94: New technology alone is not the answer

Jock Busuttil

1 year ago

Turkish chef Nusret Gökçe aka Salt Bae sprinking AI on stuff

I’m writing about 100 things I’ve learned the hard way about product management. You can catch up on the previous entries if you like.

Generative AI, large language models, agentic AI (and whatever comes next) each has its own strengths and weaknesses. What they’re not is a panacea or a magic bullet. New technology is not going to suddenly make all the challenges facing an organisation disappear overnight. Why? Because more often than not, those challenges are social not technological. Technology alone rarely solves ‘people problems’.

A repeating pattern

Adopting AI-related technologies has been the right move for many organisations¹. The tech complements and enhances what they were largely doing already, and they understand and work within the inherent limitations of the tech. Other organisations are sprinkling AI over everything partly because everyone else seems to be doing it, partly because there’s a vague hope that doing so will make things better in some unspecified way.

This pattern repeats with every major shift in technology or associated practices. We’ve seen it many times before with personal computing, the world wide web, social media, smartphones, cloud computing. We’re seeing it again with AI-related technologies. It stems from the mistaken belief that the new technology itself is the important bit rather than the new ways of working, interacting and playing that the technology now makes possible. It often also highlights the lack of a strategy beyond joining everyone else on the hype bandwagon.

Now a Director of Strategy and Transformation at Herd Consulting, former head of product at Ministry of Justice Scott Colfer observes:

Everyone’s budget is being cut to the bone, commercial and public sector — the only thing boards seem willing to invest in is ‘AI’ — so senior leaders and middle managers are going with it and making their deal with the devil, as they still get some funding out of it (even if there are strings attached).
Scott Colfer, Director of Strategy and Transformation at Herd Consulting

Digital transformation in UK government

By the time I joined the UK’s Ministry of Justice in 2014 for my first stint in government, momentum had been building behind digital transformation. The seeds had been planted in the year before with 25 exemplar services. Building on the success of GOV.UK — a single destination for the public that replaced the hundreds of fragmented sites before it — Sir Francis Maude declared at SPRINT 13: “we are determined to build fast, clear, simple digital services that are so good that people who are online will choose to use them.”

Ever an awkward term, ‘digital transformation’ was a double-edged sword. It was as useful a shorthand for the people who understood what it involved and why it was important, just as it was for the people who didn’t and repeated it often enough to try to hide that fact.

Alex Pentland observed in his book Social Physics that we tend to ape the behaviour of our peers when we see those peers being rewarded for that behaviour in some way. Near the beginning with digital, there was a core of experts brought in from outside government, mainly (but not exclusively) at Government Digital Service (GDS). They had already demonstrated the benefits that could be achieved with GOV.UK, and with senior-level backing from Francis Maude they had a strong mandate.

The departments and other government bodies closest to GDS set out to replicate their success and accolades. To begin with, these adjacent organisations often did not have the relevant skills or understanding, however their proximity to the core of experts allowed the knowledge to transfer pretty well, and they too achieved reasonable success.

However, not all government bodies were as proximate or ideologically aligned with GDS. The more distant layers of the onion copied the language and rituals of digital transformation and hoped for similar success, but lacked the underlying understanding. The further away from the core of experts they were, the more that ‘magical thinking’ took hold. These organisations sprinkled ‘agile’ and ‘digital’ liberally but superficially over everything, often with the help of expensive consultancies which were equally clueless but offered a compelling shortcut.

Predictably, without the relevant underlying skills and understanding, merely going through the motions didn’t deliver the anticipated results. Naturally it was the concept of digital transformation that was to blame rather than their superficial understanding and imitation of it. But by now, the damage was done and GDS was seen as responsible for promoting the idea in the first place.

Tech doesn’t solve people problems

Simon Wardley would often use weighing scales as an example of the dysfunction that would arise from departmental silos trying to solve problems locally rather than holistically:

A particular department has to produce a count for another group, it gets this count through weighing machines. What happens is it gets lots of paper forms and it’s too time consuming to count up the paper forms. What it does, it weighs them to calculate the number of paper forms, puts them into a system and reports that number at the end of the month to another department. It wanted to spend a lot of money converting its manual weighing scales to digital, it’s part of a digital transformation program. Exciting!

