I talk to company leaders every day about their AI visions. Most of them are starting from the complete wrong premise.
The vast majority are rooting their investments in software-era logic - looking for automations that fit into their existing, rigid infrastructures.
But AI isn’t a new technology to plug into your old stack. It’s a fundamental shift in how your organization will operate - and if you don’t shift your mindset, you’ll struggle to build AI-native workflows and teams. Here’s what you need to wrap your head around now to move quickly.
The 3 big changes you need to grasp as a leader
We’ve been SaaS-washed over the years to view work through UIs and subtasks. AI is going to change all of that in three big ways.
1. AI will become the platform
We’re already seeing signs of a future where you open up an LLM to accomplish a task instead of a browser. ChatGPT recently launched integrations with companies like Expedia and Booking.com to help plan your travel - more of that is coming.
Companies that rely on their app or website to generate revenue (via ads, etc.) are going to be rendered unnecessary soon. AI won’t be a plug-in, it will be the environment. It will sit at the center of digital interaction and slowly remove the need for apps and third-parties by making that functionality native to its makeup.
2. Rigid software will become fluid data interaction
Using software has always been about manipulating databases. The meetings we schedule, the files we send, and the programs we run are all housed in databases, and we recall that data to perform tasks. We're used to websites and apps that make us manually push data through rigid, step-by-step workflows.
Databases won’t go away, but AI will add something radically new - a middle layer that sits between us and the database, and eliminates friction by making database manipulation intent-driven instead of declarative.
In essence, instead of telling the system what you want it to do, you’ll tell it what you want. Instead of following a rigid workflow, AI will build its own plan for achieving your requested outcome.
This is already happening – Claude can dynamically build UIs for you from a prompt, agents can autonomously create deliverables – and it will only become more ubiquitous.
3. Knowledge will be encoded into tools
This one is going to trigger your nervous system: the most valuable workers of the AI era will encode their own expertise into AI systems.
It sounds self-sacrificial: Why would I train an AI to do my job and risk getting laid off? But don’t worry about that. The need for human novelty will never go away, because AI isn’t good at coming up with new ideas.
Instead, AI is going to take over the mindless work – the hours of coding or data analysis – and allow humans to do the big picture thinking and strategizing.
Why all this matters: AI will replace your current infrastructure and rewrite how work gets done on your team – and sooner than you probably think. You’re at risk of being disrupted if:
- Your business creates interface friction - through non-AI browsers or sites, rigid infrastructure, or tools that don’t do the job for people
- You don’t adapt your processes to more efficient, AI-driven ones faster than the competition
How to set strategy around these capabilities
Today’s leaders need to:
1. Hold two mindsets at once. Your business still runs on SaaS (or sells SaaS) so you have to maintain it, but you can’t deny a world in which AI replaces the need for your current infrastructure or product. So straddle both realities: Invest in the future while keeping the present running.
2. Invest in AI strategy, even if it feels early. If you aren’t working on your AI strategy now, you’re shirking your responsibility as a leader. Even mechanics are using AI diagnostic tools – there is no single business that will not be disrupted by this technology. Existing frameworks like Porter’s Five Forces can easily apply to AI, so no excuses. Figure out how you’ll be disrupted and make a plan.
3. Orient toward long-term transformation. The expectation of instant ROI is a software-era artifact. Will you have some easy-to-see, easy-to-measure gains? Yes, of course. But monetary wins are not the only reason you’re doing this. Your long-term goal is to survive and hopefully thrive in an AI-driven world, and that might not lead to a lot of feel-good metrics in the short term.
4. Stop relying on tech people to do this for you. Most leaders are scarily under-informed about what AI can do. I dazzled a group of them with one of my most basic agentic workflows. This is a problem – you can’t set a vision for something you know barely anything about and the cost of ignorance is high. Spend 30 minutes a week experimenting, attend events, hire expert consultants – just take personal responsibility for understanding the transformation that is already happening.
Here’s your homework
Put this prompt into the LLM of your choice and get started:
“I am the [JOB TITLE] of a [COMPANY SIZE] company in [INDUSTRY]. Help me draft an AI strategy that identifies opportunities for automation, potential competitive threats, and areas where we can use AI to create new value for customers. Include short-term (6–12 month) priorities and long-term (3–5 year) transformation goals.”
Horses to trains
You’ve probably heard people compare AI to Henry Ford’s famous quote - “If I had asked people what they wanted, they would have said faster horses.”
It’s a good comparison for this reason: Obviously trains sped up how quickly people could get to their destination, but they also changed far more than transportation. Towns no longer needed stables, which displaced stable builders. Supply chains for hay and carriage making became largely obsolete.
And they introduced a whole new industry: coal mining. People had to design new clothes for railroad workers versus horse riders, telegraph systems for communication between stations, and time zones for schedule coordination.
AI is this kind of change. It’s not SaaS on steroids, the same way a train is not a faster horse.
New industries will be built around this tech, the way we do business will fundamentally change, labor in general will look a lot different. And, like the businesses who continued to operate around horses surely failed in a train-powered world, the businesses who refuse to adapt to AI will fail too.
