02/01/2026
Don't be a Dinosaur!
We recently attended Bigdata London, where a session by Data Puppets really stuck with us, not because it was flashy, but because it offered a clear mental model for where AI in business is actually heading.
They framed AI’s evolution in phases, which helped cut through a lot of the noise:
1️⃣ Generative AI
AI assists humans in generating content, code, insights.
👉 We are already here. Productivity gains are real, but humans still decide and act.
2️⃣ Agentic AI
AI starts to act, not just respond like planning steps, using tools, executing workflows under constraints.
👉 This is where we are entering now: early, narrow, supervised, but meaningful.
3️⃣ Agentic + Physical AI
AI decisions extend into the physical world within logistics, factories, supply chains, robotics.
👉 This will roll out gradually in controlled, high-ROI environments.
4️⃣ Autonomous Organisations (AOs)
AI runs entire business functions end-to-end, with humans shifting from operators to governors.
👉 Directionally important, but still aspirational at scale.
What struck me most is that technology isn’t the main bottleneck anymore.
The real challenges are:
Trust
Governance
Data quality
Organisational design
But most importantly, "Why this matters for 2026?"
2026 won’t be about “who uses AI” but it will be about who redesigned their processes, roles, and decision-making around AI.
The winners won’t just add copilots. They’ll build human–AI systems where AI operates & Humans supervise, set boundaries, and own outcomes
In short:
AI is moving from assisting people to acting for people.
And organisations that don’t adapt their structure.. "not just their tools" will fall behind.
