From Chatbots to ContentOps: How AI is Coming to Your CMS (Whether You Like It or Not)

“Do androids dream of electric sheep?”

Imagine Phillip K. Dick asking ChatGPT to answer this question. Would the Blade Runner films have been any different?

Unfortunately, the acclaimed science fiction author has long since passed – and might be turning in his grave. In his absence, we asked ChatGPT about those same sheep, and here’s what it said:

“The answer to the question ‘Do androids dream of electric sheep?’ is a matter of interpretation and speculation, as it is ultimately a metaphorical question about the nature of consciousness and what it means to be alive.”

Talk about existentialism.

Having artificial intelligence ponder its own existence seems a little scary. Turing Tests aside, we’re all wondering how smart these fancy algorithms really are – and if we should be worried.

Without a doubt, ChatGPT has catapulted AI into the social zeitgeist. Unless you’ve been living under a rock or scanning the horizon for spy balloons, you’ve probably seen it blasted across media.

But this fascination isn’t new. Just a few years ago, artificial intelligence was the toast of the town and part of the investment hype machine. Heck, in early 2018, Gartner projected the global business value derived from AI would reach a staggering $1.2 trillion that same year.

Of course, back in those days (cue the pre-Covid tumbleweed), the promise of AI was focused on automation, particularly at the enterprise level. From cybersecurity to big data, companies were banking on AI to improve customer experience, increase sales, and reduce costs.

These goals are still key to its value, but the game has changed.

Now it’s all about… content?

Well, not just content. But that’s certainly been the inflection point thanks to the wildly successful launch of ChatGPT. Suddenly, the average person could sit down at a computer, ask it anything – literally anything – and have answers magically appear before their eyes.

That magic was infectious. In its first five days, ChatGPT accrued over 1 million users, causing system outages and latency due to unprecedented demand. For context, it took Facebook 10 months to get to the same numbers.

But if AI isn’t new, then why is ChatGPT so groundbreaking? Because, despite all its mystification, it has singlehandedly demystified AI. It’s made an exotic technology feel simple, accessible, and beyond the domain of just large enterprises. And let’s face it: the thing has been fun to play with.

Just ask it about those electric sheep.

As ChatGPT was spitting out grandma’s time travel novella and writing your kid’s college thesis, it was quietly being engineered into high-utility applications with big use cases. In fact, over the last month, two of tech’s largest rivals – Microsoft and Google – started an epic battle royale over which platform will reign supreme in the new AI-powered era of intuitive search.

If you’re in the content game, you might be wondering how these new generative technologies will impact your content strategy. In this article, we’ll take a closer look, starting with the race to enhance search. We’ll also explore the efficacy of AI and some real ways it’s already transforming content operations – and your CMS.

BING-Bong! Traditional search is… dead?

Well, not yet. But the guillotine is looking mighty sharp. Here’s why.

Earlier this year, Microsoft announced it was investing billions into a partnership with OpenAI, the brains behind ChatGPT and DALL-E (a natural language image generator). While the stated goal was to improve the search experience for users, the company saw an opportunity to grab market share from Google, the undisputed king of search. As it currently stands, Bing owns less than 4% of the search market – a pittance in comparison.

The “New Bing” officially debuted on February 7 during a press event in Redmond, Washington where the Microsoft unveiled its next-generation large language model powered by OpenAI. Microsoft Edge, the company’s browser, will also incorporate the same functionality with an “AI copilot for the web” which they expect to become part of the overall Microsoft ecosystem.

So what might cause the guillotine to drop on traditional search? In a word: users. It all depends on how they respond and adapt to a natural language search experience. People like choice, and traditional search delivers a wide range of options based on variables like keywords.

But people also crave convenience, and a natural language engagement could shift perceptions. If we see the same groundswell of response that ChatGPT ushered in, Google could be looking at some stiffer competition.

The real question remains: will this window of opportunity give Bing an edge in search wars? Or will “New Bing” go the way of New Coke?

No-holds-Bard

When Google decided to name its new AI solution Bard, it was not without a sense of irony. In Celtic lore, a bard is a poet known for reciting epics about great journeys. In the oral tradition of Shakespeare, these stories would often have a hero… and sometimes, a fool.

