text. The neural networks are trained on millions of examples of sentences and their translations.
This involves feeding the network pairs of sentences (one in each language) and adjusting the network’s parameters to minimize translation errors. This process is called training. During training, the network learns to recognize patterns in how words and phrases correspond between languages. It develops a statistical model of these patterns, which allows it to predict the most likely translation for any given sentence.
For example, if the network sees the word "dog" in an English sentence, it learns that the Spanish word "perro" is a likely translation based on the patterns it has learned from many examples from its training data.
How LLM Models work (Large Language Models)
Large Language Models (LLMs) are one level above translation models.
They are a type of artificial intelligence that can understand and generate human-like text - like ChatGPT, and Bard. They are built using deep learning, a subset of machine learning. Specifically, they use a type of neural network called a transformer. Transformers are designed to handle large amounts of text data and can learn complex patterns in language. They consist of layers of nodes (neurons) that process input text and generate output text. Each layer learns different aspects of the text, such as word meanings and sentence structure.
To build an LLM, engineers feed it massive datasets containing text from books, articles, websites, and more (ChatGPT used as much data as possible from the entirety of the internet). The model then processes this text and learns how words and sentences are structured. It looks at how often words appear together, the context they are used in, and the relationships between different parts of a sentence. This process helps the model develop a ‘statistical understanding’ of language.
How Generative AI ‘Thinks’
When you ask ChatGPT a question, here’s what happens:
Input Processing: Your question is converted into a format the model can understand.
Context Analysis: The model looks at the context of your question, considering the surrounding words and phrases.
Pattern Matching: Using the patterns it has learned, the model predicts what comes next in the conversation. It calculates probabilities for each possible next word or phrase.
For example (see image above), if you start with "To sleep well," the model might assign probabilities for the next word like 0.25 for "you," 0.15 for "one," and so on (as seen in the image above).
Generating a Response: The model selects the words with the highest probabilities to generate a coherent and relevant response.
If you ask about the weather, the model knows from training that words like "sunny," "rainy," and "temperature" are likely to be relevant. It calculates the probability of each word fitting well in the context of your question and chooses the most probable one, mathematically. This is called Natural language processing (NLP). Okay, that’s enough with the technical stuff.
Why AI can't think like Humans:
Now that we know how AI works, we can confidently say that despite appearing intelligent, it does not think or understand the way we do. It doesn't have thoughts, feelings, or consciousness. Its responses are generated based on probabilities and patterns learned during training. Essentially, It predicts the next word in a sequence, one word at a time, similar to how a predictive text works on your smartphone’s keyboard when you type something, but on a much larger and more complex scale.
In simple words, generative AI (text) is nothing but a word-guessing machine. It guesses what you’d like to hear back based on the context in which you are asking the question. Its goal is to give back an answer, even if it means that it needs to make something up. That is why sometimes, even the most advanced AI models hallucinate, giving you rubbish when you ask for something simple.
Just look at the examples below:
Why AI is not a Threat to the Marketing Job Market
Think of AI in marketing like the introduction of the assembly line by Henry Ford in the early 20th century. Before the assembly line, cars were assembled by hand, which was a slow and labor-intensive process. The assembly line revolutionized manufacturing by mechanizing several steps, drastically increasing the efficiency and output of car factories. This is what happened after it was introduced:
- Increased Productivity: The assembly line allowed more cars to be produced with greater speed and consistency.
- Raised Expectations: With faster production capabilities, manufacturers were expected not just to produce more cars, but also to innovate continually and improve product quality.
- Expanded Workforce and Skills: The efficiency of the assembly line created more jobs in ancillary areas like parts manufacturing, assembly line maintenance, and quality control, expanding the workforce.
Just like the assembly line, AI in marketing automates repetitive and data-heavy tasks and reduces the time it takes to get to a certain point. This doesn't replace marketing jobs; instead, it frees marketers to focus on more strategic and creative aspects of their jobs. Automation and AI will increase the baseline capabilities of all marketers, shifting the focus back to creativity, innovation, and differentiation.
We are definitely not going to work fewer days in a week, neither are we going to lose our jobs to AI. If anything, more output is going to be expected from us. Agility is going to be one of the core qualities that us marketers will need to develop and embed in the teams we work with.
In a world where large output is demanded on strict timelines, this has come as a saving grace. Creativity isn't machinery, but AI is, technically, just machinery. It’s giving our creative minds a rest, allowing us to do what creatives do best—focus on creativity—and let AI handle the mass output. Things will move faster than ever - campaigns will go from ideation to deployment in shorter timeframes, creatives will be tested & optimized on a larger scale in real-time, and messaging will be iterative and highly competitive. Media buyers will focus more on optimization rather than spending hours creating reports. Content writers and copywriters will write more in less time. Graphic designers will bulk-create from their set brand templates. And strategists will spend less time on research and data analysis.
Example: Tomorrow.io
Recently on our podcast, Scratch: CMO Interviews, we interviewed Dan Slagen, the CMO of Tomorrow.io. Tomorrow.io has managed to disrupt the $4.7Bn weather intelligence category with just a 4-person marketing team and a six-figure budget.
They were one of the first companies to hire the "world's first AI marketer" - essentially creating a job description around all the AI-driven tasks needed for the role. Dan mentioned how AI plays a huge role in their marketing strategy. For every marketing activity, they sit down as a team and analyze the detailed steps involved to determine where AI can be leveraged versus using human resources. Even incorporating AI for 30-40% of the tasks helps them save significant time and allows their small 4-person marketing team to punch above their weight. What we can learn from Dan is that we need to build set processes and responsibilities if we want to integrate AI into the way we work every day.
