Shared Intelligence – Change Management for AI



The people impact of AI is too big to leave to chance. Change Managers will need to be on the front line, addressing negative outcomes of AI on the workforce and delivering the promises. Introducing Shared Intelligence – Change Management for AI.

I spent an excellent two days this week at World Summit AI 2017, in Amsterdam. There was a fascinating mix of technical companies, academics and thought leaders in AI ethics.

The talks were excellent, the tech was mindblowing, the barriers and considerations to equitable implementation, clearly sizeable.

The event has firmed my belief in the need for action in the field of Change Management to address the challenges of automation, ML and AI generally. With this blog I’m launching a new initiative: Shared Intelligence.


We’re not at General AI, but there’s a LOT happening


Right now companies are, broadly, using AI to either extend their offering or to improve their current processes. The former tend to attract the headlines with incredible stories of robodoctors and the like; but it’s the latter where I’m seeing the challenges right now.

I’ve now been working on large change programs for 20 years. In that time some patterns have emerged, and with these a healthy familiarity with the realism of claims made at the outset of transformation.

An aspect of my role that never gets mentioned in the role spec, but which is always required, is to fight for recognition of the people impact. This has been my consistent experience through mergers, outsourcing, automation and re-orgs. A Change Manager needs to fight his/her corner. Ones who don’t, end up sending out a bunch of project newsletters, organising webcasts that no-one wants and doing the go-live email.

Why? Because those aspirations of upskilling, moving on to value-add, etc. are really “difficult” compared to the perceived simplicity of IT delivery, the new org chart, and so on . So when the going gets tough on some aspect of delivery the project team tends to narrow their focus. The sponsor or steering committee take shears to the case for change and pare it back to the basics: reduce cost or harvest revenue uplift.

When this pressure arrives (and it always does), the change manager needs to take away problems, and fast. If the response is “people change is difficult, it can’t be planned like a project” you’re going to be relegated to those go-live mails. You need finite, simple, effective solutions that require finite, predictable and realistic effort. That means pragmatism, compromise and accepting you’re not going to be able to help everyone.


We need simple interventions, right now


As I listened to the presentations at WSAI17, and talked to the various experts, it struck me that there were two timeframes at work. The technical folk had solutions ready right now. Those who took responsibility for standards, ethics and getting this right for society had a longer path. Not due to any lack of effort, but due to the challenge and complexity.  This stuff is absolutely 100% required, but this presents a significant problem.

The problem is that this technology is getting implemented right now and it’s affecting people right now. This isn’t a case of heartless CEOs willfully slashing the workforce; the majority leaders I’ve spoken to have 100% positive outcomes in mind. It’s what happens when aspiration and messy reality meet.

The case for change starts with good intent for the staff: moving to better things, upskilling, new value, happier times; but that’s not what results.  The positive aspects of outsourcing were also exposed to the raw heat of implementation, and I witnessed the upside of business cases evaporate too many times.


Make the difference on the front line


So, to us change managers. As the proclaimed masters of “the people stuff” in projects we need to work with efficiency, diligence and heart if we’re to mitigate. We’re entrusted with getting it right for the people. If we can offer smart, workable, cheap and effective interventions when the heat starts to rise, we’ll have success. If we produce grief-curve charts with no end and 20 column stakeholder XLS containing half the company, we will fail. And our failure will be loss for those impacted.


I’m taking action.


Shared Intelligence is, to my knowledge, the first initiative to address AI impact through Change Management. I’m reaching out to my contacts and colleagues in AI, Ethics, HR and Change to help me develop pragmatic, workable interventions that can be used right now

I’m not developing a mega, end-to-end framework or similar, there’s enough of those competing for attention. The aim is to offer a set of simple, effective, tools that address the unique challenges of automation change.  I want to make this as easy as possible for a change project to enact. I firmly believe that businesses want to do the right thing, but that they need help doing so. I hope you can help me help them.

Leave a Reply

Your email address will not be published. Required fields are marked *