2030 - A Ruthless Future?

AI is creating a power-law economy where the best users pull further ahead, get ready for a more ruthless future.

Chris Dunlop

Jul 19, 2026

We are standing at an exact midpoint.

ChatGPT launched four years ago, in late 2022. The end of the decade is four years away.

Imagine going back to 2022 and describing 2026 to yourself

Models that write production code
Models that plan multi-step work, and run for hours without supervision.
AI agents
Claude

You would have switched off. The description would have been too far-fetched to engage with, so your brain would have filed it under science fiction and moved on.

It’s absolutely insane what has happened in 4 years and yet I’m sure your past self from 2022 would have said “oh well if all that is happening, then the business world must look entirely different” and you’d say, “well…. not exactly…. you see, it kind of looks exactly the same.”

It’s all a bit baffling. This effect goes by two names for me:

One is the matrix, where there are some people that have taken the red pill and can see what the world looks like and others who haven’t taken the pill. Those people have never tried AI and have absolutely no idea what is going on.
The other is Formula 1, where I think the future will belong to people that can operate at an extremely high intellectual capability.

If I could sum up how I feel about the next four years in one picture, it would be this.

By 2030 we know that things will be even better, the models will be smarter, faster, and cheaper and will be able to do more than ever before.

But the reality is that right now, if you aren’t using them to their full potential then who cares. Who cares how good it gets? It’s already so smart. What does it matter if it’s ChatGPT 8 or Opus 9.7 - you don’t need those to draft an email or do 99% of the tasks you are currently doing.

This is a huge problem. As it means that you aren’t benefiting from the progress.

Instead, it will funnel to those who can and that is why I think it will be a ruthless future.

So my prediction is that, from this point on, this will be a rich get even richer world and a smart get smarter world and the gap between the two will grow and grow.

There is already evidence to support this.

90% of code on Cursor is written by 10% of users.

The 99th percentile developer is producing 46 times more code than the median developer on Cursor.

I know hardly anyone that has even used Cursor. I would say Cursor users are already a power law themselves. It’s no exaggeration to say that the 99th percentile Cursor user is probably 1,000 times ahead of the median Non Cursor user.

Then we go to Stripe

Top decile solo founders are earning 61 times more revenue than median founders, up from 34 times 4 years ago.

The evidence of all of this is clear, we are living in a world of Power Laws. The benefit of AI has not been this widespread equal thing, where we all hold hands as businesses and get 5% more productive together. It is certain people, with the right attitude extracting the full value out of the technology to their advantage and everyone else is essentially in the same position as they were before.

So I don’t need to be Nostradamus to say that I think 2030 will only accelerate this gap and in fact I think this paints a pretty ruthless world picture.

You have already seen the ruthless world play out in the stock market for SaaS companies, where it is just a see of red where ever you look.

Maybe they are oversold? Maybe they aren’t? I don’t know the answer, but it shows a clear example that we have entered a new era and things are uncertain.

This is the ruthless side of everything. If you are on the wrong side of trends, like Figma is, then your business can get absolutely pumped.

And guess what, this power law is playing out in the stock market as well. With the top 7 companies accounting for the majority of returns of the S&P500 in the past year.

So what do we do?

Are we all stuffed?

Do we just sit back and invest in Alphabet and NVIDIA and then wait for them to make all the money while we just chill?

The season of learning

Well, the first thing to realise is that you have time. AI is not mainstream yet.

So if you are in the camp where you aren’t a 99th percentile Cursor developer and you don’t know anything about AI then you need to dedicate some time to learning. But learning comes with a big caveat.

There is Harvard research showing you cannot optimise learning and performance at the same time. If you try to treat a moment as both a learning moment and a performance moment, you do worse at both. Most leaders I meet are trying to perform their way into 2030 and trying to ‘learn a little bit about AI’ as well. This approach will fail.

You are better off to enter a season of preparation and learning. Honestly put aside your growth targets for the rest of 2026 and just learn about this new technology and actually immerse yourself in it, rather than dabbling.

Like how a sports team will lose some games in the short term to setup to try to win a world cup in a few years.

