Less T-shaped, more tree-shaped
As I have written about recently, the job market is in a state of significant change and uncertainty at the moment, and my personal circumstances are a reflection of that.
At the end of this week and after 7½ years, I will be leaving Deltatre for, well, nothing right now, and while that is scary and liberating in roughly equal measure I think it's also an opportunity.
One thing that I think is genuinely changing is the type — or shape — of individual that brings the most value, and I think (and hope) that this favours people like me.
The T-shaped individual has been a stalwart of tech teams for years (decades, even), but I think it's time to change the typographic analogy. It's time to be more tree-shaped.
A chance to begin again in a golden land of opportunity and adventure!
Things are changing rapidly at the moment, and what seems like the answer now might not be the answer in 6 months time.
I am fortunate in that I expect to be able to cover core household expenses for a decent period, so I actually have a strong opportunity to be highly flexible in a way that I could not really do with a 3-month notice period and restrictive non-compete clauses.
If an opportunity comes up tomorrow then I can take it, whether it's a fractional role, a short term contract, consulting, runway modelling, or a permanent role. I am currently planning on taking a deliberate 6-month sabbatical, but if the right thing comes up or I have a bright idea, then I can run with it.
In the meantime, I will be watching the wider market and tech ecosystem carefully.
(Un)employment in a changing market
I guess that in the spirit of writing openly, authentically, and in my own voice, I should probably start with the unvarnished truth.
The tech job market in 2026 is a shit-show - the traditional pyramid model is crumbling as companies are trying to do more with less, ostensibly looking for smaller, hyper-efficient teams.
What that actually means is that many companies are either using AI as an excuse to cut headcount, actively defer hiring, or are simply sitting on their hands until the wider situation becomes a little clearer and less hype-driven.
Many companies outside the tech community are also looking on in puzzlement, because AI is generating a lot of noise and fanfare, but isn't really touching their day-to-day reality.
My personal view is that those companies cutting heads and replacing them with raw AI output are creating incredibly brittle systems that are going to bite them on the arse down the line. They think they are building a "lean machine", but they're actually losing institutional memory and systemic judgement, clear-cutting their future capability for sustainable growth.
The team of the future was always the future
I don't think the answer has ever really been a team of pure specialists dragged forcibly down from their ivory towers, or a wide but shallow team of generalists with a sufficiently broad capability to be able to succeed, but rather a small enough group of multi-skilled, highly experienced practitioners who possess the deep structural "roots" — judgement, domain mastery, and collaborative capability — to deliver exceptional work.
It's basically the old-school "skunkworks" idea with a fleet of robot butlers1 to help out (albeit ironically with everything but the refreshments).
The best people I have worked with over the last 25 years have not been single-domain specialists - they have been those (apparently rare) multi-skilled individuals who come together with complementary and overlapping areas of expertise and knowledge.
AI makes those kinds of people way more powerful, and it also makes them precisely the kind of people that can safely anchor the transition to increasing use of AI tools.
Going organic
Maybe it's the impending change of employment status and the echoes of "garden leave"2, but I think we need to shift the entire conversation from a cold, mechanical team architecture to an organic, self-sustaining ecosystem, which is exactly what an engineering team needs to be whether they're on the AI bandwagon or not.
Whilst the Latin alphabet gave us the T-shaped individual, in Japanese and Chinese character typography, there is a typographic progression that's beautifully logical:
- 木 (Ki/Mù) = One tree — the multi-skilled, deeply rooted individual.
- 林 (Hayashi/Lín) = A grove — a small, tight-knit cluster of trees.
- 森 (Mori/Sēn) = A forest — the full, thriving ecosystem.
By using 木 → 林 → 森, we can bypass the rigid, artificial, geometric models like pyramids or diamonds and introduce something more organic and interesting for the AI era.
For me, it's an immediate response to the challenges and opportunities provided by AI. That includes the pragmatic need for fewer people to maintain the same productivity — which presents socioeconomic challenges of its own that I have written about previously — and a structural response to how we plan for the future.
