For all the ‘big’ reports from Deloitte, BCG, PWC etc, I prefer the more measured views from people within organisations who are doing smart stuff and learning from it. One I particularly liked, as it chimed exactly with my own experience on implementing AI within small and large organisations comes from IBM.
Let me introduce you to Nickle LaMoreaux, the CHRO (Chief Human Resources Offices) at IBM.
She sees HR as ‘client zero’ for lots of initiatives. I like this up front statement, as HR is so often simply reactive or sidelined into initiatives that are not focussed on business improvement.
It just seemed reasonable from someone who is actually doing stuff, not just talking about it of doing surveys. She’s practical and has to deal with the bottlenecks, cultural resistance and hierarchies within a large organisation. We should listen.
What to do? The 3 cs!
1. Consumer-grade experiences: delightful experiences enabled & enhanced by technology.
I loved this. Go for time saving and increasing quality but if you want to effect change, go for something visible, that touches people. GenAI clearly does this as the billions of uses per month show that dialogue works. It is simple, engaging and delivers what people want. It gives them agency.
2. Cost efficiencies
Cost efficiencies are so 2024, so she also went for a solid project and measured the efficiencies in terms of time saved and quality. “HiRo saved 60,000 hours for Consulting managers in a year”. That’s what I want to hear, a solid project with measurable outputs. She unashamedly wants to see a high return on investment. This is refreshingly businesslike.
3. Compliance
Spot on. This is the one area where AI can be leveraged to reduce time and, increase quality in processes and training. As she rightly says, it is “perfect fit for AI”. Complaince has become a nightmare at the national and international level. AI can really help in terms of processes, support, learning and keeping everything up to date. This is one area where the data side of AI can really help.
Identify Some Quick AI Victories
Don’t procrastinate – that’s the easy choice. AI is the here and now. The tools are available, use cases identifiable so don’t give into to the ‘wait and see’ attitude. She explains how to proceed by winning some smaller battles, not turning into a strategic war. This is correct. This is new, powerful technology, it has huge potency but also some vagaries. We need to proceed clearly but also with caution. What’s the point of creating an unnecessarily negative climate for future progress on the back of some failed and over-ambitious masterplan?
Choose “High-volume, repetitive tasks. Processes employees don’t enjoy. Moments that matter for employees.” This is so right. Large organisations are full of processes, which are tedious, bureaucratic, have bottlenecks, old technology solutions and are ripe for automation. They are often deeply embedded practices. Focus on these and the quick wins will come.
How to do it?
1. Start small and experiment
This will vary across organisations but I doubt there’s a single organisation, small, medium or large that will not benefit from the use of AI across a range of uses. These need not be grandiose or abstract, such as skills recommendation prediction or large data analysis projects. They can be simple and manageable.
2. Learn as you go, fail fast & be agile
Can’t emphasise this enough. This technology can be implemented fas=st against a goal. Don’t get bogged down in huge project plans – get on and do mit as the technology will get better as you process. It is not that you may have to pivot in some way on the technology and approach – you WILL have to pivot.
3. Lead with use case, not technology
This often comes down to finding a pain point, bottleneck or hated process for a quick win. It doesn’t have to be big just impactful. Make sure your goal is clear.
4. Cover off data and security issues
You have to establish trust and allay fears in the projects chosen. This often comes down choosing a technology partner whose vision, goals and purpose align with your own. But some simple FAQs to vapourise the myths and calm fears is wise.
5. Build advocates out of your employees
I have seen this for real. Young employees, let loose to use this technology and show their worth. People who do process know what’s needed to redefine, short-circuit and improve those processes. Give them the agency to suggest projects.
Clear goals
I’ve seen projects founder when it was not clear what the final goal was. AI projects tend to shapeshift and that is fine in terms of swapping out LLM models used and changing tactics – this is normal and easier than people think. But don’t founder on the rock of vagueness. Be clear about goals
I’m not a fan of old-school SMART objectives, as what is needed is a solid goal and sub-goals. This FAST framework, from MIT, sums it up perfectly:
Conclusion
We are in the first phase of seeing the benefits of GenAI in organisations. Before we head off to climb that great single peak in the distance, take time to conquer a few foothills. Choose your hills carefully and once chosen, make sure you focus on them. Stay ambitious, measure that ambition and shout it from the hill top when you get there! Thanks Nickle.
No comments:
Post a Comment