Fish Food: Episode 621
Using AI to challenge your thinking, Ben Evans on AI eating the world, the X-odus, AI for strategists, Microsoft's reinvention and scamming the scammers.
This week’s provocation: Perspective blindness, and using AI to challenge your thinking
‘20 years after my own graduation I have come gradually to understand that the liberal arts cliche about teaching you how to think is actually shorthand for a much deeper more serious idea. Learning how to think really means learning how to exercise some control over how and what you think. It means being conscious and aware enough to choose what you pay attention to and to choose how you construct meaning from experience’
David Foster Wallace, from his fabulous ‘This is Water’ commencement speech
Back in June I wrote about why cognitive diversity on teams is such a superpower, particularly when it’s combined with psychological safety. A paucity of diverse perspectives or the environment in which people can easily express those views can easily lead to team-wide perspective blindness, the tendency we all have to be unaware of our blind spots.
This week I read a thought-provoking post from the neuroscientist Chantel Prat, talking about how our experiences can hard wire our brains into habitual thought patterns, and how we barely notice how these filter our perception of reality. We need, she says, to be more aware of how our inputs shape the way that we think, and more deliberate in thinking beyond our perceived limits.
It made me think about perspective blindness at the individual level. In his book ‘Rebel Ideas’, Matthew Syed says that ‘we perceive and interpret the world through frames of reference but we do not see the frames of reference themselves’. This, he notes, means that we tend to underestimate the extent to which we can learn from people with different points of view.
I keep wondering about the role that AI can play in undermining this kind of perspective blindness. Right now, most GenAI use seems to be pretty executional - focused on time saving or efficiency. Largely this is helping us get to places that we probably would have got to anyway, but just making it considerably easier or faster. But what if we can use AI to really challenge our thinking, or to change our frame of reference entirely, or understand a totally different point of view?
I’ve been playing around with prompts to see if I can use AI in a more substantive way to open up new ideas and thinking. There’s still much progress to be made but I thought it may be useful to list some of the ways that I’ve found most useful to do this.
Norm switching: here you can prompt the GenAI tool to explore the norms of a particular situation or context and then get it to apply those norms to a completely unrelated context. One example that I’ve used is exploring the norms, practices, and innovations in one sector (Formula One racing) and then prompting the AI to apply those to another sector (Banking).
Perspective role reversal: You can flip the script by getting the AI to imagine how someone with a drastically different perspective would approach a particular problem or situation.
Metaphor exploration: I use metaphors a lot in presentations, so you can use a ‘If X was Y’ approach to reframe your thinking. For example: ‘If the process of launching a new product were like planning a rock concert, what would be the equivalent of the 'headline act,' 'venue selection,' and 'ticket sales'?’
Constraints-Driven Creativity: Introducing specific constraints or extreme scenarios to stimulate different thinking. Constraints can force different thinking (my favourite example of this is Audi at Le Mans). So as an example you might get the AI to imagine that it had to redesign something using only 5% of the current budget, and ask what would that look like.
Counterfactual and ‘What-If’ Thinking: AI can be used to explore hypothetical scenarios. An example: ‘What if our industry had to operate entirely without digital tools? How might companies adapt and thrive?’
Reverse Engineering Success: the tools can be used to analyse successful outcomes in unrelated fields or industries (randomly I quite like doing sports examples), and then reverse-engineer how they were achieved. These lessons can then be applied to a different scenario, business or sector.
What would nature do? Nature often provides elegant solutions to complex problems. So you can ask the AI to consider how lessons and principles from natural systems (anything from weather, to animal herds, to fungi, to Starling murmurations) can be appled to another context. For example: ‘How might a forest ecosystem's principles for resource sharing and interdependence inspire a new way to structure partnerships in X sector?’
Combining Unrelated Concepts: creativity is combinatorial, so asking an AI to deliberately collide two unconnected ideas can sometimes uncover surprising connections.
There’s aways a bit of back of forth involved after the initial answer but the opportunity here is to take one or more aspects of what has been suggested and to dive deeper to get to really interesting ideas. This has definitely taken me to places that I probably wouldn’t have got to by myself, and I think has been really valuable in getting beyond the perspective blindness that we’re all susceptible to.
