4 ways AI will change software — and what the Industrial Revolution taught me about it.
When chatGPT was first released publicly (2 years ago), I had the same reaction as most people.
First excitement, then a wave of fear.
I mean, Artificial intelligence wasn’t new to me, It was one of my core subjects while studying for my degree in Computer science and I had even written a paper on the application of AI in Nanomedicine.
However, what was new to me was how quickly and easily we got AI tools that were ready for mass adoption.
It was clear that the world was going to change very quickly and I had to understand what it meant for the future of tech.
To understand this change, I looked back to times in history when we experienced similar leaps in technology to see what I could learn from it.
I went back to research the work of tech pioneers like Alan Turing
I went back to research the rise of the Industrial Revolution and visited a few museums dedicated to this period.
I’ve closely observed how people use AI today.
If the advancement of AI continues at the current rate, then this is how I believe it will change how we build software products.
1. The end of ‘busy work’.
The first thing AI will put an end to is ‘busy work’.
I describe ‘busy work’ as those activities that are part of most jobs but don’t have any real value.
For example, a product manager’s job often involves writing user stories/tickets, while an engineer’s often involves setting up their dev environment.
These things make you feel like you’re working but are usually a poor use of your time or skills.
This will come to an end.
AI will shift the value of work from mundane tasks to focusing on what truly matters.
This happened a lot during the Industrial Revolution.
For example, before the automated flour mill was invented, you needed a lot of people to do repetitive tasks like pouring grain, but after the invention, there was a focus on speed, quality control and the taste of the flour.
It is already happening with AI and software today, but I predict a future where things like writing tickets or setting up dev environments by yourself will only happen on very special occasions.
Any tech worker who hasn’t yet understood where the true value of their role lies needs to figure it out now.
PS: It’s not writing user stories
2. Distribution will become more important than ever.
When massive railway expansions started during the Industrial Revolution, it greatly benefited those who were able to take advantage of it.
Businesses that were able to strike good deals with the railway corporation(s) were able to outcompete others due to cost-effective distribution.
By distribution here, I mean how you get your goods and services in front of your customers.
Even looking back to the Roman Empire, controlling trade routes like the Mediterranean was an important aspect of distribution. However, in modern times, knowing how to use Facebook ads efficiently or having a large audience on social media is how distribution works.
Distribution has always been important.
As it becomes easier to build software using AI, distribution will become more important than ever. This is because it will become relatively cheap to make or copy software products, so the more important thing will be how to get those products into the hands of customers.
Everyone will be able to make software but not everyone will know how to get them in front of customers efficiently.
Do you have a large following online? Do you have a strong professional or personal network? Are you part of any communities that could give you a unique advantage? Do you know how to run effective ad campaigns?
You need to figure out what your distribution advantage will be if you intend to remain in the business of making software.
3. Software will become very personal.
Before the sewing machine was invented, most people had 2 sets of clothing; one for work and the other for special occasions (Sunday best).
There was also very little style or customisation available for most people because it took so much time and manual effort to make the clothes you had in the first place.
When the sewing machine was invented during the Industrial Revolution, it changed how people dressed.
For the first time, the masses could access a wide variety of clothes and choose how they wanted to dress based on their style or personal needs.
I believe AI will do the same for software.
My experience in the software industry has shown me that most software users only care about the outcome they want and the tools they need to get that specific outcome.
Most users don’t care about other features of the platform or how other people use it, they only care about their use case and often disregard every other thing.
This is even clear in the evolution of the SaaS industry.
The first generation of SaaS platforms (e.g SAP, Salesforce) attempted to do everything.
Then we had the phase of unbundling where businesses picked one use case of a particular general-purpose software and focused on that use case (e.g Hubspot CRM).
Then we moved into vertical SaaS where a lot of platforms focused on one business vertical or industry e.g CRMs for Medical practitioners.
More recently, we have moved into the micro-SaaS era, where there are platforms that are so niche that they sometimes do only 1 thing, but do it really well.
Software is getting more personal and my prediction is that people/businesses will start to ‘generate’ software that helps them achieve a specific goal that they have.
Think about it.
Why do I have to pay for a tool that has 20 features when I only ever use 7? And of those 7 features only 80% of my use case is covered exactly how I want…
AI will change this.
Small teams and departments in organisations will start using AI to ‘generate’ software to solve specific problems or achieve specific goals that they have.
4. Taste will become more important.
I describe having good taste as having a strong sense of what good, bad or great looks like.
It was the difference between Apple and Blackberry in the early days of the iPhone, even though both companies had large amounts of money in the bank.
The difference between Boeing Space and Spacex…
and what could be the difference between the winners and losers in this AI-led software development era.
Since building software will become relatively cheap, having a strong taste will become very important in the software business.
In the past, you could get away with building poorly thought-out software but you won’t be able to get away with it anymore.
My prediction is that there will be less emphasis on how to build from a technical point of view and more focus on deciding what to build and how it should work.
There will also be more emphasis on why it should be built.
People with great taste in software will be the winners in this new AI world.
Start developing your taste today.
Start developing a clear sense of what good, bad and great software feels like. It will give you the ultimate edge.
Thanks for taking the time to read this. Please let me know if you think I’ve missed any strong themes on how software products will change in this new reality.
Thanks to my friend Faith for editing this!
