Category: Rich Pictures

“What if everyone did this?”

You already know the golden rule. The platinum rule says treat others like they want to be treated.

Useful rules, but not enough to address big questions like how do we make our organisations more sustainable?

Taking action disrupts the status quo. It introduces risk and uncertainty – and the possibility of loss. It seems safer to stick with what we’re doing now.

Change has to be worth changing for.

That’s why policy is so important. Leaders who shape the rules of the game ask “what if everyone did this?”

They create incentives that help organisations choose action over inertia.

But the rules aren’t clear right now. Many companies we speak to feel stuck – wanting to take action but finding it difficult to make a business case and get the resources and support they need.

If we want to build more sustainable companies – we’re going to have to do more to make the sustainable path the obvious one.

I keep seeing that AI has a cost problem: SaaS is expensive to build but cheap to scale; LLMs start cheap but costs go up as you grow and use them more.

I’m not convinced by this argument.

Say I’ve got to come up with an economic forecast.

There are a few ways I could do this.

Subscribe to a Bloomberg terminal for 30k that has all the data and analysis I need.

Hire someone to read and summarise material for maybe 100-400 a day.

Or run some PDFs through an LLM as a starting point for 20p.

You still need quality control so the real cost of an LLM is compute + review time costs.

But LLM + time is cheaper and possibly higher quality than time alone if you use the tools well.

What this means for me is that when I come across a business problem my first question is:

“Can an LLM help me do this better or faster?”

Before reaching for more expensive options.

Because the economics of how we deliver knowledge work are changing – and we’ve got to figure out where human judgement and input adds the most value.

This is the time of year when I’m looking at numbers and prepping reports.

And every so often I’ll see someone say that reporting is a waste of time. Stop doing it.

We’ve been managing reporting for the UK’s Streamlined Energy and Carbon Reporting (SECR) since 2019, and before that set up a trading desk for the Carbon Reduction Commitment (CRC) scheme – so been in the game for a while.

And now we have some data.

An evaluation by the government has found that “there is evidence to suggest that SECR has led to reductions in energy use and GHG emissions from organisations in scope of the regulations”.

There are more details in the report but what I want to focus on is the RADA loop – something we see in practice.

Reporting is a necessary first step. Gather data, build an evidence base, and publish findings.

Doing this gets attention – from internal and external stakeholders. Putting any statement out there needs input and approval from stakeholders across the organisation.

But how do they interact? They’re pulled into discussions, into conversations. Working groups are set up to talk about what this means now and in the future.

And that leads to action. Everything from, this isn’t important so we’re going to do the minimum to this is an existential issue and we have to get the positioning right.

And the actions then flow back into reporting, where we can monitor what’s going on.

In a decade of doing this, we’ve seen teams grow from small starts to making big impacts. Careers have developed. We’ve gone forward and back. But corporates are making a real difference – and that should be recognized.

We’re not going to solve climate change in a year. But we will make a difference if we use the decades wisely.

I started watching an AI demo yesterday that promised to show me how to build tools insanely fast.

I switched off after a few minutes.

For one simple reason – it skipped over the real work.

The demo assumed data was readily available – clean, structured, and good to use.

But when you build real world applications, getting the data is the hard part – that’s what takes the time.

It’s easy to analyse clean and tidy datasets.

It’s much harder dealing with a mess of files, formats, layouts and data entry styles.

There used to be a saying about graphical user interfaces. They make it simple to do simple things, and impossible to do complex things.

I’m seeing a similar thing with AI. We’re presenting simple things like dashboards and charts like they’re breakthrough technologies, when they’ve actually been around for ages.

The demos I want to see are different.

Show me AI working with messy data and helping with the practical problems organisations face every day.

That’s when AI use stops being theatre and starts to create real value.

Most companies don’t need new software. They need to get better at using what they already have.

I’ve seen this called “Service as a Software”, rather than the traditional SaaS phrasing.

Traditional development is based on building solutions to client problems. But developers are often separated from where the work is being done.

But if you actually go and spend time with the people doing the work, or try doing it yourself, you’ll find that it’s less about a solution and more about improving what they’re doing already.

For example, I’m willing to bet that any potential client of ours has some kind of carbon reporting system in place.

It’s probably built using spreadsheets. It was probably made in-house or by a consultant. It probably takes a long time to collect and enter data. And it probably has broken formulas. And it’s probably stressing people out as they try and get their reports out.

A solution oriented approach says – let’s get rid of all that. Here’s some software that we’re hosting – put everything there and it will be fine.

There is an alternative approach – an inquiry-led one.

This is where we go and look at what’s being done now, talk through the situation, and see how we can improve it – make things better.

Every single one of the issues I’ve listed can be solved using tools you already have, and the whole process can be improved with simple processes.

Why not wrap these processes in software and offer them as part of your service to clients?

