Can Writing AI Solve Your Organization’s Writing Problems?
Artificial intelligence (AI) for writing and content production has come a long way. Given that AI has been in the news, it may be worth revisiting the state of AI writing: can AI now offer a solution for businesses struggling with the quantity or quality of their writing needs?
Some seem to think so. The recent release from AI research and development organization OpenAI of an impressive new AI tool called ChatGPT, a leading-edge natural language processing algorithm that is capable of producing human-like responses to inputs and prompts, has made waves. The quality of its output has some commentators wondering if ChatGPT’s admittedly impressive content production is a harbinger of the end of human writers.
AI development is impressive, but it still falls far short of what would be needed to replace human writers.
ChatGPT and similar tools are legitimately exciting, but the technology isn’t as far along as it may seem at first glance. To start, this kind of tech development is an area where the 80/20 rule (sometimes called the Pareto Principle) applies, which stipulates that 20% of the inputs lead to 80% of the outputs.
In other words, even if ChatGPT and other writingai tools are already 80% of the way to being able to just replace human writers in business settings, the lion’s share of the development work and technical challenges still lie ahead. It will take 80% of the development effort just to close the gap and get the tool to 100%. This is partly why, for example, for the past 10 years people have been saying that self-driving cars on a mass scale are only five years away (and they still remain many years away).
Plus, ChatGPT’s output isn’t as sophisticated as it might seem. As Barron’s writes: “ChatGPT Is Amazing—and Totally Overrated.” These technologies just haven’t yet reached the point of being able to create new ideas or provide new insights. In fact, even if what ChatGPT and similar tools can do is eye-opening, what they can’t do is still significant:
- Truly create. These tools can synthesize, aggregate, and regurgitate; but it’s not the same thing.
- Be totally fact-based; ChatGPT will make up data points if it can’t source actual evidence.
- Understand or convey nuance.
- Understand or leverage human psychology.
- Judge its own quality.
As researchers at the University of Washington write, “[These] systems do not have any understanding of what they are producing, any communicative intent, any model of the world, or any ability to be accountable for the truth of what they are saying. An overlay of apparent fluency does not, despite appearances, entail accuracy, informational value, or trustworthiness.”
In fact, one journalist even tried to use ChatGPT to write an article for her. “I asked ChatGPT to do my work and write an Insider article for me,” she says. “It quickly generated an alarmingly convincing article filled with misinformation.”
So where does AI fit into content production?
Over time, AI will become capable of more content production and the limitations we describe apply today but will be overcome in the years (or decades) to come. Technologies like ChatGPT will probably catch up to human writers, step by step, and be able to handle increasingly sophisticated writing projects.
As a result, it’s simply true that many kinds of writing projects will eventually be fully replaceable with AI.
But the keyword in that statement is “eventually.” Given where we are today, conversation about AI is firmly on the hype train.
That said, AI versus human content production isn’t an either/or. Ultimately, AI isn’t going to outright solve writing problems nor is it going to replace human writers anytime soon, but it does offer tools and resources that can help make the process of writing a little less painful and a little more effective.
As a result, tools like ChatGPT should still best be seen as helpers and facilitators. They can give human writers a starting point, overcoming one of the biggest hurdles in writing: the blank page. They can help with brainstorming and ideation and maybe even with information discovery. As a result, with their assistance, human writers can work faster and with less effort – and that too will be a boon for businesses. It just doesn’t yet herald the imminent end of the need for human writers.
