A rising demand for Outcomes & Flow

I’m sharing here some abbreviated notes from a talk I delivered for Manchester Law Society last week.? It was nice to go back home (special thanks to Fran Eccles-Bech and Rob Elvin for the invite) and reconnect with (and occasionally provoke) a community that was such a core part of my own development as a student, trainee and young lawyer.??

Manchester is renowned as a city that spearheaded the “industrial revolution” and I was keen to call on the idea that we are on the precipice of both technological and societal change that will be at least as significant.

Note 1: What united us in the room?

Whether we are practising lawyers, barristers, legal ops professionals or academics - we found common ground that:?

A. Our “training data” was essentially comprised of:?

• Our education
• The matters we (or those we have learnt from) have experienced handling?
• Our firm or team’s knowledge and intelligence?
• Legislation and case law (which we use to navigate to our truth)
• The workflows and processes we use to problem solve

B. That we would typically “apply our training data when prompted” by our clients, counterparts or organs of the state in order to help a person achieve an outcome - whether that’s an acquittal; custody; asylum; big business contract or the establishment or funding of a corporate entity.

C. How we achieve these outcomes is a priority for our thinking (it often determines the basic economics of our practice), but this is often not a priority of our clients.

Note 2: Are we unique (as legal professionals)???

We explored the breadth of our practice areas (in the case of law firms) and our work at the intersection of a variety of business teams or societal sectors, each providing us privileged insights across a range of social settings.??

More profoundly, though, there was consensus that we are a breed that is trained to be obsessed by recording, curating and maintaining knowledge and developing process and policy to guide and inform the behaviour of others.??

We playfully speculated that these seemingly very simple behaviours and insights (to us) might be the currency of the world of tomorrow.??

We consulted a very basic definition of Generative AI from WikiPedia, and circled back to what we had just discussed and noted the uncanny familiarity and likeness of a technology that is capable of learning the patterns and structure of information it is fed and applying that learning in response to a broad range of new settings and environments (identified to it through prompts).?

Note 3: Augmenting our training data through GenerativeAI?

Reflecting on how we had defined our “training data” we discussed the ease by which this could be collected and fed to a large language model.? The challenge for this didn’t feel insurmountable, nor did the idea that we might be well placed to train and prompt for the efficacy of any output being generated through an application being built on top of it.

With this established we hypothesised on how applications developed on our training data might begin to augment our capabilities, disrupt our existing working practices and reorient our areas of focus.??

Note 4: The TravelPerk Case Study on “Outcomes”

I told a couple of stories of a busy, creative and compelling 12 months at TravelPerk deploying AI powered applications (chatbots and virtual coaches) to both fanfare and failure, but with a core focus on our learnings.? TL;DR - We have seen a substantial increase to the number of successful outcomes we are delivering back to our superusers (who have 7x scaled from 70 to 500 internal superusers) and we, as a team, are becoming more and more obsessed by the accessibility of our knowledge and the environments our “superusers” might need it.??

What might have happened? Well the business continues to grow and scale, but I’d hypothesise the following are equally significant: (1) We are no longer the “arbiters'' of the truth or process (our content is); (2) our content curation and maintenance strategy is pretty good (but this won’t be unique to our legal & privacy team); (3) people are co-piloting speculatively with us now, transforming a once formal dynamic, which is also allowing us to occasionally spot or anticipate issues before they arise; and (4) we’re becoming seen to deliver quicker outcomes (contributing more flow and less work within a busy hyper-scaling organisation).

Note 5: Quicker Outcomes is great, right??

Yes, our historic capability to deliver outcomes was always constrained by: (i) time - both on the problem and how our outcomes were designed and surfaced; (ii) environment (people would have to find us); (iii) reactive and bespoke workflows; and (iv) the speed with which certain outcomes might be processed by organs of the state/judiciary.? It feels reasonable to assume that over time these constraints will release somewhat, though not necessarily in unison and a lot will depend on adoption across the profession.??

But, “innovator’s dilemma” - the whole economic basis on which we value our labour and expertise is being disrupted if outcomes can be delivered quicker, with less work and more flow - and we discussed that without the accompanying arrival of a well defined new AI business model there will naturally be hesitancy to join a race to transform and pivot away from profitable business models with deeply embedded organisational structures and working practices…

Note 6: Do we have a professional duty to..

One of the more interesting topics of the day was raised by a fellow speaker, Sir Geoffrey Vos who speculated:?

“whether lawyers and others in a range of professional services will be able to show that they have used reasonable skill, care and diligence to protect their clients’ interests if they fail to use available AI programmes that would be better, quicker and cheaper.”

In our session, we were honing in on some of the core competitive advantages we might assume through deploying and developing AI enabled applications if these allowed us to extract ourselves from the process of delivering our knowledge and expertise, we touched on the potential to:

• augment our own productivity
• expand the surface area of the outcomes we could reasonably deliver deliver/new problems we could solve/customer profiles we might serve/markets we could operate across

Final Note: There will be consequences to each of our exploration….?

Across the profession we are all aligned on this, and we considered the potential:

• For new roles and business models to be created as old ones are lost?
• To become leaner, more productive and capable - a small team or business now has recourse to superpowers that used to be the preserve of the big corporate elites
• To become the problem solvers we once aspired to be?
• For the personalisation of our “outcomes” to now truly determine the commercial viability of our businesses, practices or teams
• For both wonderful opportunities and material disruption to our working lives in the next few years

And while we’re on that journey of exploration, we should reflect on...

• The fact that our superusers or clients will still require outcomes (we are simply beginning to develop the capability to deliver more of them).
• That, as we enter the age of our “industrial revolution” we shouldn’t let the opportunity to shape it pass us by.