It's January 2023 and something has shifted. Not just in technology, but in the conversation. Every executive I've spoken to in the last six weeks has asked the same question, in slightly different words: "What does this mean for us?"
What You Need to Know
- ChatGPT crossed 100 million users faster than any consumer product in history. That's not hype - it's signal. The demand for intelligent software was always there. Now there's a visible interface for it.
- 2023 will be the year enterprises start taking AI seriously. Not because the technology is new (it isn't), but because it's now visible enough that boards can't ignore it.
- The biggest risk this year isn't moving too slow. It's moving too fast, without clarity on what problem you're solving.
- Enterprises that invest in data readiness and AI literacy now will have a 12-18 month advantage over those that wait for the technology to "mature."
100M
users reached by ChatGPT in under two months - the fastest consumer adoption in history
Source: UBS Global Research, January 2023
The Shift
I've been building digital products for 15 years. I've watched the internet go mainstream. I was there for mobile, cloud, APIs, microservices. Each of those waves had a moment where the conversation changed from "is this real?" to "what do we do about it?"
For AI, that moment was November 2022.
ChatGPT didn't invent anything new. Large language models have existed for years. GPT-3 was available as an API since mid-2020. Researchers have been publishing about transformer architectures since 2017. But none of that mattered to enterprise leaders until there was something they could try themselves, in a browser, for free.
Now they've tried it. And the questions are coming fast.
What I Think Happens This Year
Q1-Q2: The scramble. Boards will ask for AI strategies. Vendors will pivot to "AI-powered" everything. Consultancies will publish frameworks. Most of it will be reactive and surface-level. Enterprises that already have strong data practices and clear operational challenges will be positioned to move quickly. Everyone else will be writing slide decks.
Q3-Q4: The reality check. Early experiments will show that enterprise AI is harder than consumer AI. Data quality, integration complexity, governance requirements, and change management will slow things down. The gap between "impressive demo" and "production system" will become painfully clear.
By year end: A small number of enterprises will have meaningful AI capabilities in production. A much larger number will have learned expensive lessons about what it takes. Both groups will be better positioned for 2024 than the enterprises that sat this year out.
What This Means for NZ
New Zealand has specific advantages and disadvantages in this shift.
Advantages: Small organisations can move faster. We have strong domain expertise in sectors like agriculture, healthcare, and financial services. Our regulatory environment is more adaptable than Europe or the US. And we have a history of punching above our weight in technology adoption.
Disadvantages: Small talent pool. Limited AI research infrastructure. Distance from the centres of AI development. And a tendency to wait and see rather than move early.
My view is that the advantage goes to organisations that start now - not with AI implementations, but with the groundwork. Clean your data. Map your knowledge. Identify where decisions get bottlenecked. Understand where your best people spend time on work that doesn't need their expertise.
That's the foundation. The AI tools will keep improving. What won't improve on its own is your organisation's readiness to use them.
A Note on the Hype
I'm not naive about hype cycles. I watched blockchain go from "this changes everything" to "well, maybe not everything" in about 18 months. Some of the energy around AI right now has that same breathless quality.
But here's the difference: I spent three hours talking to ChatGPT, and it fundamentally changed how I think about software. That didn't happen with blockchain. That didn't happen with VR. That didn't happen with any previous technology trend.
This one is different. Not because the hype says so, but because the technology actually works in a way that previous "next big things" didn't.
The question isn't whether AI will transform enterprise operations. It's whether you'll be ready when it does.
2023 is the year to get ready.