- Steal These Thoughts!
- Posts
- How 3 Smart Companies Are Using AI For Learning And Performance
How 3 Smart Companies Are Using AI For Learning And Performance
Ross Stevenson
March 26, 2024 • Estimated Reading Time: 14 minutes
Today’s Thoughts ☠️
Ahoy there 👋,
I’m torn on what to share with you today.
My problem is never coming up with ideas. It’s knowing which are meaningful and which to kill.
I spend more time killing ideas than creating them.
I’ve settled on something I believe we’ll find impactful (tell me if I’m wrong). Not far back, I spoke about practical applications of Generative AI in L&D. I want to delve deeper into this.
Today, we’ll explore how 3 smart companies are using AI for learning and performance.
→ Much love to today’s sponsor, Sana ❤️
In today’s chat👇
Est reading time: 10 mins
TOGETHER WITH SANA
The 8-Step Blueprint To Get Your Learning Tech Business Case Approved
Here’s the picture.
You’ve found the platform you know will enable and enhance learning at your company. But, you need budget approval from the big board of stakeholders.
Fret not, as Rita Azevedo, Ex-Head of Learning at Klarna has you covered.
Check out her guide to learn how to:
🛠️ Build internal buy-in by conducting internal user research
👀 Reduce risk by testing your shortlisted tools with a team
📣 Strengthen your case through persuasive storytelling
THE BIG THOUGHT
How 3 Smart Companies Are Using AI For Learning And Performance
When someones does something useful with AI
I don’t need to tell you that we are at peak AI hype.
Despite all the hype, I’d bet less than 1% of organisations are meaningfully using any Gen AI tool. So, we have this weird divide. Great tech but no clear impactful applications for the working masses (yet).
By meaningful and impactful AI application I’m specifically referring to business performance and thus increased profits.
I’ve seen no direct connection between a ‘ultimate prompt’ and impact on the bottom line. The good thing? Time is on our side.
I’m delving more into the Gen AI → business impact topic by reading a new book from Verity Harding, Director of AI & Geopolitics at Cambridge Uni, called “AI Needs You”.
Let’s get into the good stuff.
We were all surprised by generative AI in 2023.
But, the next 2 years (24-25) is where we take the fabulous tools we have from gimmicks to meaningful assistants. You don’t have to start from a blank page though. We have plenty of examples and case studies available now.
Here’s some of the best ↓
BHP Mining: Improving leadership frameworks with ChatGPT
TL;DR
BHP, the mining giant, leveraged ChatGPT to revamp its leadership framework, ditching traditional consultants for AI (sorry consultants).
They fed ChatGPT a 90-page culture assessment and other crucial documents to get insights on improving leadership approaches to match new cultural goals.
ChatGPT provided insights to refine these leadership qualities
It made the often confusing HR language more relatable to frontline staff too
The problem they were solving
BHP were keen to update its leadership framework to align with new cultural aspirations.
Normally, this task would require hiring consultants for a project spanning weeks. Which, as I’m sure many of you know, takes a lot of time and a lotta money!
The goal was to make the leadership model more reflective of where BHP needed to go. Plus, they needed to improve the language used in HR documents to make them more understandable to frontline teams.
I think we’ve all stared at HR documents wondering “What does this mean?”. Using AI to decipher the HR lingo is a big win in my eyes.
"Normally this is a job you’d farm out to consultants, it’s a six-week project. We thought, ‘Why don’t we give it a go using ChatGPT?”
How they did it
Instead of going the conventional consultant route, BHP turned to ChatGPT.
They uploaded a 90-page culture baseline assessment and HR policy documents into ChatGPT.
CGPT was tasked with analysing these materials to suggest how BHP could reshape its leadership to better meet its future objectives. The process involved identifying gaps and opportunities and rewording documents to make them more accessible to all employees.
"The chat started with some fairly basic processes that we’re reviewing, where are the gaps and opportunities. It then moved to improve the language because one of the critiques we have in the way we write these documents. So you say to ChatGPT, ‘Express that or rewrite that in a way that [a] worker can relate to it and engage with it.”
Sheahan also noted the AI's practical advice, "Not only did it pick up the things that were missing, it said a framework is one thing but it’s the implementation that matters.”
Return on Impact
ChatGPT proved to be more than up to the task.
Yes, PWC, McKinsey, Deloitte and the rest of the crew. Be scared.
CGPT offered meaningful suggestions that BHP found valuable. The AI's ability to digest extensive documentation and provide actionable insights surprised the team.
Reflecting on the AI's analytical capabilities, Sheahan was impressed: "That’s quite an abstract, deeply analytical question - and its response that came back absolutely blew my mind in terms of its ability to look at it.”
The Empire Bank
Morgan Stanley: Your local AI banking nerd
TL;DR
Morgan Stanley launched 'AI @ Morgan Stanley Assistant,' an AI tool to improve financial advisor’s access to the bank's vast database of research reports and documents.
Built on OpenAI's GPT-4 software, it streamlines the boring admin and research tasks, allowing advisors to focus on client interactions.
MS were one of the early adopters of GEN AI.
The problem they were solving
Morgan Stanley needed a speedy way to access and analyse the bank's stack load of over 100,000 research reports and documents.
Can you imagine the stress of finding all that in archaic bank tech? I’m having heart palpitations just thinking about it.
Financial advisors needed a more efficient way to find all that money-making information. No one wants to sift through vast amounts of data. The other problem was more time searching meant less time with clients to get that dollar, dollar!
