From vision to action: How can you build on the UK’s AI Plan?
The UK government recently announced a plan to turbocharge AI. The press release from the Prime Minister’s office explains that “artificial intelligence will deliver a decade of national renewal.”
Our first reaction to this news? We’re thrilled to see our government take such a proactive and positive stance toward technological change. After all, getting ahead of transformation is not something governments are always well-known for doing!
But of course, there’s more nuance to it than that. I teamed up with Dr Shahzia Holtom, Global Head of Technical Excellence at Microsoft, to put together some thoughts based on our decade of experience guiding companies through digital transformations.
The good
Here’s what we think is promising about the plan:
It comes directly from the Prime Minister and urges every member of the Cabinet to make this a top priority for their Departments.
Within organisations, there are often many competing priorities. If the most senior leader isn’t helping people understand that something is a critical priority, it’s all too easy for people to fall back on business as usual. When the leader sets a clear priority, it creates clarity and momentum for the team to rally around. The fact that the Prime Minister reached out to each member of his Cabinet also creates a shared accountability that gives this plan much more of a chance to succeed.
The plan isn’t just focused on immediate benefits: It sets out a strategy that will guide us through the next decade.
Often, in the business world, there’s an overemphasis on the short term. On average, the C-suite changes every few years, quarterly business results and individual performance reviews still take priority. In these conditions, it can be challenging to think about something beyond the next year. But because this plan is set to take place over the next decade, it allows us more space to consider the wide expanse of what’s possible. It feels like a more optimistic point of view than what we often see in business about cutting budgets and doing more with less.
There’s thoughtfulness around which industries are best suited to AI transformation, including energy, where the potential for adoption and impact is significant.
We’re thrilled to see that the plan highlights certain industries, like energy, which already have maturity in terms of data and predictive modeling, which make them well suited to AI transformation.
And this is something we don’t always see in industry—there are many times when companies will invest heavily in a proof of concept, only to realize that it can only benefit a small number of users, or data quality is insufficient for impact at scale.
Watch out for blockers
As we’ve learned from our decade of experience helping companies navigate digital transformations, we know there are potential pitfalls to be aware of, too. Here are a few points we’d like to highlight.
It’s important to be aware of limitations we’re likely to face.
The VC world often looks at limitations through a few lenses: the data wall, the compute wall, and the model wall. The data wall involves getting to the point of data collection that will deliver a significant ROI to your business. It’s one thing to build a quick and dirty prototype on a small collection of data, and it’s a very different matter to really put something into production and have a business-wide impact. You need to determine whether you have the resources, runway, and money to close data gaps.
The compute wall involves determining whether you have the skills and the people to make this happen technologically.
And the model wall involves asking, once you’ve built the model, can you maintain it and monitor it?
You need an integration plan. It’s not just about putting something into production, but setting it up for long-term maintenance and adoption.
We need to always keep the people and the problem we’re solving for them top of mind.
It can be easy to get dazzled by the technological capabilities, but if they’re not solving real-world problems or they’re creating more friction, that’s ultimately not progress.
Within the business world, specifically within product, we always need to ask what user need we’re solving and whether someone would be willing to pay money for it. And as exciting as the new technology is, the answer can’t just be “I want to implement AI.” We need much more of a business case.
An invitation to you
Of course, we’ve just scratched the surface here. There’s much more to unpack in terms of what’s likely to work well and what will require further refinement. We have so many more thoughts to share on this topic, so watch this space for more updates!
We’re also curious to hear from you. Do you have a case study or success story of AI adoption you’d like to share? We’re looking for people to feature in our forthcoming book on this topic. Please get in touch to let us know if you’re interested in participating.