Welcome to The Martech Weekly, where every week I review some of the most interesting ideas, research, and latest news. I look to where the industry is going and what you should be paying attention to.
👋 Get TMW every Sunday
TMW is the fastest and easiest way to stay ahead of the Martech industry. Sign up to get the full version delivered every Sunday for this and every TMW, along with an invite to the TMW community. Learn more here.
Google has 178,000 employees and Microsoft, 221,000. But OpenAI has only 375. All three of these companies now represent the major force in the generative AI technology tech trend that’s taken off in 2023.
Yet somehow, with the release of ChatGPT, OpenAI’s consumer-friendly large language model chatbot, has caused the entire content economy on the web to change drastically. Call it timing, a smarter product or just dumb luck. But you have to hand it to OpenAI; they have seemingly cracked the code for a successful stand-alone consumer AI product.
On the heels of ChatGPT’s success, this week marks a significant turning point for the adoption of generative AI products, particularly for text. After investing $10 billion into OpenAI, Microsoft has announced its first real foray into using the technology with a ChatGPT-powered Bing search product this week.
Do you know what else happened this week?
Google also released its own version of an LLM text-based tool called BARD and, after a failed demo where the AI chatbot provided inaccurate results, Google’s stock plummeted by 7% wiping off $100 billion in a day. But mind you, this drop in public value doesn’t even make it into the daily top ten largest price drops in the public history of Google.
Yet, all signs point to Google launching an underbaked product in a rushed arms race to prevail over a hyped-up generative AI tech trend, a signal that there's a lot of value to capture in the trend in a winner takes all market.
The dehumanization hype cycle
Generative AI is perhaps the fastest trend to sweep the marketing technology industry in recent memory. It’s a transformational force happening at the intersection of experience, data, and marketing. So much so that ChatGPT has recently broken a record for the fastest-growing user base in history. It’s a little disorienting, after years of tech stagnation, to see so much change so quickly in the marketing technology industry.
The thrill of technological change in real time is exciting because there are new, shiny things to play with. But also, technology change at the scale of Google and Microsoft gives us a glimpse of the future. And although the COVID-19 tech trends were focused on a different kind of future, this time, there seems to be a lot of value in experimenting with generative AI, quite unlike the metaverse and Web3 concepts that were overhyped during the pandemic.
But there’s plenty of reason to suggest that AI is in its own kind of overhyped tech bubble. Take, for example, Buzzfeed’s recent announcement that the company is going to lean on generative AI for content creation. This is what that announcement did to its stock price:
Remember when Elon Musk was publicly backing Dogecoin? It’s a similar hype dynamic happening here where investors getting overly excited about unproven technology in new domains. In a lot of ways, the AI hype of 2023 is very similar to the hype surrounding Web3, but that technology was trying to leverage the blockchain (a decade-old technology) to do new things.
What we’re witnessing with generative AI, however, is the complete opposite in one respect – while AI has been in the background for a long time, nothing is as optimized or as seamless as ChatGPT in the way it uses natural language skills. But in another respect, this tech trend is just as dehumanizing as the Web3 movement. Web3’s central tenant was the idea that we should replace bankers and intermediaries from the financial system, seeking to replace human judgment with code.
ChatGPT and other generative AI technologies do dehumanization differently. By directly answering our questions, writing our stories for us, and creating images and videos, what it deprives us of is what makes us most human – the ability to create new things in the world. In a surprising twist of fate, AI’s purpose was always meant to eliminate human drudgery in an economy – transportation, updating spreadsheets, managing data, or customer service. But instead, the breakout-use cases are all focused on replacing how we create, the one thing that we’d want to do more of with our time in marketing.
It seems like whatever Silicon Valley is wanting to churn out is focused on replacing the good bits in your career with robots. But what happens when the cost to create content goes too close to zero? When creating, you don’t need to learn how to write, design, illustrate, edit videos, or make music anymore
The consequences of zero cost creation
The interesting thing about content – what we abstractly call the ‘output of human creativity’ – is that before the internet, only a small group of people had the resources and skills to create and distribute it. You had to pay to get your brand up on a billboard, on TV or in a magazine. It cost money, resources and was gatekept by an influential group of producers, media agencies, and brands.
