A few months back I came across the now infamous video of Reese Witherspoon talking about women and AI. Her basic message was that women are using AI less than men, and that women's jobs are particularly exposed to AI disruption. In a nutshell: women can’t afford to be “left behind”.
Now, let me start by saying I don’t generally look to Reese Witherspoon for technology strategy, or know that much about her at all, to be honest. I do know she’s done a lot of work supporting female authors through her book club and production company, which is partly why seeing her talking so positively about AI struck me as an odd fit. After all, authors and writers have been among the creative professionals raising concerns about generative AI and the use of their work.
However, amongst all the oddness, one thing lingered in my brain long after the video itself had disappeared off my screen. That phrase, left behind by AI.
Strangely enough, soon I started seeing it everywhere. It’s not just women who can’t afford to get left behind by AI, apparently. Workers need to reskill or get left behind. Companies need to become AI first or get left behind.
(From the sounds of it, AI might be moving at a pace that simply doesn’t suit the rest of society, but I digress.)
I’ve spent a large part of my career in marketing and communications, which means I know a fear-based sales pitch when I see one. And right now, I think fear is becoming one of the AI industry’s most effective marketing tools.
From FOMO to fear
The first wave of generative AI marketing relied heavily on what you might call fear of missing out, or FOMO. This is the positive side of fear-based marketing, where you tell people what they stand to gain from your product or service.
In this case, we heard a lot about how AI would transform your business, and that early adopters would gain an advantage. The possibilities were enormous and the people and companies that moved first would reap the rewards.
While there’s still plenty of that around, another type of message has become increasingly prominent. It focuses on the consequences of not adopting AI. This is when you scare someone into thinking something is going to happen to them if they don’t consume whatever you’re selling.
Thinking about this in the context of AI, we’re now hearing that companies who don’t go “AI first” will fall behind their competitors. If you work in corporate, you’ve probably consumed some version of the message that your career will suffer or you’ll even become obsolete if you don’t use AI. Essentially, you’ll get left behind.
Then there’s the version of this sentiment that I dislike most of all, perhaps because I question the inevitability of anything in this life apart from death and taxes:
AI is here to stay. You might not like it, but it’s inevitable. Get used to it.
When you think about it, this is actually extraordinarily effective marketing because inevitability changes the burden of proof in the commercial transaction.
Normally, if you want me to buy an expensive enterprise technology, you have to prove why I need it. You, as the person selling me the technology, would need to help me make the business case, which would include the anticipated return on investment.
On the other hand, if your approach is to convince me that your technology is inevitable, suddenly the burden of proof moves solely from the side of the person selling the solution. Your prospective customer is now in the position of defending the decision not to buy it.
After all, no CEO wants to explain to their board why they ignored AI while their competitors embraced it. Fear has started doing some of the work that the business case ordinarily would.
While fear-based marketing isn’t unethical in itself, in this case it should bother us because the business case is far from settled with AI, particularly in terms of its inevitability.
Yes, there’s reasonable evidence that generative AI can improve productivity in particular tasks. There’s much less evidence that those productivity improvements consistently translate into the enterprise-wide financial returns required to justify the scale and urgency behind every AI investment now being made.
Just because a tool can help one employee complete a task faster, it doesn’t automatically mean a company should reorganise its entire workforce and strategy around that tool.
And yet, here we are. Some very big bets are being made right now, with as yet untested returns.
The perfect sales loop
It gets even more interesting when the fear moves from companies to individuals.
There’s credible evidence that women use generative AI less than men. Women are also disproportionately represented in occupations exposed to AI-related disruption.
What the AI industry has done is taken those findings and turned them into a social issue. Lower adoption becomes a gender gap. The gender gap becomes a threat to women’s economic progress. And what do you know, greater AI adoption becomes the solution.
There’s just one problem with this reasoning. We don’t yet have credible evidence that women who use more generative AI get promoted more, earn more or have greater job security because of it. There’s a leap from A to B that’s yet to be proven.
Now, that doesn’t mean learning AI is a bad idea for women. What I have an issue with is the fact that we have turned an observation into a prescription without first establishing that the observation even needs a prescription.
And what should make you very cautious is the fact that the prescription happens to be commercially excellent for the companies selling AI… who are also the ones who created the problem, and are now marketing it to the world as the pressing social issue of our times.
It’s a conflict of interest of epic proportions. The technology creates disruption. The industry warns us about that disruption. The answer to avoiding the disruption becomes greater proficiency with the technology. It’s a near-perfect sales loop.
