The Highs and Lows of AI Technology, Especially for Copywriters.
On the 10th of September 2022, Codie Sanchez tweeted:
‘AI will take more jobs from copywriters than it will plumbers.’
This is a scary but valid thought. AI is incredible.
The first time I ran into a copywriting tool, I was amazed by it. I simply typed in a few prompts about the intended topic and the type of copy I was looking for and voila, out came a perfectly scripted introduction to a blog post about NFTs. I was impressed by the articulation, but I noticed something about the tone. It did what technology is anticipated to do and produced something scripted.
With the remarkable introduction out of the way, I still had to dauntingly complete the post.
Most people cringe at the idea of writing copy. It is marketing art that only the gifted can wield. A compellingly written piece of copy has the power to attract digital era currency — attention. And attention has the potential to yield the real prize — margin!
So, being able to delegate this part of advertising to intelligent software is a breath of fresh, relief-filled air.
But according to a Forbes article on AI:
“Today’s AI still has fundamental limitations. Relative to what we would expect from a truly intelligent agent — relative to that original inspiration and benchmark for artificial intelligence, human cognition — AI has a long way to go.”
Before I elaborate on this point, I wanted to introduce a related technology that I seriously struggle with — the algo–fricken–rithm.
In a CMS Wire post, Dr. Mir Emad Mousavi, founder, and CEO of QuiGig explains the relationship between the algorithm and AI as comparable to that between a car and a flying car. “The key difference is that an algorithm defines the process through which a decision is made, and AI uses training data to make such a decision.” In other words,
“You can collect data from thousands of driving hours by various drivers and train AI about how to drive a car. Or you can just code it [to say] when [it] identifies an obstacle on the road it pushes the break, [or] when it sees a speed sign, [it] complies. So, with an algorithm, you are [setting] the criteria for actions.”
On the other hand, you “would not tell the computer what to do because AI determines [what action to take based on the] data that says this is what people almost always do.”
CMS Wire elaborated that ‘AI is also known for adopting unsavory behaviors, [by] failing to discern political, social, and at times, even objective correctness from incorrectness’ [my edits].
Going back to the quote from the Forbes article, algorithms and AI can expertly capture a procedure and its observed mechanisms but cannot predict the fickleness of human emotions.
I may have had a fleeting interest in something on a personal or professional level (authors and bloggers know this especially well). And at some point in your adult life, you’re likely to Google something completely random based on a present need. An algorithm can’t interpret that.
In all honesty, that accidental click on a link that explains the process of curdling milk was not an instruction to be replicated. Or that research I was doing as a freelancer probably won’t happen again because I don’t care about the number of socks sold in Papa New Guinea in 2017 (I can’t imagine many other people do either). But somehow, algorithmic programming will continue to suggest products and pages that I am no longer interested in.
Forbes confirms my sentiments in their June 2021 article by highlighting the things that AI cannot yet do. They touched on 4 main areas:
Use Common Sense
Human beings are both relational and have a history. These things make life an experience that is lived through imperceptible observations that do not need to be/cannot necessarily be written down. It is difficult to describe what the color orange looks like, how it feels to hug someone, or the taste and sense of fulfillment gained from a cup of morning coffee.
AI can interpret statistics but cannot make the interpretations meaningful.
“The absence of common sense prevents an intelligent system from understanding its world, communicating naturally with people, behaving reasonably in unforeseen situations, and learning from new experiences.”
Learn Continuously and Adapt on the Fly
A trained AI model is rigid. It can only perform a task for which it has been trained and for it to do something new, it must be retrained.
AI learning is constrained by ‘catastrophic forgetting.’ This happens when “new information interferes with or altogether overwrites earlier learnings in a neural network.”
“Imagine sending a robot to explore a distant planet. After it embarks from Earth, the robot is likely to encounter novel situations that its human designers could not have anticipated or trained it for ahead of time. We would want the robot to be able to fluidly adjust its behavior in response to these novel stimuli and contexts, even though they were not reflected in its initial training data, without the need for offline retraining. Being able to continuously adapt in this way is an essential part of being truly autonomous.”
Understanding Cause and Effect
Machines today are phenomenal at subtle patterns and associations in data but cannot understand the causal dynamics that are observed in the real world.
It would be possible for AI to recognize that people wake up in the mornings, but it would be unable to establish if people waking up causes the morning or vice versa. The field of statistics that makes up AI is built to understand associations and not causes.
“Mathematics has not developed the asymmetric language required to capture our understanding that if X causes Y that does not mean that Y causes X.”
Reason Ethically
I’m sure we’ve all run into the ever-growing prevalence of bots on our social media platforms. AI is said to be ‘altogether amoral.’ I’m sure this is our collective experience of them. They do not empathize with posts, instead, they’re missional.
Essentially, AI can ‘recite[d] toxic statements as a result of toxic language in the training data and on the Internet — with no ability to evaluate the ethical significance of those statements.’
The experts cited in the Forbes article are transparent that AI still has some way to go but significant progress is being made.
In the meantime, I guess we’ll have to endure algorithmic assumptions on our respective feeds with some exasperation. And grimace through the bots that plague our comments sections. And avoid unsolicited affiliate links to dodgy seminars.
And edit AI-assisted copy by adding some common sense, new knowledge, reason, and ethics.
And where we can, maintain the essence of our humanity.
I love to write, and as long as a machine has not mastered what we can do so eloquently, I will continue to trust my intuition. And write with all my heart.
Sources:
https://www.cmswire.com/information-management/ai-vs-algorithms-whats-the-difference/