How they did it
Morgan Stanley collaborated with the smart folk at OpenAI to develop their assistant.
The assistant provides quick and easy access to the bank's intellectual capital. Enabling advisors to ask complex questions and receive concise, relevant answers (we hope, anyway). The development process involved curating documents and extensive testing with human experts to ensure the AI produced high-quality responses.
As much as you can control a probabilistic system, of course.
"I've never seen anything like this in my career, and I've been doing artificial intelligence for 20 years... We saw a window of opportunity that was just completely disruptive".
Return on Impact
The launch of 'AI @ Morgan Stanley Assistant' shifted a few things.
Reduced the time advisors spent on admin tasks
Increase time with clients for enhanced human relationships
That’s exactly what we should aim for.
Enhancing human activities by delegating the blockers to that. Don’t get confused with AI, use it to enhance human stuff, not lose it.
My one feedback on this one would be the name MS gave the tool.
I mean, was ‘AI @ Morgan Stanley Assistant’ the best that could be done? I suppose, they’re are a bank after all.
Consultants vs AI
McKinsey & Co: Consulting on command
The consulting gods must have heard the news about BHP above.
Companies with the ability to leverage AI to do classic consulting work are no good for business. so, what do you do instead? Send the consulting mafia? Perhaps.
Or…you get smart too.
McKinsey took the latter approach. I would have loved to see a consulting mafia vs AI, personally.
TL;DR
McKinsey & Company introduced ‘Lilli’, a generative AI tool designed to empower its team by providing quick and efficient access to the firm's extensive knowledge base.
Just like Morgan Stanley’s assistant but with a much better name. Lilli serves as a researcher, time-saver, and source of inspiration, streamlining the search and synthesis of McKinsey's vast stores of information.
Lilli represents a pretty good use of Gen AI in enabling McK & Co to leverage its intellectual capital at speed.
The problem they were solving
I’m not going too deep on this one.
The Morgan Stanley problem, and thus, the problem with surfacing content is shared here and millions of organisations.
What good is all that content if it’s locked in Dorothy’s secret Sharepoint site?
McKinsey sought to overcome the challenge of efficiently accessing and synthesising its vast knowledge resources. Conversational AI is a tool that could quickly bring together its best insights to support client engagements, without the limitations of traditional research methods.
Sorry, Dorothy. This means the secret info hoarding on Sharepoint is over.
Remember to delete those AI-generated cat pics!
How they did it
Simple. They paid someone to create a Large Language model (their own ChatGPT).
They graced it with the name Lilli and threw it out into the McKinsey-verse. For (probably) the first time, the firm’s knowledge and capabilities were only a conversation away.
Lilli is designed to provide an impartial and streamlined search across McKinsey's extensive databases. This makes it easier for teams to find relevant information and insights. That line sounded overly corporate. I must say.
Erik Roth (or obvious Eric as I named him while writing this), a senior partner at McKinsey, preached the benefits of Lilli, stating:
"It aggregates our knowledge and capabilities in one place for the first time and will allow us to spend more time with clients activating those insights and recommendations and maximising the value we can create".
Or in human words: We’re outsourcing the shit stuff to make more money with humans.
I get you, Eric.
Return on Impact
As with the folks at Morgan Stanley:
Less time running around trying to search for that ‘one golden document’
More time schmoozing human clients for that sweet-smelling paper dough (aka money)
Dorothy can also find her cat pics faster
Final Thoughts
Opportunities are everywhere
This is a long game, not a short game
Explore and experiment in your work
If you want to work smarter with GEN AI at work. Let me help you in my 2-hr Crash Course for L&D pros. Get everything you need to cut through the AI chaos.
I hope this helps.
Enjoy the newsletter? Please forward to a friend. It only takes 10 seconds.
How I can help you👇
My L&D toolkit stack + AI For L&D Lab with 50 zero-cost tools.
Work smarter with AI in L&D with my AI For L&D Crash Course: A 2-hour AI Crash Course for Next-Gen L&D Professionals. Join 300 + students to future-proof your skills and work smarter with practical use cases in L&D.
→ Access all courses at the STT Dojo.Join The Cult of Thoughts 🏴☠️: Get access to exclusive deep dives, events and perks with our premium membership - The Cult of Thoughts 🏴☠️. Your investment gets you more great content and helps me keep doing this for you until the world’s tea supply runs out.
→ Exclusive subscriber-only Monthly newsletter
→ Private catalogue of toolkits and insights
→ Workshops, events and perks
Stalk me on LinkedIn for daily insights, tools and frameworks for modern L&D pros.
SMART THOUGHTS
Stuff that caught my eye this week.
🤔 Are we in a AI arms race?
The short answer: Yes
Do we need to be? No
Verity Harding (mentioned earlier) talks to Prof Scott Galloway about the problem with politicised AI amongst other topics.
This is worth 31 minutes of your time.
💡 The corporate Gen AI upskilling conundrum
There's a big gap in AI training.
It's baffling.
You hear lots about GEN AI tools on social, yet very few people are using them in the real world.
Why so? Learn more.
🔥 Why is performance consulting so important for L&D?
A lot of L&D pros tell me they don’t feel valued.
They feel like a conveyor belt of fulfilling orders vs delivering meaningful work. I get it. That sucks.
For that to change, you have to change.
You have to move from transactional to conversational. From McDonalds drive-thru to performance. I’m not saying it’s going to be easy.
→ Here’s what you need to know to deliver meaningful performance impact.
Join the conversation