But today, the cost of distributing content and getting it onto someone else’s screen has dropped close to zero. And sure, you can say that it still costs to get that content in front of audiences and that social and search web traffic is always a gamble for how many people see your content. But still, the point remains, like many of people have experienced, you can get content distribution with such a small outlay that it’s practically zero.
ChatGPT and other generative AI tools take the internet to its extreme logical conclusion by making not the distribution but the creation of content close to zero cost.
This will of course do several things to how online marketing works. The first is that it will divide marketing in similar ways to what zero-cost distribution did in the early days of the web. Over time with books like hacking growth and internet “hustle and grind” marketing personalities like Gary Vaynerchuck, a generation of young marketers and entrepreneurs who were more digitally native became convinced that leveraging social media algorithms, creating viral content, building products that did your marketing for you, hacking SEO, and doing conversion rate optimization are what you need to do to turn attention into money.
It’s hard to size up a cultural movement that was built on top of the concept of exponential growth through accessible and cheap content distribution, but searches for growth hacking peaked towards the end of the last decade, and now it’s on a slow decline. Like everything else in marketing, trends come and go. And this one is on the way out.
The other camp of marketers rejected modernity and embraced tradition. Mark Ritson is perhaps the most vocal about the lack of strategic rigor with the first generation of digital marketers, and has even called out the need to remove digital marketing teams as a separate department in business, citing the excessive focus on digital channels:
“Finally, a new breed of marketer who prefaces their title with the tactical term ‘digital’ has inundated our discipline with under-trained, overly tactical managers who have already selected their mode of execution long before any research or strategy has been even countenanced. They already have their crossbow drawn with no clue where, who or what they are attacking.”
While the previous generations of marketers are perhaps more balanced when it comes to approaching the content opportunity of the internet, there’s no denying that the ability to distribute content at close to zero cost created a new kind of marketer that was tactical, metrics oriented, and highly iterative. It also created a whole lot of grifters and fake influencers.
Now there’s a whole movement growing around generative AI with people who are promoting the technology as the next big growth hack. It’s the new hustle culture with people promising that you can make millions by finding a niche set of customers and, through using generative AI tools, you can offer them something that scales with close to zero marginal cost to create and distribute. This is creating another ‘hustle bro’ moment where everyone is telling you that you can get rich with ChatGPT. The Verge cites one example of a get-rich-quick scheme using the technology:
“Skim their advice and some patterns quickly emerge. One common scheme is to sign up to various freelance marketplaces — Fiverr, Upwork, etc — and advertise your skills writing blog posts and ad copy. Then, when a request comes in from a client, just plug their brief into ChatGPT and send over whatever it generates. If your client doesn’t like the results, just ask them for notes, feed those into the chatbot, and send the results back again.”
One of the unfortunate side effects of when you make things more accessible is that it becomes a target for crime, controversy, low-quality content, and an endless stream of click-bait articles produced by automated content mills.
The side effect of this is that everything becomes synthetic content, which means that people will start to doubt where this content is coming from and if its trustworthy. Removing the human from the content creation process will lead to more gibberish in news feeds, more noise, and less engagement. But because humans tend to operate on personality and status, we will find new ways to display quality and authenticity.
Verifying that a human is indeed writing that blog or article seems like an absurd idea, but on an internet where 42% of web traffic is fake, and where the majority of content is increasingly fake, there will be a way to differentiate the good from the mediocre. And how we do that will be by putting a premium on human writers, creators, and thinkers.
One idea put forward by Sid Jha is to develop an AI augmentation rating, similar to how diamonds are rated by sorting real stones from cubic zirconia. Don’t underestimate people’s needs to know what they are buying, reading, listening to, or watching actually came from another human.
AI and the creativity supply chain
The real question for marketers when it comes zero cost creation is how it augments the existing ways in which marketers create and refine content. One of my colleagues reached out to me this week saying that they have just written an entire landing page in minutes using ChatGPT. Another startup recently took me through their digital asset solution that not only uses AI to find images for you, but can modify them based on your inputs such as changing the race, gender, or background of people to suit regional-specific marketing campaigns.
