If you had asked me to explain artificial intelligence or AI a few years ago, I would have laughed. I will talk about the Matrix trilogy and reflect on another Terminator sequel.
I'm ready to talk about AI today because I'm a professional writer who's currently sitting in a room with data scientists working on AI and I (sort of) understand what they're talking about.
I'm writing this because people keep talking about how AI-generated content is coming to their content marketing business. And that is simply not the case.
A Brief History of AI Content
Natural language generation (NLG) has been around since 1960S. It has been marketed for nearly 30 years. For most of its history, NLG wasn't very good. It can't read well, and the text it produces is easily identified as computer generated.
While it works for some stuff stuck to formulas or beats, that's it. The “content spinners” then rewrote the content in the hope that other computer-generated readers (i.e. Google’s search algorithm) would recognize the AI-written text as “new” content. ". But in the end, they're not that good either
A few years ago, new paradigms emerged to change AI's success, handling more natural language tasks on par with human creators.
I won't go into full semantics. As marketers, we don't really need to know the facts – we can leave that to data scientists . Bottom line: You no longer need to enter a strict set of rules and parameters for AI creators to produce quality content. Models now require much fewer rules for them to produce quality, and they can even interpret those parameters without human input. That change makes effective AI-generated content usable for a wider range of commercial applications.
You no longer need a strict set of rules and parameters for AI creators to produce quality content, @content_fairy via @CMIContent. #WHO Click to Tweet Simply put: Computers have learned to read better so now they know how to write better.
These advances have surprised many people. But it's time to go back to Earth because AI content creation is still here and that's okay. Listen to me.
AI is not as smart as you think
Let's reduce NLG to its most basic: You give it input. It reads that input and generates output.
To appear as if it was written by a human, someone needs to teach the AI software what a “human” looks like and determine the appropriate output. It requires hard work and human intervention to work. Think of it this way: AI is a student. Humans are teachers.
Example: AI language model GPT-3 has been hailed, revered, packaged, and commercialized in the last year or so. Developer and artist Arram Sabeti used it to create some sample text and posted it on the blog . They write: “I usually don't have to work so hard to come up with an amazingly coherent text with a simple prompt.
And yes, GPT-3 is awesome. It works really well in a lot of applications. But here are some exceptions:
- Writing longer than a few paragraphs is not very good.
- It can be beaten with short content. For each good variation, you get one to two points.
- Manual polishing is necessary to have good copy.
Will there be GPT-4? May. Will it be even better? I bet on that. But that will continue to be driven by innovation and human intervention.
At face value, AI-generated content will need to test infinite versions of infinite text to find the winners that deliver the highest ROI. That would require unlimited time and unlimited budget.
And that, dear human, is where you come in. You still need to handle the right message for brand voice audience yours personality yours , otherwise your content will be noisy. The real power of AI-generated content is helping to find the right language to reach your audience.
Impact on current AI content
I am a writer. I hate writing.
Okay, that's not 100% TRUE. I hate begin write. For me, a blank page is a portal to inevitable failure, not to mention a portal to completing horrendous quests I've overlooked. If you want me to sort out that junk drawer, just give me a writing assignment. On the other hand, if I could redact the content, I would feel less uncomfortable siding with a tortured anger procrastinator and more on the “Look at me, I'm a machine” aspect. write."
AI helped me start this article with an outline like this:
Item: What is an AI clone generator?
Part: Limitations of AI copy generator
Section: Why are people necessary?
Item: How Businesses Are Using AI Copy Generator
Draw result: Don't be afraid of AI, embrace it!
Supporting AI-generated content almost eliminated the dreaded blinking cursor for me. It can:
- Create something that I hate and I say, “Here let me show you how it has to be done”.
- Create relevant content that I can rework to fit the needs of the brand and voice.
- Give me a starting point in the form or an outline or an introductory sentence.
But there's so much more. Businesses can use AI-generated content tools to unify communication across departments and channels, such as email subject lines, website copy, ads, etc. save a lot of time.
The nonprofit Village of the Valley, a client of my company Anyword, creates a lot of content about what they do and to empower volunteers. Founder Brandi Powell found himself saying the same thing over and over again in different ways. AI-generated content has made it easier for her to find those new ways.
See AI is not an opponent but a teammate
Not sure if you noticed when typing in Google Docs, but sometimes it gets pretty complicated and it completes the sentences for you (if you agree by hitting the tab button.) And let's be honest – it usually ends your thoughts correctly. It's natural language generation in action, dear people.
Whether you're using AI-generated content to brainstorm ideas, say the same thing differently, or craft messages that score well on your demographic, you're making. You are still behind the wheel. The cruise control system gets a little more complicated.
Cover photo by Joseph Kalinowski / Content Marketing Institute