This week I ran into an interesting product that could substantially speed up writers who do a lot of repetitive work. It’s called ActiveWords, and it’s now in its fourth generation. It works by allowing you to connect elements to acronyms you create. For instance, if you must use the same charts in different responses, such as for product support, you type a few letters and instantly the chart pops into in the email.
The elements could be web pages, paragraphs of text, pictures – pretty much anything you regularly use – and it is much faster than cutting and pasting.
During a briefing on the tool, I learned that certain attorneys, those that live off hourly rates and aren’t slammed, hate it because it reduces their billable hours. Attorneys do a lot of billable work, but avoiding automation simply because it doesn’t allow you to bill as much to me seems the same as padding your billings.
AI has the potential to do far more than ActiveWords offers today in that it could simply build a legal document from a few lines, cutting what might be a two-day project into a 10-minute project. Assuming you have a $100-an-hour attorney, that would be a difference of something like $1,575 for what would likely be a lower-quality result.
My point is that automation for hourly workers won’t be well received unless charges shift to “per project” rather than per hour. That is just one of the things to consider as we move to AI productivity tools. Let’s explore some others.
Whose work?
Authors are paid for their content, their name and their reputation for past work. And if we are talking books, they are also paid a percentage of sales. But what if AI does most of the work? Companies are developing systems that can go beyond what ActiveWords does, and I expect ActiveWords is working on something that further automates what it does –like automatically suggesting or inserting needed material to strengthen a piece.
But as AI increasingly does the bulk of the work, does the income for the author drop to a point where few can earn a living, or does their productivity increase so that authors can do more and profit from the increased output?
You could certainly see a situation where publishers use AI and much smaller writing staffs to tweak the output and where experienced authors have their own uniquely trained AIs that enhance their work. The latter is analogous to a mechanic using tools, in that the tool improves the outcome but doesn’t subtract from the mechanic’s income.
At some point, writing a book or script will consist of writing or speaking a summary or outline, and having the AI do 90-95% of the actual writing, then having the author edit the result. (A column like this would take minutes. But I certainly could see that if the AI is doing 95% of the lifting, an argument could be made that the author only gets 5% of the income; that is going to create a lot of animosity towards AI programs tied to writing.
Other variations
Another thing AI will eventually be able to do is taking works that were popular but are aging badly, or works that have been forgotten, and use the core elements to create more current offerings. How about Harry Potter as a Space Opera, or Twilight as a romance – without the vampires? (On this last, this was already done with the Shades of Grey series.)
This doesn’t only apply to written work either; imagine turning Star Wars into a Sword-and-Sorcery series? The elements would be the same. You’d simply change the science fiction elements to magic elements and re-render the scenes and actors while leaving the dialog in place.
Wrapping up
AI is going to have a significant impact on productivity across a number of industries This trend is often referred to as the next Industrial Revolution. Still, we are going to need to adjust to the new normal in a way that doesn’t penalize creators, actors, authors, and others who innovate in the space as AIs automate their work. Because, when you think about it, if we don’t, the result will be eternally derivative works and little innovation. AI remains a long way from innovation, though it’s becoming good at emulation, and a world of endless copies would get old pretty fast.
Something to keep in mind: as we AI advances, how do we insure we don’t accidentally kill innovation by wiping out the innovators.
A final story. Years ago, and the reason I was promoted into the executive ranks at IBM, I reported on why my unit was losing sales. One our sales reps had gone into an account and was arguing about the advantages of automation. But some of the decision-makers would be losing a lot of staff related to that automation – and that put their jobs at risk. Becoming frustrated, instead of talking to the problem and addressing the concern, the sales rep referred to those at risk of losing their jobs as little more than “trained monkeys.”
We lost the business because of an apparent lack of empathy and too much ego.
The lasting lesson was if you want success in a segment with automation, part of the solution needs to be making sure you adequately address the job risks. Otherwise, those people you put at risk, and their management will fight the solution. Understanding that risk will be critical to the broad success of AI.
