January 20, 2023
Like a lot of people right now, I am struggling to get my head around where we are going with AI. Look around and you’ll see the breathless excitement at the sheer amazingness of it all. And examples of its limitations, how it screws up, how military grade AI can be fooled by somebody hiding in a box, and its sense of humour failure. All of this is demonstrably, paradoxically true.
A similar superstate of unknowability swirls around how we might react as a species to what is unfolding. On the one hand you’ll find plenty of commentators telling you exactly what will happen, but their predictions typically rely on a set of faulty assumptions including blank-slatism and setting aside the effects for the majority. They’re not always allowing for the way some people will even literally stomp all over it.
Nor is it a relearning, as we have seen with the narrative arc of the Great Workplace Conversation. We are in genuinely new territory with this stuff.
People are not going to beat AI at that game
Some commentators have suggested that what we are seeing is only superficially remarkable because all the current systems do is flatten out what already exists and present it back to us. The commentator Tom Goodwin makes the point stylishly here.
“I still hold loosely the idea that for the time being GPT3 isn’t AI, it’s AAI -average answers immediately. Every time I’ve used it I’ve had zero trust in the answer. Whatever it tells me, I fact check, and since I see others not bother, I see it as the greatest creator and distributor of false fact or misleading information we’ve ever made.”
This is quite a claim when the laundering of bullshit is one of the defining characteristics of our time.
On the same theme, the record producer Rick Beato made a typically good point below, following a conversation with Billy Corgan. His general argument is that the music industry has already accustomed people to music that is based on averages of what they are used to. And that people are not going to beat AI at that game.
This isn’t a new idea. A few years ago, Brett Domino anticipated what was coming with this formula for the creation of an Ed Sheeran song.
It might be unfair to suggest that Ed Sheeran writes to a formula in this way but a more general consequence of the tendency to flatten creativity appears to be evident in the disappearance of key changes in popular music, although that could well be a temporary shift or something that doesn’t matter if some other feature of the music serves the same purpose.
I know I’m straying into old man yells at cloud territory here, but I agree with Rick Beato’s point that we have laid ourselves open to the full force of AI by buying into Harry Styles’ and The Weeknd’s autotuned facsimile of A-Ha rather than something new or ‘genuine’.
Nick Cave has already taken exception to the use of AI to generate a ‘typical’ Nick Cave lyric.
“Judging by this song ‘in the style of Nick Cave’ though, it doesn’t look good. The apocalypse is well on its way. This song sucks. What ChatGPT is, in this instance, is replication as travesty. ChatGPT may be able to write a speech or an essay or a sermon or an obituary but it cannot create a genuine song. It could perhaps in time create a song that is, on the surface, indistinguishable from an original, but it will always be a replication, a kind of burlesque.”
This averaging out of what we create applies in a number of other sectors, including our own. If journalists, bloggers and PR people think they’ll continue to make a living charging for cookie cutter articles and op-eds about ‘quiet quitting’ and ‘hybrid working is here to stay’, a bot that can produce one in a second is coming for them first.
Similarly, the creators of library images might also be in for a rough ride, and one they’ve made for themselves. We’ve already started to use AI imagery to illustrate articles, including this. It helps to eliminate the old problem of writing about workplaces whilst trying to avoid the hackneyed imagery that picture libraries seem to think people want.
Some image creators and copyright owners are already taking a stand against the way AI generates its imagery, including Getty who are suing one firm for scraping content as raw material.
As always, such changes present us with opportunities as well as challenges. Scott Belsky wrote a piece at the end of last year to explore how we might reframe what we think of as ingenuity in response to the tsunami of artificial creativity that has been unleashed.
Creativity is not just the output, it is the inputs
“As generative AI gets better at producing content, it’s important to remember that creativity is about far more than the outcome. The striking and wondrous thing about creativity is its mysterious seeds of origin. Do net new ideas come from genuine curiosity and initiative? Mistakes of the eye? Childhood traumas? Nobody fully understands the origins of ingenuity, but we know it is a function of the arrangement of our neurons and is as individualistic as our fingerprints. The creations that see the light of day in the form of pigment or pixels or breakthrough businesses are the result of these mysterious inner workings. Creativity is not just the output, it is the inputs — the ideas and the ingenuity. It’s the judgment to know when something is good and when it’s done. It is the creative control to modify and iterate based on a career of fine-tuned intuition. It is the unique human story that brought it to life, and the story we share that gives the work meaning to those who experience it. And it is the innovation in the creative process itself that distinguishes the outcome. As the process part of creativity — chipping away at the stone or mixing the colors or iterating the pixels — becomes less of an obstacle, the other parts of creativity — the original idea, the judgment, the innovations in process, and the story — become more important than ever.”
In a forthcoming book, I, Human – AI, Automation and the Quest to Reclaim What Makes Us Human, business psychologist Tomas Chamorro-Premuzic addresses the issue and appeals for people to seize the chance to leave the formulas to AI and discover something new about themselves. In a recent interview with the Irish Times he declares that the major barrier to this is to become aware of what we are dealing with and the environment in which we now exist and adapt.
“Each time we spontaneously react to AI or one of its many manifestations, we do our bit to advance not just the predictive accuracy of AI, but the sterilisation of humanity, making our species more formulaic. The mere fact that you may not be experiencing life in this Orwellian way highlights the immersive allure of the system itself, which has managed to camouflage itself as a normal way of life, successfully turning us into a rich record of digital transactions immortalised for AI’s posterity.”
Mark is the publisher of Workplace Insight, IN magazine, Works magazine and is the European Director of Work&Place journal. He has worked in the office design and management sector for over thirty years as a journalist, marketing professional, editor and consultant.