…

“Where are you getting your paper forms from?” At the distribution site, “Well, we print them out.” “You’re printing these paper forms?” “Yes.” “You’re sending them to this group?” “Yes.” “Do you know what they’re doing with them?” “No.” “Okay. Where are you printing them out from?” “Our CRM system, which our users fill in online.”

“The users fill in a CRM system online and you print out the paper to send to this other group to count the number of paper forms printed out?” “Yes.”

All of this could be replaced by SELECT COUNT (*) FROM [table]. Now these people aren’t stupid, what’s happening is they’re optimizing their local area.

I’m seeing the whole picture, does spending millions on a digital transformation program with new shiny additional scales make sense? No. Okay, so once you learn to communicate, then you start to discover patterns.
‘Crossing the River by Feeling the Stones’, Simon Wardley at QCon London 2019 (4 July 2015, 00:13:02 onwards)

The mistake many government bodies made back then was to think that the act of making more stuff available online for people to self-serve would paper over all the flaws with the services they were providing. Replicating the process of a poorly-designed, government-centric service in digital form simply resulted in the original (and newer) problems, just via a different communication channel.

In contrast, the departments that ‘got it’ were looking at what people were actually trying to do, assessing whether the service the government currently provided was allowing that to happen easily (or at all), then figuring out a more effective way to achieve that outcome for both the user and the service provider. Sure, making an online channel available may have been a part of that solution, but it was the means, not the end. In other words, when solving a ‘people problem’ (“I have to interact with government to do this other thing that actually matters to me”) the tech wasn’t the important bit, but in many cases it happened to form part of the solution.

In retrospect part of the problem was the reiterated belief early on that digital services (solely) would deliver the promised speed of delivery and cost reductions. New technology and its associated working practices were billed as the saviour of UK government. However, the whole digital thing was arguably a Trojan horse for the actual revolution, which was to build coherent services that met the needs of users first, government second and cut across the internal departmental silos. That this happened to result in the creation of more digital things was largely due to the Venn diagram overlap of ‘more convenient for the user’ and ‘doing stuff online’.

But by then it was too late to pivot the message. Digital was everywhere. It was in the name of every new team. I wonder whether things would have panned out the same way had GDS simply been called ‘Government Services’ instead, but with the same laser focus on user needs. Or perhaps it was the promise of the new shiny things that came with ‘digital’ that helped the idea to gain traction in the first place.

A rising tide

The time to debate whether we should adopt AI-related technologies is already long gone. The fundamental technology shift has already happened: for better or worse AI is in our smartphones, our productivity tools, the media we consume and the public discourse. This particular genie is not going back in the lamp. It also means that the bar of expectations has been raised for a good proportion of users (but not everyone yet, and not to the same levels for all). You’ll notice this effect when it becomes as absurd for organisations to highlight that something is powered by AI as it is for them to say they have a website.

Similar to other fundamental technology shifts such as the world wide web, the rising tide lifts all boats: consumers and organisations all have access to these new technologies in one form or another, so simply having AI as part of your offering is not in itself a differentiator. It’s more useful to ask when, why and how would AI-related technologies genuinely address valuable user needs in a way that also benefits the organisation. The answer can and should be that AI is not always the right solution, despite the associated hype. What matters again boils down to the intersection of ‘more convenient for the user’ and ‘doing stuff using AI’.

The challenge for commercial and public sector organisations alike remains two-fold:

1) It is perennially hard to find and solve valuable user problems well in a way that also benefits the organisation

2) User and organisational needs are continually evolving and being shifted by external factors beyond their direct control

It continues to be important to start with the problem, understand as best as possible why it is occurring for the people with it, then to look at the range of possible and practical ways to solve that problem. Those problems, needs and opportunities have always been shifting — we just happen to have undergone a period of more rapid change than usual. New, shiny AI technologies (and whatever comes next) may open up previously unavailable options to solve a given problem. However, we shouldn’t get hung up on ‘AI’ as a catch-all solution any more than UK government got hung up about the ‘digital’ bit of service transformation.

Final thoughts

Everything around us will to continue to change, but some things will remain constant as long as we focus on asking the right questions and seeing any new technology for what it is: a new set of possibilities, but not necessarily the solution to every problem.

Notes #

I use ‘organisation’ throughout rather than ‘company’ because this applies equally to the private and public sectors. ↩