15 March 2024. Tech | Seven questions
Tech is turning into an ordinary business // Asking questions about the future. [#552]
Welcome to Just Two Things, which I try to publish three days a week. Some links may also appear on my blog from time to time. Links to the main articles are in cross-heads as well as the story. A reminder that if you don’t see Just Two Things in your inbox, it might have been routed to your spam filter. Comments are open.
1: Tech is turning into an ordinary business
The French writer and entrepreneur Nicolas Colin has just written a post where he revisits some of the ideas that have informed his thinking—mostly about technology—over the past decade or more, and tries to assess how these have held up.
He has nine headings and one summary section that looks at where we are in 2024, but looking across then, it turns out that there are some connecting themes, so rather than do a list, I’m going to try to make some of the connections.
(Elsewhere in the piece, he wonders if the globalisation of tech has given way to fragmentation, whether China is now turning inward, and Europe’s difficulties in finding a distinctive technology path.)
The first idea, from a co-written book, L’Age de la multitude, (written during 2011-12), was that we were entering what he called “the age of the multitude”:
Guided by (co-author Henri Verdier)'s vision, we centered the book on the premise that technology empowers individuals, enabling them to surpass the influence of organizations—a concept we believed explained the current paradigm.
Obviously he’s not the only person who had this model in their head, and Tom Peters wrote a book called A Brand Called You a bit earlier which argued, similarly, that the internet would have this effect. Colin doesn’t think that this one has turned out so well.
My doubts are inspired by two main observations. Firstly, we assumed this empowerment would compel companies to prioritize exceptional user experience. However, has this idea turned to ‘shit’? Cory Doctorow captured the shift in the concept of the “enshittification” of platforms.
(I wrote about this almost exactly a year ago on Just Two Things here.)
(Detail from The Downfall of Capitalism, by Jack Hastings.1 Photo by Phil Beard/flickr. CC BY-NC-ND 2.0)
His second observation—I think this one is a bit sketchier—is that large language models could be used to emulate or mimic the collective power of individuals.
This connects to another of his key ideas, in the work of Fernand Braudel, who viewed capitalism as the process of integrating capital into the production system:
This concept refined my understanding of corporations as entities balanced precariously among three stakeholders: customers seeking better quality at lower prices, employees demanding higher wages and improved working conditions, and shareholders aiming for greater returns on their investments.
His perspective is that we are at a moment where employees have been better able to protect their interests (he may overstate this point), while shareholder influence remains as strong as ever. Customers get the wrong end of what’s left:
customers/users often appear to be the disadvantaged party. Braudel’s vision of capitalism and the framework of the tripartite corporate contract definitely stand the test of time, and highlight a significant evolution in the balance of power within corporations.
And this may be a different way to understand the dynamics that sit behind Cory Doctorow’s enshittification thesis.
One of his ‘key ideas’ is about the work of Carlota Perez, and her ‘five technology surges’, which he used to inform the thinking of L’Age de la multitude’, at a time when we were still clearly in the middle of the ICT surge.
(Source: Carlota Perez)
He used this as the basis for the idea that this fifth wave might lead to the development of a new type of business, different from the production led models of the 20th century. They would be different
not only in their products but also in their organizational structure, culture, and stakeholder relationships, including with customers, employees, and shareholders.
Earlier on, some of the signs were there, he thinks:
Initially, major tech companies seemed to embody this new paradigm by prioritizing exceptional customer service, treating their highly skilled employees incredibly well, and placing shareholder interests lower on their agenda.
He argued back then that the “true tech company” provides “an outstanding customer experience, consistently monitors user engagement, and strives for increasing returns to scale.”
The latest evidence from the sector doesn’t really support this view:
Amazon, once the pinnacle of seamless service, has become bloated and is now crowding its platform with advertisements , compromising the user experience. Meta has just started issuing dividends, and Google has restricted its historically open work culture, indicating a movement towards traditional corporate models.
He doesn’t mention this, but this is exactly what Perez’ model would anticipate. Effectively in the fourth quarter of the S-curve in each surge, markets become saturated, and returns start to fall, and companies that led the way during the period of rapid market development during the third quarter of the surge start looking and behaving like conventional mature businesses.
Reflecting on Perez, Colin wonders elsewhere whether the ICT wave is running out of steam, but decides against:
However, the current debate (including my own views on the door closing on startups ) questions whether we're now moving beyond the Age of Computing and Networks. Though advancements in energy, space, healthcare, and agriculture hint at a new, different technological revolution, the AI boom—powered by advanced computing (credit to Nvidia) and networks (think large language models functioning as virtual ant colonies)—suggests we're still navigating this era, with still much to leverage.
I’m not sure about this. The more I look at AI, the more it looks like the kind of technology adaptation you get at the end of a wave, not the beginning, signalling the long run problems left behind by a mature technology. (I see that the economist Noah Smith has now joined the ranks of people concerned by the energy consumption of AI, along with that of crypto, on his blog.)
The way the Perez model works is not that technology runs out of steam, but that it’s impossible to get decent returns on the technology, which means that investment capital looks for new homes elsewhere (such as energy, healthcare, agriculture. Though probably not space).
And if it really takes $7 trillion investment to make AI work, as Sam Altman of OpenAI seems to be saying at the moment, then there’s no real prospect of returns there, ever. In fact, generally, no-one’s suggested a credible business model for AI that fuels the kind of investment that leads to a new technology surge.
2: Asking questions about the future
A sometimes colleague asked me a question about the Seven Questions interview framework, and when I had a look at it I realised that although the framework, or versions of, are widely used by futurists, it’s not much written up. The framework itself is published by the UK’s GO-Science in their Futures Toolkit (p.29), and looking at some of my recent proposals, I barely send out a proposal that doesn’t mention it—we use an adapted version—as part of the overall project method.
