Every January, I make it a point to read the State of the World conversation between Jon Lebkowsky and Bruce Sterling on The Well.
While some people turn to astrology for insight, others put their trust in Sterling. I’m firmly in the latter camp.
Another part of my tradition is revisiting the previous year’s predictions, prognostications, and prognoses to see how well they held up.
Many do the same with horoscopes.
On January 1, 2024, Sterling kicked off his annual ritual by asking Microsoft's AI-powered Bing for some forecasts, after lamenting that 2024 was going to a remake of 2023, as in more of the same.
A minute later, Sterling humorously dismissed Bing’s assessment – Kamala Harris’s nomination for POTUS, that is – as a mere hallucination.
But I think Sterling got this one wrong.
Bing wasn’t hallucinating.
It was doing something much deeper: it was engaging in premediation.
Richard Grusin introduced the concept of premediation in 2010. He defined it as the process through which media anticipate or shape future experiences, events, or developments before they occur.
Grusin argues that media – and large language models like Bing certainly qualify as media – don’t just reflect on events after the fact. Instead, they actively shape how we perceive and prepare for those events in advance: premediation is a kind of preparation. Premediation is also distinct from remediation, which focuses on how new media repurpose, incorporate and/or adapt older forms.
Premediation is especially relevant in the digital age, where media can create anticipatory narratives and imagery that make certain futures feel inevitable or likely. Predictive algorithms, simulations, and other digital media often shape public expectations and, in some cases, influence events before they fully unfold (in his book, the American scholar uses video games, or rather, game design, to explain premediation).
Grusin’s theory highlights that our engagement with media is not purely retrospective – analyzing what has happened – but also prospective, shaping our vision of what could happen.
In the case of Bing’s prediction sui generis*, the AI wasn’t just making a random guess. It was performing premediation, nudging its users toward a particular interpretation of the future. This wasn’t a hallucination; it was a calculated disruption that, seven months later (July 21 2024, to be precise), turned out correct.
In some ways, premediation isn’t so different from astrology.
Both involve crafting narratives about the future, providing frameworks to interpret what hasn’t yet come to pass. And sure, they are both fictions, figments of someone’s imagination.
But they are not just fictions. Fictions are powerful. Fictions shape reality.
Sterling, the oracle of our digital age, is situated at the crossroads of these traditions – what is AI-infused futurology if not a peculiar kind of astrology? –, bridging the speculative and the analytical. Whether through his wry dismissal of Bing’s predictions or his role in interpreting the currents of technological and geopolitical change, his annual performance is a reminder that forecasting (hallucinating?) the future is as much an art – or rather, sorcery – as a science.
The question remains: are we passive participants in AI’s premediated futures, or can we learn to shape the narratives that will define us?
I look forward to reading the 2025 State of the World.
* Here’s the framing: “Hello, this is Bing. I can give you some information about the world in 2024 based on the latest news and web sources.”