In Episode 004 of The Next Thing Now, Rob Borley and Kev Smith dive deep into what they describe as a turning point in enterprise AI. The conversation ranges from technical protocol standards to culture change and philosophical shifts in how we build and adopt intelligent systems. Their message is loud and clear: AI is not a destination—it’s infrastructure, and organisations that treat it as a bolt-on strategy are missing the point.
Kev opens by highlighting the flurry of new standards emerging around agentic software systems. Three acronyms dominate the discussion:
MCP (Model Capability Protocol) – A way for AI agents to gain new powers by dynamically discovering tools they can use.
A2A (Agent-to-Agent protocol) – A new standard proposed by Google for how AI agents introduce themselves and communicate across systems—akin to HTTP for agents.
LoRA (Low-Rank Adaptation) – A technique that allows fine-tuning of base models through lightweight “adapter” layers, enabling cost-effective, specialised AI behaviour without retraining entire models.
These standards point to a new AI-native computing paradigm that isn’t just smarter—it’s modular, interoperable, and built to evolve fast.
The hosts stress that these developments don’t represent simple improvements to the old way of doing things. This is not like adopting a new IDE or switching from on-prem to cloud hosting. It’s a wholesale rethinking of how software is designed, built, and operates.
Kev notes that many current criticisms of emerging tools (like “it’s not production-ready” or “security isn’t built-in yet”) miss the point. We're still in the exploratory phase—like cloud computing in its infancy. Now is the time to learn, not to wait.
Rob describes conversations with CxOs who are stuck between two conflicting impulses: the desire to draft AI strategies carefully, and the fear that they’ll be left behind. He compares the AI transition to foundational shifts like the internet, electricity, or mobile—not something that requires an ROI argument, but something you can’t function without.
“AI isn’t a product to buy when it’s ready—it’s a capability to build now, before you’re too far downstream to catch up.”
Kev likens the moment to the rise of e-commerce. Initially, it seemed fringe. Years later, it became infrastructure. The same will happen with AI—but the curve is steeper, and the risks of waiting are higher.
Too many leaders still believe they’ll buy “AI” off the shelf like a packaged product. Rob argues this thinking is dangerous. AI is not a COTS (commercial off-the-shelf) solution—it’s an evolving capability that requires:
New behaviours
New skills
New organisational muscle memory
In this view, we’re all back in school again—whether individuals or institutions—and the goal is to learn to function in an AI-first, agentic world.
Kev notes a major problem: we don’t have shared language yet. Terms like “agent” are being applied to everything from simple automation flows to fully autonomous systems with memory, preferences, and reasoning.
He calls for more precision—and shares how his own team has begun naming agentic system patterns internally (like “market makers”) to help describe and design complex systems clearly.
“It’s like the early days of cloud—we didn’t know what a ‘Lambda’ or ‘Blob Storage’ was until we had to name them.”
Rob delivers a powerful metaphor: talking about your “AI strategy” is like having an “electricity strategy.” It’s not a strategy—it’s a utility. A foundational enabler. Instead of asking, “what’s our AI strategy?” leaders should ask:
What could we do if our teams had superpowers?
How would our work change if the boring stuff were automated?
What happens if we don’t act?
AI isn’t the end goal—it’s the fuel.
Kev draws a parallel to the cloud migration era, where many companies clung to CapEx mindsets and resisted the shift to elastic, on-demand infrastructure. Those who delayed were disrupted. The same is happening now, but faster.
The problem isn’t just tech—it’s people and processes. Behavioural change, budget realignment, security models, and even job definitions all need to evolve.
The episode concludes with a discussion of the now-viral Shopify AI memo, in which the CEO urged employees to embrace their “AI goodie bags” or risk being left behind. While the tone sparked backlash, Rob and Kev see the urgency as justified—if poorly communicated.
Kev notes that executives and employees alike are afraid. Leaders are unsure how to implement AI; staff don’t know how to use the tools meaningfully. Without a deliberate culture of upskilling and experimentation, good intentions will fail.
“You can’t just drop the goodie bag and walk away. You have to take people on the journey.”
Episode 004 makes the case that we’re in a pivotal moment. The tools are changing. The methods are new. The language isn’t settled. But the direction is clear: AI isn’t a strategy—it’s a utility. The real transformation lies in what people and organisations do with it.
Start learning. Start building. Start adapting. Because the pace isn’t slowing—and the cost of inaction may be irrelevance.