What this does for seasoned marketers is give them an edge across the entire supply chain of marketing creativity. But there’s little reason to suggest that generative AI will completely replace search engines or valuable, factual content created by humans.
As Ted Chiang mentioned in a New Yorker piece this week, generative AI as it stands cannot replace the search for factual information. LLMs approximate words based on the frequency of their use with each other, so in a way, what OpenAI is doing is just creating a blurrier version of real facts, arguments, and prose. Chiang goes on to suggest that this might not be the best way to create original content:
“Obviously, no one can speak for all writers, but let me make the argument that starting with a blurry copy of unoriginal work isn’t a good way to create original work. If you’re a writer, you will write a lot of unoriginal work before you write something original. And the time and effort expended on that unoriginal work isn’t wasted; on the contrary, I would suggest that it is precisely what enables you to eventually create something original.”
A text-based tool can give you an additional skill set to brief an illustrator or designer, by mocking up concepts using a tool like Midjourney to “show them what you mean.” The technology can also help in the ideation phase, giving you different perspectives for a blog post. It’s also a more efficient content retrieval system for when you’re researching. It can pull insights out of a spreadsheet for you, but also provide feedback for a finished article. For marketers, there are a whole host of ways AI tools enable greater creativity.
But in reality, while there are fun and helpful in edge cases, there’s little chance that generative AI tools can completely replace marketing creativity. Rather, marketers will use these technologies strategically to speed up, refine or improve their content ideation, creation, refinement, and production:
Another way to think about the generative AI hype moment we’re in is to figure out who’s selling the shovels. When everyone is digging for gold, you’re more likely to make money by helping others find it. When the internet emerged into the mainstream, everyone wanted a slice of the opportunity. And companies like Google, Facebook, WordPress, and even hosting companies like GoDaddy were the ones selling shovels in the form of audience, access, and digital space.
Right now, there are a million and one “AI startups” that take a specific-use case, create a wrapper around it, and sell it as a hot new startup. But in reality, what’s under the hood is just a basic web framework with an API to OpenAI’s GPT-3 program.
Companies like Google, Microsoft, and OpenAI are the ones selling shovels now. The goal for the generative AI movement is not to create a commerce or media monolith like we’ve seen over the past 20 years, with the likes of Meta or Amazon, but to create a platform that owns the technology that makes all the other AI applications work. CB Insights suggests there are now billions invested in the various permutations of generative AI products in the marketplace.
But which categories are getting the most funding? That’s right: the companies focusing on generative AI for marketing use cases. The platform-model opportunity for AI startups is where all the leverage is. And right now it will be one of the three that will emerge victorious.
We’re now beyond the idea that generative AI will disrupt search; Google settled that with their announcement this week that demonstrated BARD in the experience of familiar apps and services. But in reality, Google’s in the business of selling shovels. And so the real opportunity in the technology is to create the next platform for generative AI so you can start seeking rent from all other AI startups.
In the original BARD generative AI demonstration, Google ironically favors the AI response in the experience, pushing down the human creators that published actual content in search results – a fitting metaphor for our time.
In this way, most of our experience of the internet has shades of dehumanization. Our person is reduced to a social media profile picture; our creativity is constrained to 280-character tweets or 10-minute-long TikTok videos; our content must conform to Google’s search ranking algorithm for it to get visibility; and we don’t even have much control over the kinds of email we see anymore.
Even the irony of OpenAI creating breakthrough technology with a fraction of the employees of the tech giants is not lost on me.
As the internet enabled anyone to distribute content for close to zero cost, this created a whole new class of marketer. But what kind of marketer will emerge out of the next cycle of AI automation?
In the new economy of zero cost content, the drudgery of content creation that AI solves is offered as a new format of liberation, but what is freedom without creation? Perhaps before we dive headlong into ChatGPT and other generative AI tools, it’s probably worth considering the real cost - losing our own ability to create for others.
Make sense of marketing technology.
Sign up now to get TMW delivered to your inbox every Sunday evening plus an invite to the slack community.
Want to share something interesting or be featured in The Martech Weekly? Drop me a line at firstname.lastname@example.org.