• Are we all editors now?

    As the use of AI becomes more widespread and the need for high-quality content increases, is the role of editors surpassing that of creators? This may seem to contradict the belief that AI will make editing obsolete. It’s almost a fantasy delusion that all our content will come out perfectly formed and magically ‘just how we wanted’ – that is a very questionable vision and at least feels quite far away. For the time being, many of us will be primarily responsible for editing content generated by AI.

    1. a person who is in charge of and determines the final content of a newspaper, magazine, or multi-author book.

    “the editor of The Times

    2. a computer program enabling the user to enter or alter text.

    “you can use the editor to make any changes you want.”

    Starting with editing

    Editing at the start comes in framing the prompt to create the content itself and the tone you want it created in. For example:

    <what you wish to generated> <in the style of><with focus on>

    At least now and briefly, you will get reactive prompt-based engineering for the content model. That means framing the start and the input is vital, and your edit is critical to the output. ‘In the style of’ and ‘with focus on’ are key beyond what you wish to be generated. They are where the art of editing comes in.

    Editing your content

    Even if you create fresh content not generated by AI, that content can be edited. This is fascinating, and actually, one spiral of tools as far as editing goes that I think is the most useful today as a use case. Tools, like Go from Grammarly (many others), parse your content and recommend.

    Whilst they are the editor, the results are somewhat variable, and you rarely can accept it without at least reading and interpreting. This accuracy can degrade depending on the type of content; for example, the more technical, the worse often the recommendation.

    Ending with editing

    If you do generate the content, it again will need editing. This goes along the same line as the tools to edit your content; the more narrow the field of knowledge, the often wilder the interpreting today. It’s getting better, but the generation output can be fascinatingly slightly off today.

    Anyone who has used AI tools to create content or support creation knows it’s not perfect – although that is leaping forward at an incredible speed. It also is a matter of learning as you work with it. It adds to your flow of editing and boosts.

    Reviewing (V) To examine or assess something with the possibility or intention of instituting change

    Editing (V) To prepare written material for publication by correcting, condensing, or modifying it. 

    Are we not just reviewers?

    Perhaps one change is seeing this as reviewing, not editing. There is a difference, after all. In an ideal world, as content improves, the flow would be to check over and edit more. The system and tools you use would be learning and knowing your field and terms. Generated content would more than often be more correct and need less editing.

    Editing is a natural state

    It might be a bold statement, but how AI editing tools are positioned does play to human strengths. Humans are pattern-recognising machines set up to parse and review. These tools that support different workflows and adapt when needed to fit whatever content curation we want today are snowballing in popularity because they are taking friction tasks out of our way. That’s another post, but that’s a massive point of why they work. They don’t change our workflows so much that they become unrecognisable; they take what works and build on it.

    The actual craft of editing is an art form, so learning those incredible skills is something to focus on in the future. Crafting down a prompt, simmering meaning, and getting to interpretation will unlock so much. The art of prompts isn’t in the ask; it’s in the refinements, for example, knowing how to edit and ask for style or the prompt itself.

    You could argue that we are all becoming more editors, but we probably were naturally more fitting to that role and needed to learn our strengths. It’s just at what stage in content curation we’ll be editors and whether some of us ever become anything else.

    Next editing

    Currently, there are various predictions about the future, and I am not fully convinced of the ones where humans are entirely replaced in content creation. Upon reflection, I believe that some degree of editing will be necessary at least for a while. However, with improved tools, this need for editing will likely decrease. The focus should be on creating tools that facilitate outlining in editors, support editing workflows and simplify these tasks. AI tools that take the pressure off tasks today and easy burdens felt now are instantly the most successful for a reason. Think about the housekeeping robots of the movies or cartoons – what are those pieces of editing workflows that AI can do for you? That is an opportunity.

    I believe there is a significant potential for assisting with editing tasks and identifying the need for editing. This includes providing suggestions for improvement and adjustments, as well as recommending content once it has been generated. Even basic tools that analyze writing style can be trained to identify areas for improvement beyond just spelling and grammar, old ones extended and new ones created. As the production of generated content continues to increase, I foresee the demand for editing support growing and this a strong opportunity.