If you do one thing this week…
Every year analyst Benedict Evans does a touchstone presentation of the state of tech and this year’s one is as rich in insights as ever. It majors on the state and future of GenAI (of course). Some really interesting charts showing how lots of people have tried it but a much smaller proportion are using it on a daily/weekly basis, and how software, customer service and marketing are the current main enterprise use cases (which supports McKinsey data that I’ve seen elsewhere).
Then the presentation dives into to the key questions that will shape the future of GenAI, why this is another major platform shift (the next S-curve), scaling implications related to data and computing power (no particular mention of the environmental considerations though), and how all the science and engineering questions are still moving underneath our feet. A telling quote about what this means for new businesses: ‘The consumer internet model of “launch free, go viral, work out revenue later” doesn’t work with today’s LLM cost model’. And some interesting questions around why aren’t more people using it (I liked this Steve Jobs quote: ‘You’ve got to start with the customer experience and work back to the technology’). Plus some good charts on ecommerce an retail media as a bonus. Amazingly useful.
Links of the week
Plenty of coverage this week of the mass X-odus, as the FT put it, as professionals leave the platform en masse. The main beneficiary seems to be Blue Sky, which does (for now) feel like Twitter used to pre-2016. I’m there, so do connect with me there.
Google have released a new prompting course which is designed to build on their AI Essentials course (the most popular GenAI course globally on Coursera apparently) and teach how to go beyond just ‘good’ prompting to somewhere near great. More on what it includes here. Useful.
The APG (Account Planning Group) and a bunch of senior planners have been doing some great work pulling together some research on ‘AI for strategists’, focusing on four main areas - what it means for the role of the strategist, tools and practice, quality control (potential issues relating to critical thinking), and effectiveness. I was interviewed for the research and sat in on the webinar where they went through some highlights and also launched a useful toolkit which lists practical tools that are available for planners and which will be expanded over time.
Gen AI is sweeping through many industries as we know, and design is no exception. Fish Food subscriber Rob Boyett got in touch to say that he, along with some other design industry veterans, have come together to conduct a survey across designers of all flavours to properly explore how GenAI is impacting the day-to-day job of design. It’s a completely independent survey so worth supporting if you are any kind of designer (and they’ll plant a tree for every respondent they get)
A great deep dive from Stephen Levy on the story of Satya Nadella’s tenure as the CEO of Microsoft. Related - this Bruce Daisley EatSleepWorkRepeat episode is one of my all time favourite case studies on how he drove cultural change in the company
There often seems to be little differentiation between how people talk about Chatbots and AI agents but they are not the same thing. This is a handy breakdown of the difference between RPA, chatbots, co-pilots and AI agents
Surprising fact of the week: ‘More music is being released today (in a single day) than was released in the calendar year of 1989’. Thanks to the democratisation of tools and music self-publishing (HT Faris)
Quote of the week
‘The real problem of humanity is we have Palaeolithic emotions, medieval institutions, and god-like technologies’ Edward O. Wilson (HT Paul Frampton-Calero)
And finally…
I absolutely loved this idea by O2 and VCCP. Daisy is an AI-generated pensioner who has been designed to ‘scam the scammers’. She keeps them on the phone for as long as possible, frustrating them and wasting their time so that they can’t use that time to scam real pensioners.
Weeknotes
This week I was out in Dubai working with a banking client and I got to run a workshop in the most Bond-villain’s-lair-type venue I’ve ever been in on a man-made island off the coast. It was quite something. Over the next couple of weeks I’m back on the road again out in the Middle East (but not in Bond-villain’s-lair-type venues unfortunately) and then Belgium.
Thanks for subscribing to and reading Only Dead Fish. It means a lot. If you liked this episode please do like, share and pass it on. If you’d like more from me my blog is over here and my personal site is here, and do get in touch if you’d like me to give a talk to your team or talk about working together.
My favourite quote captures what I try to do every day, and it’s from renowned Creative Director Paul Arden: ‘Do not covet your ideas. Give away all you know, and more will come back to you’.
Really useful prompts, thanks Neil