This includes setting aside a dedicated budget for innovation and approaching it like any other business initiative—start by establishing a process, and continually refine it until it improves the team's performance.
Example: Image Creation
I use krea.ai to create blog post images for our content here at Rival. Below are a few examples of the images I’ve created so far (the workflow for this is pretty simple, I go to ‘Flux’ in Krea.ai, and paste prompts that I create from GPT-4o.). Why do I use AI? Because it’s much cheaper than licensing stock images for commercial use.
Here are some examples:
Example 1:
Zoomed-out, mid-angle drone shot, photorealistic picture of a group of working professional consumers standing in a crowd in bright pink clothes in New York
Example 2:
Create a photorealistic 4K portrait of a co-founder in a vibrant startup office, radiating inspiration and drive. The co-founder, wearing a bright pink hoodie, stands confidently in front of a large whiteboard filled with colorful marketing strategies, creative ideas, and dynamic diagrams.
The workspace is alive with energy, featuring team members engaged in animated discussions, laptops open, and motivational posters adorning the walls. Sunlight pours through expansive windows, illuminating the co-founder's face and casting a warm glow over the scene.
Capture the moment as the co-founder gestures passionately to the team, sharing a vision that sparks excitement and collaboration. Their expression should reflect determination and enthusiasm, symbolizing the power of innovation and teamwork in a startup environment. The overall mood should inspire a sense of possibility, creativity, and the relentless pursuit of success.
Example 3:
Create a photorealistic 4K portrait of a marketer in New York passionately unveiling an innovative concept, surrounded by their supportive team. Capture the essence of trust and collaboration as they confidently wear a striking combination of bright pink and dark blue clothing, symbolizing their collective creativity and vision.
Conclusion
The question here isn’t “Will AI take my marketing job?”, it is “How can I use AI to get better at my marketing job?”. The first step you can take right now is to get together with your team and ask them how they are currently using AI, and see how you can create a workflow that involves AI, just like Dan did.
Think of it this way, when everyone uses AI, everyone has a boost in their ‘skill-power’. The real competition isn't just about having AI; it's about who can harness its power the best. The widespread adoption of AI will not diminish the job market; rather, it will elevate the playing field. Everyone will get access to better tools, and the real challenge will lie in using those tools to create a better outcome than the rest. This is where creativity will become the most valuable commodity, meaning that our jobs as marketers become even more important.
We will have to keep learning and adapting, finding better ways to integrate it for our teams to execute the strongest marketing strategies.
#1 Pro-tip: Do not overcomplicate it, start with something simple. Every little bit of time saved with AI brings economic benefits to the company.
#2 Pro-tip: Never, ever trust AI completely to give you a good output. Go through AI-based outputs thoroughly for errors, as if you’ve written them yourself. You need to be able to take responsibility for the outputs you use.
#3 Pro-tip: Not everything can be enhanced with AI, so if you’re trying too hard to use AI for something and it’s not working out, you’re probably better off doing it yourself.
To conclude, AI is not here to replace us. Marketing with AI is only going to improve from here, and it is our job to figure out how to reach our target audiences better through the powerful tools that come with this emerging technology.
Frequently Asked Questions (FAQs)
How big is the AI market?
In 2023, the global market size for artificial intelligence (AI) was estimated at USD 515.31 billion and is anticipated to increase to USD 621.19 billion by 2024. By 2032, it is expected to reach USD 2,740.46 billion, with a compound annual growth rate (CAGR) of 20.4% over the period from 2024 to 2032.
Best AI Courses for Marketers?
- Hubspot’s AI for Marketers Course
- CIM’s AI in Marketing
- Artificial Intelligence in Marketing by the University of Virginia
Top 5 AI Marketing Tools?
- HubSpot for CRM, marketing automation, and analytics.
- Claude.ai for content writing & data analysis.
- Marketo, which uses AI for campaign management and lead nurturing.
- IBM Watson for predictive analytics and insights.
- Adobe Sensei for content personalization and automation.
The Best AI Tools for Digital Marketers?
- Jasper AI (content writing of all types_
- Lexica Art (blog thumbnails & cover art)
- Surfer SEO (for SEO-enhanced content)
- Claude.ai (advanced information retrieval & content writing)
- Content at Scale (for generating SEO long-form blog posts)
Which is the Best AI Tool for Market Research?
Zappi ActivateIt - The AI-powered optimization feature takes verbatim feedback from consumers and selected key performance indicators (KPIs) to revise your concept. It gives you new versions of what your concept description might look like if the feedback from your survey was applied directly.
Can I use AI in Digital Marketing?
Yes, you can use AI in multiple areas of digital marketing. You can use it while creating content, emailing campaigns, gathering actionable insights from campaign performance, personalization in marketing activities, and much more.
How to use AI for Marketing?
Marketers can use AI-powered tools to boost customer engagement, drive personalized marketing campaigns, analyze consumer data for better insights, and optimize their advertising strategies for higher efficiency and effectiveness.
Is AI useful for marketing?
Though the power of AI will help marketers do much more with fewer resources, the true competitive advantages will come from effectively integrating AI with human creativity and strategic thinking.
Disclaimer: Data collected by AI is subject to the LLM host. ‘We Are Rival’ bears no responsibility for any of the outputs generated through the resources mentioned in this article.
