In the past month, I’ve heard stories from a founder aggressively trying to triple revenue without hiring more reps, a CRO watching one automation-savvy solutions engineer quietly fix months of broken GTM processes in days, and a COO at a national retailer looking to deliver faster, more accurate customer support without tying up a human for every minor issue.
Everyone’s asking the same question: Can AI do that for us? But the real question is, how do we actually make AI useful for our business? Not the flashy demos or the hype-filled press releases. The real stuff like workflows, decisions, delivery. Things that impact the bottom line in your business.
The themes are consistent across industries and org charts: AI can reduce costs, accelerate delivery, and in the right hands shift the trajectory of a business almost overnight. That kind of upside doesn’t go unnoticed. Which is why everyone’s racing to figure it out.
"The future is already here, it's just not very evenly distributed." – William Gibson
When it comes to AI, you’d be hard-pressed to find a more accurate description of where we’re at. It’s already here, and it’s already unevenly distributed. There won’t be a decades-long digital transformation window this time. It’s winner-take-all, and the sprint has already started. Some businesses have a mile head start — and with AI, that head start compounds fast.
Here are three trends I’m seeing shape this adoption race.
1. AI Agents Will "Eat the World"
Building on Marc Andreessen’s thesis that "software is eating the world", we now face the next inflection point: AI Agents. These autonomous systems don’t just run software; they think, decide, learn, and act on our behalf. In short, they run software.
AI agents are systems that combine planning, memory, and tool use to perform tasks autonomously. Unlike traditional automation, which follows deterministic, pre-programmed instructions, agents can interpret goals, decide what to do next, and adapt in real time based on new information. They are not just running steps. They are managing workflows with intent.
Model Context Protocol (MCP) functions like a universal connector for agents. It helps systems work together by standardizing how agents access tools, memory, APIs, and structured context. Think of it as the USB-C equivalent for AI, making it easier to plug into different components with minimal effort.
This is not theoretical. Companies are already putting agents to work. At Zapier, teams are using agents to handle customer service, sales meeting preparation, and engineering support. These agents pull from knowledge bases, route tickets, enrich CRM records, and trigger follow-up actions without needing a person to guide each step. What used to be built with step-by-step workflows is now powered by systems that can reason, learn, and act based on broader goals.
Agents are not just an evolution of software. They represent a new layer of intelligence that is already reshaping how work gets done.
2. SMEs: A Trillion-Dollar Opportunity in Efficiency
For small and mid-sized businesses, this isn’t sci-fi, it’s a competitive edge.
McKinsey and PCG research shows that SMEs report efficiency gains of up to 33% when adopting AI tools. In product development alone, AI-driven projects outperform others by 27% on key KPIs.
And momentum is growing:
57% of SMEs already use AI, according to Forbes and PCG surveys.
90% cite efficiency as the leading benefit.
Here’s how to get started today:
Deploy general-purpose AI tools like ChatGPT, Claude, or Gemini. These are the entry points.
Orchestrate simple automations using platforms like Zapier. Link AI to tools like Slack, Google Sheets, or your CRM.
Scale into agents: evolve from single prompts to workflows that handle multi-step reasoning, action, and follow-up.
This is a playbook for immediate wins. Start small, iterate fast, expand when you see real impact.
3. Enterprise-Grade AI Orchestration: What Matters
Enterprise adoption doesn’t mean more tools. It means orchestration at scale, with security, and with people in mind.
Based on reports from McKinsey, Deloitte, and OpenAI, here’s what the most advanced organizations care about:
Reliability at scale: Enterprise AI efforts are now governed like core infrastructure. AI must be secure, explainable, and observable.
Ecosystem thinking: Orchestration layers are being built to handle everything from data routing to human-in-the-loop approvals.
Interoperability: Standards like MCP allow agents to interact across tools like Slack, GitHub, and Postgres without friction.
AI Fluency across teams: McKinsey reports that 81% of employees are ready to use AI, but many orgs are behind in training and governance.
What we’re seeing early on:
Enterprises are building "AI fabric" or managed layers that orchestrate tools, context, and security.
They’re embedding safety-first principles: reliability, auditing, and limiting actions are table stakes.
They’re investing in literacy: Not just launching AI, but upskilling employees, reworking processes, and measuring what matters.
Early conclusions
AI agents are positioned to eclipse software as we know it. Their adoption will be as transformational, and as inevitable, as the move from desktop apps to cloud.
SMEs that move now are in the best position to win. The tools are affordable, available, and effective. The path is clear: start simple, automate fast, and learn by doing.
Enterprises have a more complex challenge. Not just adoption, but orchestration. The ones who build structured, scalable AI fabrics and invest in team fluency will have a long-term moat.
I’m watching these shifts happen every day in the field. Some companies are sprinting, others are standing still. Regardless, the race is on.
If you’re not experimenting with agents or automation yet, now’s the time. The head start won’t last forever. It's only a matter of far less time than you think.