The reason why is that like I said, I believe that from this point on, the gains in AI will only go to the top of the power curve and so you have to understand what is happening to benefit from that. If you have never installed Cursor or Codex or Claude Code then you will be missing out on the gains as everything is currently being plowed into those platforms.

This is the season for re-evaluation. When the internet was growing, Jeff Bezos quit a freaken hedge fund to start an ECommerce book shop. If he quit a hedge fund, you can reassess what you are doing.

Become AI complete

What is the goal with the learning? When have you learnt enough?

While of course it is a constant process, I believe that after a few months you will develop a good mental model of the AI world. But here are the key targets that you are aiming for.

To develop your own intuition and judgement. You need a radar to understand changes. Maybe you have a little mastermind group to help you with that. Maybe you set aside time every month to test the latest things and run some experiments. You need to be active in this period of change.
You want to become AI complete.

At Cub we have a name for the exit condition of a learning season. We call it being AI Complete.

You are AI Complete when the AI getting better makes you better automatically, with no further redesign required. We set this as a test for the long strategies we built with the Olympic Team: when we planned in 2020 for the 2028 Olympics, the requirement was that every capability gain in the underlying technology flowed straight through to the performance of our system. Or in other words, we didn’t need to compete and rebuild something the underlying AI labs were building.

Google did a similar thing, they would plan for Moore’s law to carry them and make things profitable in the future. That’s why when Gmail and Google Photos started, they had such crazy storage volumes at the time. Google was banking on storage getting cheaper and turning businesses that weren’t viable into viable entities. The same goes for YouTube.

Individuals need the same property. When the next model lands, can you pinpoint how it has uplifted something that you are doing? If you can’t, then that is a worry. That means you aren’t using the system to its potential.

Even better, do you have a task that is extremely expensive today, that you know will get cheaper to do when the models improve in 4 years?

Business fundamentals are still important

Here is the part that is reassuring. Even though we are entering a ruthless era, technology does not exempt anyone from the fundamentals.

Hamilton Helmer’s Seven Powers still applies. Counter-positioning, scale economies, cornered resources: the laws of gravity are intact. If you are in a structurally weak position, the tools will not outrun the physics. If you are in a strong one, the tools are the biggest amplifier of that strength you will ever be handed.

The strategy work still matters. Where to play and how to win are still the questions to ask. The learning season is what earns you the right to answer them well. So make sure in that season that you read some books.

Plan for your future context

Last week I wrote that context becomes the company. If you missed it, the short version is that the choice ahead is usually framed as a binary: hold everything permanent, or dissolve everything into fractional, assembled-on-demand capability.

After you have spent time learning, I would spend time thinking about what context means in your business and where your sources of power will be derived from.

So what do you want to be and what do you want to do?

In 2020 we ran an exercise I recommend to every leader. We asked the most outlandish question we could construct: with unlimited money, what would we build? This is actually hard to answer, as no matter how hard you try, you will still limit your ambition.

Dreaming like that is genuinely hard. You never get to be in that mode. Kennedy chose the moon because it was hard. Shackleton advertised terrible pay, bitter cold and small chance of return, and men signed up. Most mission statements today would embarrass both of them.

Which brings me to the only question that matters for the next four years. A line I keep returning to from Dharmesh, the founder of HubSpot: I like AI less for what it is and more for what it makes me.

So. Who are you, and who do you want to be? The tools will let you be almost anything. That was never true before. It is true now, and it is the most demanding gift a generation of leaders has ever been given.

What is the most ambitious version of you and what you want to do? What is your lifes mission?

I think if you put this together, you plan for a season of learning, you read the fundamentals of business strategy, you plan for your future context and for a really ambitious project or mission then no matter what is thrown your way, you will be prepared for success.

A lot is set to change in the next four years, just make sure you capture some of the upside.

Ways To Work With Me

I’ve started to get more and more enquiries from SubStack about work so I thought I would make it clearer on how we can help you. We currently have slots for 2 new customers, starting in September.

Through my agency Cub Digital we can help with the following:

AI Development, website development, strategy.
If you are in financial services then we have a new AI agents tool. Particularly good for analysis.
We are accepting sport enquiries and have a range of AI products for different use cases.

Feel free to send an email to chris@cubdigital.co.nz or send a message on LinkedIn.