On February 6, Google was wearing the jester’s hat. During a hotly anticipated press conference introducing its new AI to the world, Bard made a blunder – and it was a mistake that cost Google $100 billion.

After an underwhelming debut, Google is off to rough start in the AI game. They’ve been building significant hype around their answer to Microsoft for weeks, so expectations were already high.

Despite the faux pas, the company still boasts over 90% of the search market – and that ain’t bard.

Additionally, Google has been weaving AI into its suite of technologies for years, with some spectacular results. Google Lens, Google Maps, and Google Translate have been widely adopted, so Bard felt like another installment in their innovation pipeline.

Given its epic fail on such a large stage, many are wondering if Microsoft can claim more market share in the aftermath. But don’t count Google out just yet. The bard’s tale has just begun.

“ChatGPT, how will AI impact my content and CMS?”

Since the news of ChatGPT broke the internet (the one we’re currently familiar with), we’ve been as curious as everyone else. That’s led to a lot of testing (er, playing?) with the product.

It’s hard not to be impressed – and terrified – at the quality of the content ChatGPT can produce. From code to prose, blog posts to long-form narrative, it seems capable of it all.

If you haven’t tried ChatGPT yet, do it soon. The tool is still free to use, but it’s introducing a tiered subscription plan in the coming weeks.

To be fair, ChatGPT has some notable problems. Accuracy has been one of its chief detractors, and the challenges around regulating plagiarism and other social issues are far from solved. That said, it’s impossible not to think of ways this incredible tool could improve content management – and the broader implications AI might have for all things CMS.

So how did we explore that question? You guessed it: ChatGPT.

For kicks, we asked it to “provide a detailed summary of how AI is being utilized in content management systems today, as well as an outlook to the different use cases of AI for content management systems in the future.”

What this handy new tool generated was alarmingly decent:

Was ChatGPT right?

As “robo-summaries” go, ChatGPT was pretty on target and did a good job framing its response. Remember, it’s scaping data from a wide range of sources, and this is where the issues with accuracy come in. Additionally, the content was a little surface – which could make it easier to embellish and build upon.

Sure, the style and voice were a little dry, but we didn’t specify any tone in our request. We considered asking for an answer in the form of a haiku, but that would have limited the scope to a few stanzas.

Conversely, we asked our human editors to cross-reference this list and provide a few ways that AI is already impacting content operations – and in many cases, integrate with your CMS:

1. AI chatbots: Many CMS platforms already offer ride-along chatbot integrations that are incredibly intuitive and help streamline the customer experience. These tools can reduce the burden on people by creating more efficient workflows for answering questions and solving problems. There are also an array of third-party chatbot services like Intercom, which announced it had been building features using the new ChatGPT technology in a variety of ways. With one such feature, support personnel can now “summarize” a complaint or issue when sending it up the chain for review.
2. Automated image scanning and tagging: Humans forget all kinds of things, especially minute details. With accessibility being a key issue for websites, providing alt tags (alternate text) for images is essential – but often overlooked. It’s also important for maintaining good SEO. There are CMS platforms that can scan images and provide tags automatically, mitigating these accessibility and SEO issues. And by harnessing third-party cloud services like AWS Rekognition, most cloud-first CMSes can add this feature without significant retooling.
3. Smart Content Enhancement: By integrating with third-party tools like Grammarly, a CMS can enable a user to quickly improve their content quality. Other tools are also capable of analyzing sentiment, identifying keywords, and automating content categorization – making it easier for content managers to engage with their audiences.
4. AI-powered search: Much like the battle waging betwixt Microsoft and Google, CMS platforms are leveraging elastic search utilities powered by AI to enhance the user’s search experience. For example, Swiftype provides an AI-powered enterprise search platform that indexes data across multiple sources, helping employees or customers find the answers they need faster and without hopping between multiple apps or platforms. Voice-enabled search is also growing, with tools connecting to Amazon Alexa and Google Home to improve the search experience.
5. Predictive analytics: For years, Google and other services have been leveraging AI to augment their predictive analytics. Amplitude, a product analytics platform, offers powerful AI predictions that can help brands maximize lift by optimizing their targeting workflows.
6. Content generation: One of the most obvious applications for ChatGPT with a CMS is creating content – and some CMSes are already looking to do this in a meaningful way. For example, AI has provided some great insight on how a headless CMS is best suited for delivering AI-generated content in a modular way, with use cases for auto-generating FAQs, optimizing messages, cross-item summaries, comparisons, and more.
7. Generative imagery: When we say “content,” that includes images as well – and ChatGPT can instantly render illustrations and graphics in different styles based on the parameters you feed it. DXP platforms like Optimizely have added generative image engines to their experiences, enabling users to produce custom artwork and images for website pages, blogs, and other applications.
8. Marketing and advertising: While a CMS isn’t necessarily a direct hub for managing ads, websites thrive on a properly functioning marketing funnel. For years, Google Ads has been driving traffic using AI-powered capabilities. AdRoll’s retargeting also leverages AI to bring visitors back to specific products and pages, and marketing automation tools like Adobe's Marketo can improve targeting, progessive profiling, and more with AI features. And email systems like Sitecore's Moosend leverage AI to track and manage contacts and dynamically build segments for marketing campaigns.
9. Personalization: Still the “Holy Grail” of content marketing, personalization is becoming more attainable through a range of AI-enabled capabilities, from deep learning to natural language processing. Bloomreach and Iterable offer personalization engines that go beyond A/B testing to provide relevant content and experiences for customers at every stage of their journey. And platforms like Solodev offer cloud-powered services with AWS that provide machine learning recommendation engines that surface results based on user iterations.