That’s Service as a Software. The service comes first, because that’s what matters to clients. The software is implementation, but the service delivers the impact.

And that makes the difference between services that make people’s lives easier and platforms that just give them more work to do.

AI is changing the shape of how organisations are staffed.

The traditional shape is a pyramid.

Leaders at the top. Their reports below. A hierarchy of workers making up a growing base.

But AI is (maybe) decoupling time from output. You can produce the same or more without adding people.

That turns a pyramid into more of an obelisk.

Few leaders. And a slimmer working core.

Then you have the point.

Newer firms don’t need to hire to grow in the way they did a decade ago. They may stay small, a few core contributors delivering outsize value.

The one that people are worried about is the diamond, where you still have leaders at the top and an experienced middle, but the base tapers to a point.

Companies stop hiring junior talent, or hire fewer juniors – expecting AI to take over these jobs.

It’s early days. I think we can see the obelisk and point shaped organisations springing up around us. The diamond – less sure about that. But we’ll see what happens over time.

Are there any other shapes you’ve come across?

The answer to what’s going on with real-world geopolitics is in a game of monopoly I played with my son.

He is fiercely competitive. In this game, he landed on a property that I wanted.

Rather than just letting me buy it, he forced a bidding war, driving the price higher.

I had money. I could have kept bidding. But I bailed out, thinking it wasn’t worth the price.

And that was a mistake.

From that point, I was going to lose the game.

The key to monopoly is four green houses equals one red hotel.

Where are the properties that matter on the world board?

It’s where the resources are, and the ones that matter are the rare earths and critical minerals that underpin today’s technology.

These are the properties where prospective hoteliers need to get a foothold and start building their green houses.

It’s where the conflict is happening – armed conflict, and the threats of conflict.

The three big players in the game are the U.S, the E.U and China.

It’s coming down to how much they’re willing to spend and do to get control of these properties.

How they play the game will shape the economic machine for decades – and dictate the balance of global power.

This is not the time to blink, or bail out of the game.

I’m still trying to figure out the role of vibe coding in my infrastructure.

Stever Robbins had a great metaphor in his post exploring this issue.

https://www.linkedin.com/posts/stever_i-dont-know-who-needs-to-hear-this-but-activity-7421941317032775681-qosf

For years we’ve built our systems like bridges.

Look around you. Everything you see has been designed and engineered.

It is understood.

LLMs change that.

Now we have engines creating code that we don’t have enough time to understand.

Relying on these systems is like stepping off onto a tightrope. Will it hold? Can you stay on? And what happens if you fall?

But tightropes are cheap and easy to install. And maybe it doesn’t matter if you fall off sometimes, as long as you can pick yourself up and carry on.

I suppose the best thing is to be pragmatic.

Build, test, learn, apply.

I was reading about the smile curve yesterday and how it’s changing the rules of the game.

This is the idea that we open our wallets only when we must do something, or if we really want to do something.

Taking buying a house, for example – one of the biggest decisions we’ll make.

The Economist wrote that there are only two reasons to buy a house.

Either it’s cheaper to buy than rent, or the house is perfect and you really want to live in it.

A so-so house, in the middle of a so-so place – you’ll pass on that.

The bottom of the smile is where products and services go to die – the things that are nice to have, but that you can’t really justify right now.

And this means we have to design products and services differently.

A service business like ours uses tested systems and processes to do work that clients have to do because of government rules.

Another business I know massively increases sales for its customers through a platform that quickly matches orders to the cheapest suppliers.

Who wouldn’t want to sign up to that?

Growth, it turns out, is increasingly at the extremes.

As business builders – we need to figure out where we sit on the smile curve and then head in the right direction.

We’re building a firm that’s remote first and makes the most of AI.

But what does that mean in practice?

Twenty years ago, we hired 30 people to build a team.

Ten years ago, we hired 5.

Now – we don’t need people to fill roles. We need them to bring creativity and knowledge.

The roles we hired for previously – data entry, spreadsheet wrangling, manual checks – they’re starting to disappear.

We’re able to do more and more of that kind of work with technology.

But you can’t just think of AI as a cheap employee – a robot doing exactly what was done before, just for less money.

You have to first change how things are done to generate value.

Take making bread for example. I’ve spent much time over the years kneading together flour, water, salt, sugar and yeast to make a dough.

If I were to design a replacement how would I start? Would I try and create a machine that replicates what I do?

The KitchenAid mixer doesn’t do that. It has a spiral tool that does the job in two minutes.

I haven’t hand-kneaded dough in a couple of years.

In the same way, I think we’ll move to a different, more collaborative approach in the way we work with others, at least in boutique consultancies.

Because in every business, we have to redesign how the work is done – and that starts with recognizing the value that people bring, and how it’s different from the value that AI brings.