- Feb 17, 2021
- 96
- 75
- Awards
- 4
- First Name
- Jake
Hey! Welcome to Local Marketing Insider where 7,000+ local marketing professionals get better at reputation strategy & more 2x a month.
Subscribe.
The topic of AI-generated writing content has been buzzing in the marketing world.
AI writing tools use a topic or input to produce copy variations automatically. They can brainstorm article title variations, write article outlines or expand on existing content by writing additional paragraphs.
According to search data aggregated by Exploding Topics, the search trends database owned by SEO industry leader Brian Dean, “AI Content” search volume is up 614% in the last 5 years.
Last week AI technology Jasper.ai announced a $125M fundraising round valued at $1.5B, which currently has 70,000 customers and aspires to double that in 2023.
We seem to be on the edge of a new era of AI-assisted marketing writing.
On the surface, the popularity seems kind of obvious. Writing good content, often, is hard.
Having something insightful to say is just part of the content marketer’s job.
You need to be creative on a regular schedule (for me it’s every two weeks), and you don’t publish something just once, each article needs to be broken up and reshaped a dozen different times for distribution among various channels and target audiences.
It’s a time-consuming process and business moves fast. AI can help you make it easier and more efficient.
It’s a potential remedy for writer's block. A creative lubricant and output accelerant.
But it also scares people. Am I reading something that a computer wrote? What does it mean to be informed and influenced by AI?
I believe the fear around AI writing tools is misplaced if used appropriately. For example, photographers have long had software tools to assist in the editing of images without issue. But this belief is grounded in the idea that AI writing tools are at most an assistant, not a full replacement for content creation.
To help explain this, I’ll start by covering some basics: how do AI writing tools work, what are they good/bad at and why are they a hot topic now?
How Do AI Writing Tools Work?
Using the vast database of content that has been published on the internet, AI writing assistants work by attempting to predict which words are most likely to go after certain words based on what humans have published in the past.Think of these tools as meeting a very-well-read friend at a coffee shop to help you brainstorm ideas. You can bounce ideas off them, explore new angles, and generate new phrasing variations.
Here’s an example outline generated by Rytr:
Input a topic or block of reference text and they will produce a rewritten version of that paragraph or an expansion.
I particularly like this example, the beta version of Lex, a word processor with integrated intelligence features, shows off its title idea feature:
What Are AI Writing Tools Good At? What Are They Bad At?
What they are good at: facts, history, eventsWe must keep in mind that the AI behind these tools is primarily powered by a combination of basic language training and the data available to them across the internet. Therefore, they are particularly effective at writing about widely agreed-upon topics - history, events, facts. Content with enough consistency and volume that the algorithms can be effectively trained.
For example, AI writing tools will do well with a concept like review gating, which is well-established with a consistent definition.
What they are bad at: debated, unique, opinion-based, creative, new topics
On the other hand, AI’s usefulness breaks down with topics that are debated, unique, opinion-based or creative. Additionally, newer topics or ideas that are evolving faster than the AI models can be trained will render the technology less effective.
As an example, the topic I wrote about in LMI #044, Google’s recently launched feature “multisearch near me” likely would not have the breadth of information available to power AI content on the topic.
Why Now?
Because of OpenAI.The data company launched the GPT-3 API in June of 2020, which makes it significantly easier for developers to build commercial apps on top of a prebuilt natural language processing technology.
Jasper.ai, which I mentioned in the opener, is built on top of OpenAI.
Instead of needing to build the technical infrastructure to launch an AI product, instead, companies can plug in OpenAI’s technology and package it up in a variety of commercially friendly use cases.
OpenAI has made the inclusion of intelligence significantly more accessible and affordable.
How To Use AI Writing Tools Most Effectively
Think of them like a writing assistant or recommendation engine, not a replacement for marketers.
Yes, technically you can input a topic and the tools will spit out an auto-generated article, but if you want your work to generate a lot of interest and attention, the automated output will not be good enough. The writing will be a bit uninspired, lack a strong point-of-view, and will likely have some phrases and sections that simply don’t make sense.
But AI writing assistants can be really useful when you’re stuck. They can help you brainstorm a variety of titles or point you in a new direction when you’re not sure where the next paragraph should go. These tools can spin up some new ideas you can use to get the wheels turning and hopefully kickstart your creativity to take over from there.
AI writing tools are not copywriter replacements. They’re supplemental - man + machine. As Fadeke Adegbuyi explained in her article An AI Might Have Written This, the tools should be considered “thought partners rather than creative adversaries.”
Don’t be distracted by marketing messages promising auto-generated blogs. I don’t think that is the core use case long-term.
When it comes to determining what makes for good writing and effective content, expertise will still be the deciding factor in creative fields, but increasingly, the ability to create the best work with the support and expert utilization of technology will likely be the most valued.
In photography, the experts of today are judged upon their creative utilization of digital cameras and digital editing tools - not their ability to create the same image with film and a dark room.
Writing with the support of artificial intelligence tools seems primed to be the benchmark for future marketing writing professionals, where competency + speed wins.
List of Tools To Explore
- Jasper.ai
- Lex (beta)
- Neil Patel AI paragraph rewriter
- Sudowrite
- Hemingway App
- Rytr
- Outwrite
- Copysmith
- Copy.ai
- Grammarly
See you in 2 weeks - Jake, Marketing @Widewail
Exciting news! We've revolutionized our review process with the power of ChatGPT and the OpenAI API. Say goodbye to manual posting, as our new online reviews are now automatically shared on Facebook in real-time.
