Over the past few years, the increasing adoption of artificial intelligence by corporations has led to a growing discomfort in the workforce. Many (most?) white-collar workers are anxious, carrying a dull, constant and ever-heavy fear that their job might be the one automated next. Meanwhile, new workers worry they may never find a role in a quickly evolving economy. This is somewhat ironic: both office workers and students are themselves steady, even avid, users of AI tools. They see first-hand how few irreplaceable skills they have, thus feeding their anxiety in a self-actualising technology-enabled doom loop.
I’ve spent the past 5 years helping Fortune 500 / FTSE 100 / CAC 40 companies “implement AI”. This takes many shapes, across functions (HR, Supply Chain, Marketing, Sales…) and methodologies (Chatbots, custom algorithms, OpenAI wrappers…). Through it all, I’ve learnt that there is only one thing that algorithms will not replace : friction.
NB: The above statement is aimed at the professional sphere. Algorithms will of course not replace real creativity, human connection, small acts of kindness, etc. But you catch my meaning.
Friction is conflict, in thousands of shapes and forms. And conflict cannot be automated. It is too human.
It can take many forms. Physical friction is the easiest example, and the one often wielded in high-brow conversations about the “future of work”. “I have a friend who’s an electrician and they’re making a killing”, “we’ll always need plumbers”, “robotics is a lot harder to master”. And that is true. But those conversations often use friction and manual labour in the physical world interchangeably, and that is incorrect. And we do so because in the mind of the white collar worker, manual labour is hard. And we are lazy. So we cannot think of something that is hard, yet not manual.
A few examples of non-manual friction, which AI can do, but not as well as you, as long as you put your back into it :
-
Choose for yourself
-
Write good science-fiction (yes, even at work)
-
Be a friend at work
-
Tell hard truths
-
Read a book from cover to cover
-
Write something just for yourself
-
Be mean when needed
-
Make typos
-
Listen without speaking
But also (less proffessionaly important, and yet…) :
-
Go to a parent-teacher conference
-
Ask your local government to fill a pothole
-
Do something (anything) that won’t reward you
-
Make a sign, then hold that sign in a crowd
-
Do some civil disobediance
-
Participate in a mild rioting (my bio does say I’m french)
-
Kick a robot dog
-
Make really ugly art
-
Write jokes
-
Go to a different country
-
Make imperfect lists
-
Create a scrap book
-
Pretend to be someone else
In the “age” (lol) of AI, doing the legwork is where the value is. Sure an AI could likely do the above in 6 months, or 2 years… but by consistently doing it you are practicing muscles that are atrophying. This has value in the world, and this alone will have value in the work sphere.
We can see that we’re getting lazier and lazier with time; we may have illiterate college graduates. So knowing how to pay attention for long periods of time (e.g the ability to read a book), will put you in a good spot when the job hachet comes down.
hell, we’ve become so disensitised to laziness (ours and others’) that we’ve stopped expecting the real world to work. We assume it will be slow, broken, or possibly on fire. That’s just the baseline. But friction isn’t the enemy! It’s information. It tells us where things are straining and where care is needed and where attention should go.
I’m reminded of the Wall-e (the Pixar movie) — AI has made humans so lazy and squishy that they find it impossible to stand up to press a button. A button-presser has a lot of value in that society, should machines break down… be that button presser.
I will be difficult. That’s the point. You can’t beat an AI for most tasks. But whatever an AI can’t do, which is difficult to do is, gold. If AI could do it, it would already have (different beliefs on that!)
So create friction where you can. Seek it. What is hard is worth it.