The reason that it is widely used is that it was used by Shell in their scenarios process, and it is written up by the late Kees van der Heijden in his book Scenarios: The Art of Strategic Conversation. In turn the book is probably the best single explanation of the Shell scenarios method.2 The book seems have fallen out of favour now, but there is a lot in there that is valuable.
Van der Heijden devotes a few pages in the book to the importance of good interviewing, in which he emphasises the importance of open-ended questions and careful listening.
So this is the first important thing about Seven Questions: it is a guide to a semi-structured interview in which it is OK to go off script to understand a point the interviewee is making.
It’s worth explaining the questions, of course. They are, broadly, designed to let the interviewer first ask about how the interviewee sees patterns of change, and uncertainties, elicit some normative views on these, and then move to discussing the sorts of decisions that need to be addressed.
Van der Heijden doesn’t include the actual questions Shell used in the book, but there are other accounts of these. (They may not be completely reliable, of course—photocopies of photocopies, and all that.) This is the version cited by Gill Ringland in her book Scenarios, sourced to Shell via her colleague Gareth Price.
The Vital Issues(the Oracle) 1. Would you identify what you see as the critical issues for the future? (When tbe conversation slows, continue comment) Suppose I had full fore-knowledge of the outcome as a genuine clairvoyant, what else would you wish to know?
A favourable the outcome 2. If things went well, being optimistic but realistic, talk about what you would see as a desirable outcome.
An unfavourable outcome 3. As the converse, if things went wrong, what factors would yoax worry about?
Where culture will need to change 4. Looking at internal systems, how might these need to be changed to help bring about the desired outcome?
Lessons fron past successes and failures 5. Looking back, what would you identify as the significant events which have produced the current situation?
Decisions which have to be faced 6. Looking forward, what would you see as the priority actions which should be carried out soon?
If you were responsible 7. If all constraints were removed and you could direct what is done, what more would you wish to include? (The 'Epitaph' question).
Shell had adapted these from a book by Amara and Lipinsky, of the Institute for the Futures, published in 1983.
It does seem that Shell used these a lot. At a meeting to mark the 25th anniversary of the book’s publication, Peter Schwartz says interviews took ”30 to 40 per cent of his time”.
The same paper says that the questions are “unimprovable”, but that depends on what you’re trying to do. This might also be an example of the Shell scenarios “halo” at work.
Gill Ringland, in her book Scenarios, says that she added a question for a project that she did with British computer company ICL, about the specific future of the company.
My own experience with using them, having started out (as you do) following the published script exactly, was that there were a couple of problems with them.
The first was that they are too focussed on the business and its issues. In my practice, that comes later in the process: you need to get a shared view of the landscape before you get to the implications for the organisation.
The second was that I found there was a problem with the order: people found it cognitively difficult to jump in cold with the so-called ‘Oracle’ question, partly because it takes a bit of an effort just to get your head around it.
So in my practice I’ve re-worked these. Some of the Shell questions work well, others get you too close to the organisation and its specific perspectives too early in the process. This isn’t the full description I use, but this is the flow:
The past: Looking at (this overall system)3 what the are main factors over the last 20-30 years that have shaped where we are today?
The future: Looking at the next (20 years), what new factors do you see emerging?
Good outcomes: What will organisations that succeed in this future environment have done well?
Dismal outcomes: What will organisations that don’t do well have done, or failed to do?
Decisions: If an organisation is going to succeed, what are the decisions it needs to be making now?
Uncertainties: (our version of The Oracle question, which we think of as the ‘Dr Who’ question): if someone from 20xx (insert year of choice) who knew everything about how things turned out in (system being discussed) what are the two questions you’d want to ask them about how they did turn out?
President for a Day: (our version of the Epitaph): If you were President for a Day, and could do a couple of things that would make a positive difference to the future (of the system) and had the resources to make them happen and make them stick—what would they be?
In his book, van der Heijden notes that normally in futures work, asking people to think about good/bad outcomes is not a good idea. But in the context of the interview it works well:
In the elicitation interview the discussion of good and bad worlds tends to be powerful in triggering ideas of what could be important factors to look at, leading to the discovery of underlying driving forces. These tend to be productive questions.
That’s been my experience too, and in practice these reorganised questions give people to approach things in different ways during the interview.
But pulling it away, mostly, from the specific questions about the organisation also moves it—in Donald Schon’s phrase—from “problem solving” to “problem setting”. Since you do these interviews early on in the process, you’re still trying to understand the system and the landscape. And people tell you plenty about what’s happening in the organisation between the lines.
NOTES FROM READERS: MINERS
In response to my piece about miners and the miners’ strike, and some music associated with this, David Bent mentioned to me the rather more recent—and more elegiac—record by Public Service Broadcasting, Every Valley.
He helpfully included a Wikipedia summary as well:
The album's story depicts the history of the mining industry in Wales, more specifically chronicling the rise and decline of the country's coal industry. The band's lead songwriter, J Willgoose, Esq., described the album's premise as an allegory for today's "abandoned and neglected communities across the western world", which have led to a "malignant, cynical and calculating brand of politics."
j2t#552
If you are enjoying Just Two Things, please do send it on to a friend or colleague.
Beard notes that “Hastings worked alongside Diego Rivera in 1930s Detroit and New York as a mural painting assistant. He absorbed none of Rivera's compositional mastery - his enthusiasm exceeded his talent.”
Which I think everybody knows by now is not the same as the 2x2 double uncertainty matrix method that was popularised so aggressively by Global Business Network.
Insert subject of the project here.