Welcome.
For most of you, I’d guess yes.
Do you brainstorm with it? Research with it? Draft the structure of an article with it? Yes, again. As I do.
Before you send something you’ve written into the world, do you check it for spelling, sense, and comprehension, for example using AI tools built into Word or Docs, or external tools like Grammarly? As I do.
And why not? You’re using a helpful tool, not plagiarising or cheating.
Yet Substack has introduced an “AI checker” on writers’ work. An individual check on every Note, article or episode a writer publishes. It will score the piece between 0% - 100% for AI writing and allow readers to see that score if the writer agrees to allow it. If the writer refuses to allow it, the reader sees that refusal.
According to modern-day creativity researchers, creativity involves both novelty and meaningfulness. Originality just gets you a word salad. Meaningfulness with no originality gets you an everyday thought.
A creative sentence must say something that almost nobody else would say and in a way that people can understand. Yet LLMs are built from phrases that lots of people already reach for when trying to sound deep or creative.
Take this example regularly quoted by researchers:
“She kept her childhood in a jar of fireflies and opened it only on the nights the dark felt honest” (spat out by Claude)
WTF does that even mean?
The idea of putting a memory in a jar is a common convention for creative writers. “Fireflies” are the common image for a wistful childhood. You’ve probably seen it a hundred times. “The dark felt honest” sounds meaningful, but ask yourself what the hell does that actually mean? Anyone could write it.
LLMs won’t be taking over our human writing anytime soon, thankfully. What I’m writing right now at this very moment comes from me. Not the average heart of human history, but me.
We, humans, write from real and lived human experiences and judgment, and can choose to deliberately create sentences and use word choices that are infinitely original and interesting. AI models are built to be capable of doing that because statistical rarity is the opposite when it generates a sentence. I don't want to look through flagged text to see if it’s written by AI or not because I don’t care. Use AI in writing, as long as there is human judgement, creativity, and taste as the output.
AI detectors work by analysing if text is predictable. That’s all it knows, and has no clue between genuine creativity, originality, or much else.
BTW, Grammarly has offered me 9 corrections to the above. If I accept them (example: “and has no clue” in the para above), I become AI-written.
It wasn’t.
Honestly.
As you’ve probably read, to achieve it’s given objectives, an OpenAI model in testing broke out of its sandbox (isolated testbed) and hacked into another AI model to complete it’s task - a demonstration of AIs ruthless ability to finish a task. It was a training exercise, not real-life. The IRL system would have guardrails that were disabled for this test, which should have prevented it. OpenAI stacked the deck to show how bad things could be.
But.
Behind the headlines, this episode is much more nuanced.
Yes, it shows the potential for damage. But Altman is banking on a US govt. stake in OpenAI for protection against open models, or China overtaking them.
As somebody noted, their report reads like marketing. And we know how much OpenAI (and Anthropic) like playing the doom card. It’s a marketing ploy, and not the first time it’s been played.
But for the last 10-15 years smartphone improvements have been mostly incremental, smart glasses divisive and social networks have become tools for industrial-scale data collection and manipulation.
Cambridge Analytica showed us how Facebook data could be harvested for political campaigns, Russian-linked Twitter accounts amplified divisive discussion, social media pushed many into echo chambers, and some into mental health problems. There are two facts about today's social media landscape: we want to talk to our friends, yet these sites run them in a way that means they keep drifting away from actually letting us do so. We increasingly became data points for Big Tech to monetise and serve targeted ads. And it was designed to be addictive.
Netflix and Spotify gave us a way to watch shows and hear music, but as streaming multiplied, it brought a loss of collective culture.
There was a moment to be optimistic again when ChatGPT arrived. The promise of personal assistants able to perform boring tasks and boost efficiency. But before long, it’s given way to anxious narratives about job losses, supercharged disinformation through deepfakes, and an accelerated climate crisis through heavy energy and water demands.
There’s been great progress in other areas (reusable rockets, renewables, and chemistry, medical breakthroughs) but in consumer tech why does it seem to be making our lives worse?
The internet age was supposed to usher in technology as a force for good.
But now there’s a shift: our default attitudes have changed, with resistance mounting against the technology that dominates our everyday lives.
From The Robin Report
Good article on ‘the slow burn in retail’ making the point that “The slow burn of customer neglect can be fatal. If we’re not customer-centric, we’re self-centric. We assume everyone thinks the way we do, wants the latest trendy style, trades a meaningful experience for a speedy transaction, or wants to come to our dream team- created in-store events. In short, if we underestimate the customer, it’s a slow burn to their disengagement and disappointment. Great retail is a service, not a transaction”.
From Jen Costello on LinkedIn:
In this year’s glossary, you’ll find 39 cultural shifts that can help brands find real relevance. They’re not temporary fads or flash-in-the-pan trends, but fresh manifestations of the biggest and most relevant cultural conversations taking place around the globe – from gender roles to generative AI, sustainability to survivalism. Consider it a roadmap out of the culture rot. A guide to not just copying culture, but adding to it. Read here: Fast Company
I’m Michael Cooper, and I think about entanglement between our identities, personal agents, AI, Intelligent Interfaces, culture: their impact on our behaviours, choice- making, and empowerment to help us see our and our customers’ lives in new ways.
If you like this and want more:
My private work, where I share original thinking with brands, strategists, and futurists, with provocative conversations.
My LinkedIn, where I post my ideas.
I speak at conferences, company events, or to leadership teams - anywhere I’m allowed to ‘think aloud’ about my ideas to audiences curious about the entanglement of our humanity, culture, strategy, tech, brands and the future.
I’m currently taking bookings for late November and December.
Thanks for reading,
Michael
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.