How can we use the Gartner hype cycle to understand how new technology causes harm
The sudden explosion of free and accessible generative AI tools over the last couple of months has finally prompted me to write a post that I’ve been thinking about for years.
The world is transfixed by dizzying possibilities of ChatGPT and it has me thinking about that oversimplified marvel, the Gartner hype cycle.
I am guessing you're already familiar with the Gartner hype cycle, but in sum, it illustrates the way that a new exciting technology is gobbled up by the world.
Broadly, there are three major phases:
1. a big wave of excitement
2. a cratering of interest as we come to grips with the reality of what's actually possible and
3. some stable state when it becomes ‘normal’ technology that adds some value.
Pretty straightforward. But we know that hype has other effects. And that hyped technologies leave a trail of harm behind them. So how can we use the hype cycle to understand these dynamics?
I have been thinking about two other cycles that sit alongside this one: a value creation cycle, and a harm cycle.
💸 Looking at the value creation cycle first: hyped technologies create value.
But for who? And at what point in their integration into our everyday?
New technology creates a lot of value for a tiny group at the point of greatest hype. This value is created no matter what ultimate value the technology creates for society. Consider early investors in Uber. They made a killing, and they picked a loser.
Even if web3, the metaverse, or generative AI never actually become well-established, valuable parts of society, they will still create tremendous value for some people and sectors.
VC firms pump in cash and then pass their investments like a hot potato to the next wave of investors at a major mark-up. Or use public excitement to drive resources to highly risky ventures. Experts make careers out of helping people understand these new marvels. Bankers and lawyers make money with every acquisition, IPO, and funding round.
And over time, a technology will likely find a narrow use case where it adds real value.
The key thing about the value creation cycle is that hype is translated into money, power, and attention for people already in a privileged position vis-a-vis new technologies.
This incentivises the very people who should know better – because they’ve seen it before – to feed into the hype. No matter the harms that accumulate, or the expectations that go unmet, they will have their payday.
What's the big deal? Some people make a quick buck, and maybe a technology doesn't pan out to be as transformative as we were promised.
The big deal is the second cycle.
The harm cycle.
😞The harm cycle starts at the same time as the hype cycle: a new technology is released prematurely and negligently kicks off a new wave of harm.
Hang on. Harm from a chatbot? How?
A few questions for you:
What would happen if a machine read the internet and then used what it learned to provide confident answers to complex questions?
What if it pretended to be sure of its answers?
What would happen if anyone anywhere could ask anything and think the answer was true, because computers aren't wrong?
Questions about medical advice?
Or historical analysis of race or gender or income inequality or politics or power?
Doesn't take long to be concerned.
And that concern only grows when we think about it being available everywhere, all at once, and integrated into existing systems that we see everyday.
What happens when individuals start to experience negative effects of this thing?
Nothing. At first.
As the harm accumulates, we will enter what the Hype Cycle labels 'the slope of enlightenment’. This is where the backlash begins. The harm will no longer be deniable by those peddling the possibilities. And the conversation will shift.
From one of limitless possibility to cautious optimism. And over time any media coverage that mentions the magic, will be sure to include a head nod to the harms.
What happens as ‘we’ ‘learn our lessons’ about the new tech?Unfortunately, harm doesn’t plateau because the technology doesn’t go away.
It’s too late for that. It’s embedded in ways we can’t undo. So the harm continues, and over time, becoming quotidian. A price of entry in the modern world.
The embedding of immature technologies into our social systems and infrastructures leads to harm that is much more challenging to reverse than to initiate.
It’s fun and thought-provoking to sit around and consider all of ChatGPT’s possibilities — but the harm is coming. In fact, it is likely already here. Once it becomes visible – or is forced into view – it will be too late.
The boosters will have extracted tremendous value from the hype cycle in the shape of cash and attention; and the peons will be stuck navigating a world transformed.
Not into a plateau of productivity, but into a world layered with yet more tech-enabled indignities.