What about the future of AI, content, and CMS?

In the movie Blade Runner, “replicants” – the android antagonists – are modeled after humans. They are nearly identical to us and capable of advanced cognitive processing. But they lack one crucial trait: the ability to feel emotions and empathy (spoiler alert: that’s not entirely true…)

In many ways, ChatGPT is emulating humans by building a natural language “bridge” between questions and answers. It's replicating our own responses.

But is it dreaming of electric sheep? Of course not. That said, in a few years, we can expect to see some dramatic leaps in its ability to produce natural language content for a wide range of applications. This could have a profound effect on how we manage content operations – and ultimately how our CMS helps us publish in an omnichannel world.

Here are a few things we might see on the horizon:

• Real-time accessibility and SEO enhancements: Imagine AI providing on-the-spot analysis of your content to improve quality, fix grammatical issues, repair links, and ensure accessible content – all without human oversight.
• Integration with natural voice capabilities: What if you could “talk” to your website through your home assistant? In the future, you could improve or enhance blog posts and articles in your CMS through a voice assistant and publish across channels in seconds.
• Automated personalization: If a page isn’t driving traffic, AI could step in and promote with programmatic finesse. That might be additional ads, A/B content, experiments, or other optimizations.
• Automated content analysis: Websites can be patchworks of missing content. That can impact leads and conversions. In the future, AI might be smart enough to spot the problems, alert humans or automatically fill the gaps, and instantly enhance performance.

From bards to bots, the future is automated

If this all feels like a round of Rock 'Em Sock 'Em Robots, it’s understandable. ChatGPT has thrown us all in a robotic ring, and many companies have responded to the hype by tacking on AI at a feverish pace.

AI is also coming to your CMS – whether you like it or not. Automation has been improving processes for quite some time, and we’re already relying on an ecosystem of third-party services that leverage AI.

For many content professionals, there’s a palpable fear and uncertainty around displacement. For example, artificial intelligence is emerging all over the sports industry, with generative content delivering real-time scores, in-game activity, and post-game analysis. These were once the domain of beat reporters on the sidelines.

AI might be displacing these activities, but the challenges around accuracy will continue to shade the trust factor. This is where humans can leverage the technology to improve productivity and enhance quality (think Grammarly) without relying wholly on the machines.

There are also ethical and legal questions about AI-generated content. If we can create commercials or children’s books with ChatGPT, where does that leave us from a trademark perspective? What about plagiarism? On the service side, will relying on chatbots help or hurt the customer experience?

As much hope is being put into these tools, the output of generative AI tools isn’t quite mature enough for prime time. This is especially true of imagery – which is a far cry from the beautiful original artwork being produced by humans.

In honor of the classic Rock ‘Em Sock ‘Em Robots, we asked OpenAI’s DALL·E app to create a “content battle” between humans and robots in a semi-realistic style.

Here’s what we got:

Maybe we can’t fight the future, but we can certainly put a sock in it for now.