Since generative AI became headline news and a household topic in late 2022, organisations have been exploring how to wield its potential to meet their business needs. The truth is that, in an attempt to get ahead of the curve and stand out from competitors, a lot of companies are rushing into mistakes without having the necessary conversations.
To discuss the importance of the right people and skills when implementing AI, I spoke to James Hutt, who’s helping organisations navigate technological evolution as a consultant at Paradigm Junction.
Thank you. Our reason for being is to help organisations understand what technology can do and its impact on their operations, from business strategy to their people and ways of working.
There are two common major problems for our clients. The first is that technology is evolving much faster than the rate at which people become senior in an organisation. This means that executive teams, like the ones we work with, are being asked to make decisions about ways of working, operating models and technology stacks with which they’re unfamiliar.
The other is that there's so much information out there about technology but most of it's irrelevant to many organisations. It requires time and dedicated expertise to filter this information and decide whether it affects their business.
As of late, all of the conversations have been around AI.
Senior stakeholders first need to ask three questions that'll determine really where they should be going.
How significant is this? Does this completely upend your value chain or change your way of working?
The next is, how urgent is this? Will competitors challenge you soon, or can other people emerge, adopt new technology and do what you do?
Lastly, how are you set up to succeed? Do you have in-house expertise? Is your data estate in good order?
Your IT function is going to be crucial for understanding what’s possible technically and the dependencies you already have in your organisation, particularly if you're a mature company. HR directors are going to be looking at what skills you have in your organisation and how those are going to change.
The next role that I'd say is crucial is a senior generalist or someone in a senior commercial role. They're going to know whether speeding up certain processes matters. If we can do this 100 times more, can we reach 100 times more customers with it? Will it move the needle on our bottom-line results?
Then probably the next group is the risk functions, people in legal data security or cyber security. The best analogy I've heard in this space is about the brakes team on a Formula One car. Often people talk about brakes as slowing cars down and people talk about these risk functions as slowing down the change. Actually, they're the team that allows you to go quickly by knowing that you're going to be able to stop safely. Involving them right at the end is far too late - you need to involve them in the design process so that they can say, “if we have this in place, then we're safe, we're good to go”.
Lastly, if we're looking at projects rather than a strategic overview, someone with hands-on experience in the function will be absolutely crucial. They’ll be able to tell generalists whether a solution won't work in reality because it misses crucial nuance. They're the ones who are going to have real understanding of how your process really hits the road for customers.
Let's start with technical skills. I saw a brilliant job advert the other day that asked for 10 years’ experience with generative AI, which was surprising given that the paper that launched this field only came out about six years ago! So, everyone is in some sense new to this. And part of what makes it such a difficult field is that everyone is learning on the job about best ways of working.
The best adjacent skills are data science and data engineering. Without these, I suspect you're going to be buying things from the market. In that case, you're looking at a different set of technical skills - the ability to monitor, evaluate and assess the impact of technologies. These will be crucial because, otherwise, there are really sophisticated sales efforts out there and you're going to fall foul of them.
I mentioned HR, but it’s better to say a broader area of change management, which is a skill that can be a bit difficult to pin down. There are three areas: identifying what you're going to do, unifying people behind doing it, and managing the friction that is caused by doing the new thing. What does that look like day to day? Expertise in mapping your processes, systems management, identifying requirements and communicating between teams who typically speak their own jargon. This is essential for bringing on board technology that changes the way we work, rather than sits on an employee's desktop and doesn't get used.
Finally, lots of people in your organisation are incentivised to raise a red flag if they think that something's risky. It’s up to leadership to look at the mitigations in place, weigh everything up and, in some cases, push through.
This generative AI wave has attracted so much attention because, for the first time in a generation of technologists, we’re looking to change how we work rather than simply replacing work that's already going on, wrapping it up in a box and automating it. And that change is really not a very technical one.
I get slightly worried when I see a CTO leading this. Yes, we're using computers to do it, but we also use emails to have conversations - that doesn't mean that all emails should come under the purview of an IT team! This is much more about commercial direction, business strategy, the value you provide.
There are several examples of technologists saying, “wouldn't it be really cool if we built a product and one of our teams could do this?”, but they don't realise that people don't do that very often. People spend more time doing something mundane that isn’t really cool, but it could have a much bigger business impact.
Those cross-functional stakeholders can be really good at translating between those requirements and what matters and what's possible.
An organisation called Caution Your Blast worked with the Foreign Office to build a tool that would assist their contact centres. The tool would read messages, quickly identify a helpful pre-written response and send it out, rather than wait for a customer service agent to get around to it. One of the contact centre leads was involved in the project throughout. They understood the existing processes and were essential to assessing whether the answers generated by the new tool were actually good or simply looked good on the surface.
It’s been estimated that this will save time for hundreds of people over the next few years. The real hope with this is that citizens, when they get in touch, are going to be able to get easy, quick answers and free up time for contact centre agents to work on the more difficult cases.
Another example that we've been part of was a group of workshops with Chief Information Security Officers from a number of the UK's biggest companies, being run in partnership with Crossword Cybersecurity. We helped them identify the problems that are going to emerge as companies adopt generative AI, and then what should they start working on now.
It goes back to this Formula One brakes analogy. If you take two years to start looking at what problems there are and what mitigations you need in place, your opportunity to have impact is going to be much more limited. You only really get the chance to say “yes” or “no”, rather than something more nuanced.
To explore the right solutions for your organisation’s needs regarding AI and other tech solutions, get in touch with Hays or Paradigm Junction today.
James Milligan Global Head of Technology Solutions
James Milligan is the Global Head of Technology Solutions at Hays, having joined in 2000. In his role, he is responsible for the strategic development of Hays' technology businesses globally.