What makes it even more effective is that AI companies no longer need to deliver the message themselves. We’re seeing influencers on social media repeating it, and consultants telling businesses a version of the message. You might follow career coaches who are talking about it. Eventually we even start repeating it to each other.
That’s how you know marketing has done its job - when the sales proposition starts to feel like common sense.
What if we’re future-proofing the wrong skills?
As anyone juggling a busy life and career will know, professional development has an opportunity cost. Every hour spent learning one capability is an hour unavailable for another.
For someone with limited time, is learning the latest AI tools really a better career investment than deepening their professional expertise? Developing management capability? Becoming better at making commercial decisions?
Maybe it is and maybe it isn’t. We simply don’t know the answer to that yet.
There’s also emerging research suggesting we need to consider how AI is being incorporated in our professional development, if we choose to do so. There’s a difference between learning how to perform better with AI; and using AI to become better at the work itself.
AI can help someone complete a difficult task they couldn’t previously complete. Let’s say you previously had trouble writing well - now, you can produce written content with AI. Does that actually make you a better writer, or enhance your writing skills? Almost certainly not. This is because the technology itself can remove some of the cognitive effort through which you learn to complete difficult tasks in your own right.
In one randomised study, students with unrestricted access to ChatGPT performed substantially better while they had it. When the AI disappeared, they subsequently performed worse than students who had learned without it.
This isn’t conclusive evidence that AI will deskill the workforce over time. We simply don’t have enough long-term evidence to know what happens to professional development over a career.
Again, that’s the point. We don’t know yet. So why are we accepting the messaging about AI fluency as though we already know it’s the only way to future-proof a career?
When predictions start creating reality
When you create a narrative around inevitability and urgency without enough evidence, there’s a risk that the prediction starts influencing the decisions that make the predicted future more likely. (Yes, a bit of a brain bender there, I know.)
Tell workers AI skills will be essential and they prioritise them. Companies invest in AI training and redesign jobs around the technology. Work that once developed underlying expertise gets automated and workers become increasingly reliant on AI to perform it.
Eventually, AI proficiency really is essential, and we point to that reality as proof the original prediction was right. Except by then, we’ve helped create the conditions that made it true. And right now, the AI industry has an outsized role in shaping the decisions that will determine what the future of work looks like.
Do we really want to let that happen?
Inevitable is doing a lot of work
I say this often, but I’m not “anti AI”. I think there are many promising applications and I also think companies should experiment with AI and invest where they can demonstrate that it creates value, without losing sight of their broader obligations to society.
What I reject is the idea that because AI exists, the direction of travel has somehow been decided for us.
Nothing about how companies use AI is inevitable.
Nothing about which skills we choose to develop is inevitable.
Nothing about which parts of our jobs we automate is inevitable.
We still have choices. If we treat AI as inevitable, we risk turning those choices into foregone conclusions before we have enough evidence to make the right decisions on our own terms.
At the end of the day, perhaps Reese Witherspoon will be proven correct, and women who don’t embrace AI really will be left behind. Or the companies reorganising themselves around AI today will dominate their industries tomorrow. Maybe the workers spending their professional-development time becoming AI experts are making exactly the right bet.
I’d be the first to admit all of those things are possible. What I find remarkable is how quickly possible became inevitable - and how conveniently that inevitability benefits the very people who are selling us the future.
One thing I’m thinking about this week
I had one of those moments this week when you realise you might be close to making it as a content creator: people have been stealing my content.
To be fair, it’s happened before in isolated incidents. This week there just seemed to be a small outbreak of it across different social platforms.
My first reaction was irritation. My second was genuine confusion. As someone who takes my content seriously and would never knowingly steal someone else’s ideas or words, I don’t really understand the mentality. Have you no pride? Do you not want to have ideas and a voice that are actually yours?
It also felt strangely relevant to this week’s AI discussion. AI has made it easier than ever to replicate and repackage other people’s thinking. Which arguably makes original thinking more valuable.
There’s also a very obvious flaw in building your content around copying someone else. By the time you’ve spotted my idea, copied it and put it in front of your audience, I’m already working on the next one.
So copy someone else if you must. It won’t change the fact that you’ll always be two steps behind them anyway. 🤷♀️
Things I’ve made (and you can buy)
If you enjoy my content and would like to support my work, here are a few things I’ve created.
🛒 Track Changes On Shop - Coffee mugs and drinkware for corporate rebels
📖 Do Give Up Your Day Job - Guide to corporate exits and transitions to self-employment
