How We Created A Virtual Workshop To Help A Group Of Brilliant Thinkers Explore 'Positive' Futures Around AI
from the it's-not-all-dystopia dept
Having spent two and a half decades writing about innovation, one of the things that’s most fascinating to me is how little most people can envision how innovation can have a positive effect on our lives. Perhaps it’s a lack of imagination — but, more likely, it’s just human nature. Human psychology is wired for loss aversion, and it’s much easier to understand all the ways in which technology and innovation can backfire to take away things we appreciate. History, however, tends to show that the positives of many innovations outweigh the negatives, but we’re generally terrible at thinking through what those benefits might be.
Part of the reason is just that it’s impossible to predict the future. There are just too many variables, and too much randomness. But, part of it might also be our general unwillingness to even try to imagine positive futures. But imagining positive futures is one tool for actually getting us to move in that direction. Even by suggesting what interesting innovations and societal changes might happen can inspire individuals, organizations, institutions, and movements to try to make what was first imagined into reality. And we sure could use a bit of positive thinking these days. This is the story of how we attempted to help create more positive visions of the future — specifically around artificial intelligence.
As some of you may recall, a few years back, we did a fun project, called Working Futures to use a (more fun) type of scenario planning to explore possible futures for work — and then turn those scenarios into entertaining science fiction. As many people know there are all sorts of concerns about what the future of work might look like. We’re living in disruptive times when it comes to innovation, and in the last few decades, it’s created a massive shift in the nature of employment, and there are many indications that this trend is accelerating. Historically, similar shifts in work due to technology have also been disruptive and frightening for many — but all managed to be worked out in the end, despite fears of automation “destroying” jobs.
However, simply saying that “it will work itself out” is incredibly unsatisfying and, even worse, provides little to no guidance for a variety of different stakeholders — from actual workers to policy makers trying to put in place reasonable policies for a changing world. The Working Futures project was an attempt to deal with that challenge. We created a special scenario planning deck of cards, and ran a one-day session which helped us build a bunch of future scenarios. We gave those to science fiction writers, and eventually released an anthology of 14 speculative future stories about the future of work (which is rated quite highly in Amazon reviews and on Goodreads as well).
Late last year, some people associated with the World Economic Forum and Berkeley’s Center for Human-compatible AI (CHAI) reached out to us to say that they had been engaged in a similar — but slightly different — endeavor, and wondered if we might be able to lead a similar scenario planning process. The two organizations had already been working on a series of events to try to imagine specifically what a “positive future” for AI might look like. We all know the doom and gloom and dystopian scenarios. So this project was focused on something different: explicitly positive futures. The end goal was to take some of these positive AI future scenarios and use them as part of a film competition from the X-Prize Foundation (not unlike our Working Futures project, but with films instead of written fiction).
They asked if we could take an approach similar to what we had done with Working Futures and run a workshop for around 90 attendees — including some of the top economists, technologists, science fiction writers, and academics on this subject in the world — and… they said they’d already invited people for the event just two weeks later.
It turned what would normally have been a quiet time in December into a frantic mad dash as myself and Randy Lubin (our partner in our various gaming endeavors) had to put together a virtual event. We’ve obvious done scenario planning events — including ones about the future of work. And even Working Futures was designed to be generally positive. But what WEF and CHAI were asking for was even more extreme, and required a real rethinking of how to put together a scenario planning program. Traditional scenario planning doesn’t put any conditions on the potential scenario outputs — so creating scenarios where the goal is for them to be explicitly positive presented a few challenges.
Challenge 1: Directing scenario planning towards a desired style outcome is a pretty big departure from how you normally do scenario planning (starting with driving forces, and following those wherever they may lead). There are risks in doing this kind of scenario planning, because you don’t want to preset the end state, or you lose the value of the open brainstorming and surprise discoveries of scenario planning.
Challenge 2: Something we had discovered with Working Futures: explicitly “positive” futures sometimes feel… boring. They make for a tougher narrative, because good stories and good narrative usually involves conflict and tension and problems. That’s much easier in a dystopian scenario than a utopian one. And if the end result of these scenarios is to drive useful story-telling, we had to consider how to create scenarios that were both interesting and “positive.”
Challenge 3:: With Working Futures, we did the scenario planning in a large room in San Francisco, and we had a custom card deck that we had made and printed, that everyone could use as part of the scenario planning process, to experiment with a variety of different forces. In this case, we had to manage to do the workshop via Zoom. This was a separate challenge for us in that while we’ve all done Zoom meetings (so, so, so many of them…) throughout the pandemic, for good scenario planning, you want to make use of smaller groupings, and we hadn’t had as much experience with Zoom’s “breakout room” feature. This presented a double challenge in itself. We had to create a series of exercises that people could follow — meaning with enough scaffolding in the instructions that they could go off into groups and do the creative brainstorming, but without being able to easily see how they were all doing. And, we had to keep the whole thing interesting and exciting for a large group of very diverse people.