The single tree (木): the multi-skilled individual
Instead of a single, brittle "T-spike," the modern practitioner is a 木.
A tree is a much more interesting and nuanced expression of individual capability than the latinate "T", which looks dull by comparison.
The roots provide deeply anchored human judgement, systemic thinking, and deeper context — the unseen and intangible things AI just can't simulate (yet) and which stop everything from falling over.
The trunk represents their core discipline and foundational principles, providing the structural core that grows slowly over time and provides the strength and resilience for everything built on top of it.
The branches represent multiple adjacent and overlapping skills and domains — data, product, AI, design, typography, orchestration, accessibility, interpretive dance, you name it — that reach outwards and grow in every direction over time.
The grove (林): the AI-native team unit
In the old world, you needed a big pyramid of specialised labour to build a system. In an AI world, a huge breadth of available information3 is more accessible and task execution is highly compressed so you don't need a crowd, you need a small team of highly capable, multi-skilled practitioners.
Every tree-shaped team member possesses significant individual capability that can be genuinely augmented by AI, so a pod of three or four 木-shaped practitioners can understand the bigger picture, dive in confidently, minimise costly context-switching, and rapidly deliver what used to require an entire department.
They stand close together, overlapping their branches, communicating clearly and deeply, and with significantly less bureaucratic friction than a traditional team model.
The forest (森): the sustainable organisation
This is where we solve the "pipeline problem" I raised when I talked about the changing structure of teams.
A pyramid is static — if you remove the bottom blocks, the top collapses — whereas a forest is dynamic and regenerative.
Stretching the metaphor a bit more, we can think of it as several complementary layers:
- The canopy — the senior practitioners providing coverage, strategic direction, and framing.
- The undergrowth — where new talent is nurtured. Instead of forcing juniors to do mind-numbing boilerplate work (the old entry-level execution), they are planted within the grove. They learn by observing how senior trees anchor themselves, process signals, and apply judgement to AI outputs.
The ecosystem is healthy because the small units are autonomous but share the same underlying structure and "nutrients" in the form of the organisation’s core culture, standards, and shared knowledge.
Cultivating for healthy growth over time
In the short term, leaders are playing a game of economic gravity. AI allows a smaller number of multi-skilled, high-capability individuals to collectively maintain — or even exceed — historical productivity.
The pragmatic reality for businesses is that this is an irresistible competitive pressure. If a tight-knit grove of four generalists can ship what used to require an eight-person pyramid, teams will shrink.
That has real impacts outside of individual corporate headcount decisions, and we have to call this what it is.
A widespread shift toward smaller, hyper-efficient teams means fewer total roles across the industry in the short term, until the wider economy catches up. This creates an immediate socioeconomic bottleneck — a tightening job market for mid-levels and a steep barrier to entry for anyone trying to break into tech. It’s an opportunity for the highly experienced (assuming they are not distressingly old like me), but a massive challenge for the broader workforce.
The longer term view is where the metaphor becomes essential, because you can't grow a sustainable future on a plot made barren by immediate pragmatism. If an organisation only optimises for the immediate horizon, they are essentially clear-cutting their own future.
When teams shrink to just a few senior generalists relying heavily on AI execution, the organisation becomes incredibly fragile. If those key individuals leave, they take the "roots" — the foundational context, systemic judgement, and institutional memory — with them. AI can reproduce the code they left behind, but it can't reproduce the why.
Planning for the future means realising that we can no longer rely on the old "onboarding by osmosis" model where juniors learn by churning through boilerplate tickets, and we have to actively plan for change as well.
Leaders have to intentionally design and tend the ecosystem to grow new talent, to adapt to a changing business climate, and to fill the inevitable gaps created by experienced talent moving on to new challenges.
That means shifting the junior role from a delivery mechanism to an apprenticeship of judgement — teaching them how to validate, critique, and orchestrate AI outputs rather than just writing syntax.
The immediate adoption of AI gives us smaller, faster teams today, but without sustained intent to cultivate the wider forest we are creating a vulnerable monoculture and borrowing capability from the future at a very high interest rate.