In the end — somehow — we succeeded in overcoming all three challenges, and created a really amazing workshop. The feedback we got was astounding. The key ways that we worked to overcome the challenges and to create something useful was a realization that we’d start with a few more “broad” ideas to get people thinking generally about these kinds of distant future worlds, and with each exercise we’d focus more and more narrowly, building on the work in earlier exercises to help craft a variety of scenarios. The very first exercise was more of a warmup, but one that was still important to get creative juices flowing: figuring out new abundances and scarcities in such a world.
To me, this was a key idea. When we think about big, disruptive changes brought on by technology, they often involve new “abundances.” Cars make the ability to travel long distances “abundant.” Computers make doing complex calculations abundant. The internet makes information abundant. Yet, the more interesting thing is how each new abundance… also creates new scarcities. For example, the abundance of information has created a scarcity in attention. As you think through new abundances, you can start to recognize possible scarcities, and it’s almost always those new scarcities where you find interesting ideas about business models and jobs. So we had participants explore a few of those (here’s one example that came out of the exercise):
In the second exercise, we asked the breakout groups to build a “qualitative dashboard” to guide humanity in this new, positive future. We assumed that there would be a focus on optimizing certain aspects of life, and we asked the teams to develop a “dashboard” of qualitative concepts that should be optimized, and from there what quantifiable measures might be used to see if society was reaching those milestones. Here’s one example:
Of course, recognizing that whenever you try to “optimize” a particular value, it almost inevitably leads to unforeseen consequences (usually from focusing and optimizing too narrowly on a small number of quantitative values, and missing the bigger picture), we then had the teams present their dashboard to a different team, and had those other teams provide an analysis of what might go wrong with such a dashboard. How might optimizing on one of these items go badly awry.
The third exercise was, in part, an attempt to deal with the problem of a utopian world being too boring. We had the teams focus on figuring out what was “the final hurdle” to reaching that “positive” future, and we used a tool that we’ve used a few times before: news headlines. We asked the breakout groups to effectively write a narrative in four headlines, starting with a negative headline demonstrating a major hurdle preventing society from reaching that positive future. Then the second headline would note some positive development that might, possibly, overcome the hurdle. The third headline was a setback: in the form of some kind of resistance effort that might block the hurdle from being cleared, followed finally by the last headline: a story that showed evidence that the hurdle truly had been overcome.
From there, we started to really focus in. The headlines created a sense of this “world” that each group was inhabiting, but we wanted to look more closely at what kind of world that was. The fourth exercise explored what were the new essential institutions, participatory organizations, and social movements in this new world. The idea here was to think about what would life actually look like in this world. Would there still be “jobs” or would your daily activity look radically different?
Again, mindful of both the potential “boringness” of utopia, as well as the fact that perfection is impossible, in the middle of this exercise we introduced something of a “shock” to the worlds that were being built — telling participants that a major earthquake had struck, with millions of people wounded, possibly dead, trapped or missing — and with major infrastructure disrupted. We asked the teams to go back and look at the institutions, organizations, and movements they’d just discussed to see how they reacted and how well they handled this shock (and if new such groups formed instead):
The final exercise of the workshop was designed and run by WEF’s Ruth Hickin, diving in even deeper, and asking participants to explore specific individuals within these scenarios. Each person was assigned a future persona, and had time to explore what that persona might think about this world — and then had each of the participants take on that role, and have a discussion with the others in their group, in character, trying to answer some difficult questions about their obligations to society, and whether or not they could find meaning in this future world.
While Randy and I planned the whole event in two frantic weeks, making it actually work required a bigger team of incredibly helpful people. Caroline Jeanmaire at CHAI and Conor Sanchez at WEF helped organize everything, and gave us great feedback and guidance throughout the project design. The two of them also helped keep the event running smoothly, following a very detailed run of show to make sure breakouts happened in a timely and clean manner, and that we could re-assemble everyone for the group discussions in between each of the breakouts. They also brought together a group of facilitators who helped guide each of the breakout discussions and keep everyone on track.
Another incredibly handy tool that made this all work was Google Slides. Between breakouts, we’d assemble everyone and discuss the next exercise, including an example slide. Then each breakout room had their own set of “template” slides that had the instructions (in case they hadn’t quite followed them when we explained them), an “example” slide to get inspiration from, and then the template slides for them to fill out. It turned out that this had multiple useful features. Perhaps most importantly, it helped us record the brainstorming in a way that would live on and which could be used later for the X-Prize competition. But it also allowed us to “peek in” on the various tables while the exercises were happening, without having to jump into their Zoom breakouts, disturbing everyone. As the breakouts were happening, I would flip between the different group slides, and see if any of the groups appeared to be struggling or confused, and then we could send someone into that breakout to assist.
In the end, as noted, the event turned out to be a striking success. We’ve received great feedback on it, and are exploring possibly running it again (perhaps in a modified version) in the future. Randy also wrote up a post describing some more of the nuts and bolts of the workshop if you want more details about how we pulled off the whole thing.
From my end, the biggest takeaway was that a well-crafted event could create truly brilliant and inspiring ideas for what the future might look like, and it was somewhat humbling to see how our framing and scaffolding was embraced by such an eclectic and diverse group to generate such fascinating futuristic worlds and scenarios. Hopefully, some of these futures will inspire not just films or stories about this future, but will also inspire people to work to make something like those futures a reality.