Vol. I MMXXVI
London Edinburgh
1706
Studio
arXiv:1706.03762
Vaswani et al.June 2017
The paper that opened the era
A house for the intelligence era

Most of what is written about artificial intelligence misses the point. The story is not the technology. The story is the people, and the work, and what survives the transition.

Essay № 01 Spring MMXXVI
Featured

Attention Is All You Need.

A field changed when eight researchers focused on what mattered. The same lesson now applies to organisations.

In June of 2017, eight researchers at Google published a paper that quietly changed everything. They argued the field had been building the wrong machinery for years. You did not need recurrent networks. You did not need convolutions. You needed attention. The architecture they proposed was simpler than what had come before, not more complex. The thesis was that focus on the right thing was more powerful than the elaboration of the wrong things.

Eight years later, every modern AI system descends from that paper. The transformer architecture they introduced underpins every model you have heard of. The lesson, however, is not really about machine learning. It is about how progress actually happens in any field. Most of the time, it happens by removing complexity rather than adding it. By attending to what matters and ignoring what does not.

We are at the same moment again, this time with organisations.

The intelligence era is arriving, faster than most leaders are prepared for. Every company we work with is being asked to deploy agents into their operations, redesign their workflows, rethink their headcount, restructure their org charts. Almost none of them know how. The big consultancies sell strategy decks that gather dust. The system integrators run two-year projects that miss the point by the time they ship. The AI-native firms speak in tokens and embeddings rather than outcomes. Everywhere, the noise is overwhelming. And in the noise, the actual question gets lost.

The actual question is not how to deploy more technology. It is how to focus an organisation on what genuinely matters, and to design the agents and the people and the work around that focus.

This is what attention means in the context of an organisation. Not the cognitive load on your inbox. Not the executive bandwidth lost to meetings. Attention is the firm's collective capacity to direct its energy at the things that produce real outcomes for the people it exists to serve. Most companies spend their attention badly. They optimise the visible at the expense of the meaningful. They mistake activity for progress. They confuse the instruments of work, the apps and the org charts and the quarterly metrics, with the work itself.

The agentic era will reward the organisations that can focus their attention on what matters. It will punish the ones that cannot. This is not a technology problem. It is a leadership problem, an organisational design problem, a culture problem. The technology is the smaller half of the work.

§ § §

We named this firm 1706 Studio because we believe the parallel is real. The paper that opened the AI era was not the most technically complex piece of research published that year. It was the simplest one with the right insight. The firms that win the next decade will not be the ones with the most technology. They will be the ones with the clearest attention.

Our work is to help leaders find that clarity. Not to sell them more software. Not to add to the noise. To help them see, with care, what genuinely matters in their organisation right now, and to design the people, the agents, and the work around that.

We believe the human is at the centre. Always. The agents serve the work. The work serves the people. The firm exists to design the relationships between all three.

This essay will be replaced quarterly with new thinking. The reading list grows over time. The annotated paper is below, for those who want to read what we read. None of this is service. It is an open invitation to think with us.

If any of it lands, you know how to find us.

Signed Craig Hepburn
On behalf of 1706 Studio
Frontispiece A timeline of attention

Two histories, converging.

Attention has been studied seriously for one hundred and thirty-five years. The field of artificial intelligence discovered it formally in 2017. The two lineages converge on the same question and answer it from opposite ends. Our work sits at the join.

THE COGNITIVE TRADITION philosophy, psychology, neuroscience William James Principles of Psychology 1890 Donald Broadbent filter theory of attention 1958 Daniel Kahneman Attention and Effort 1973 Iain McGilchrist The Master and His Emissary 2009 THE JOIN 17·06.2017 THE MACHINE TRADITION machine learning, agentic systems 2014 Bahdanau et al. attention in neural translation 2017 Vaswani et al. Attention Is All You Need Now the agentic era The same question asked from opposite ends. The same answer arriving in the same year.

One hundred and thirty-five years of attention The cognitive tradition asked how the human mind directs focus.
The machine tradition asked how a system might learn to focus its own.
Both arrived at the same answer in 2017.

§ 02 The method

The architecture of attention, applied to organisations.

/01i.
Attention Audit where leadership focus actually goes
Most companies spend their attention on the wrong things. The first job is to see clearly where it goes today and where it should go tomorrow.
/02ii.
The Encoder understanding the organisation as it is
We absorb the operating reality, encode it into a working model, and represent it back to leadership in a form they can act on.
/03iii.
Multi-Head parallel perspectives, integrated
Specialists working in parallel from different angles, operational, financial, cultural, technical, market, synthesised into one coherent view.
/04iv.
The Decoder strategy generated from understanding
Where understanding becomes action. The operating model, the deployment plan, the things that actually need to ship.
/05v.
Residual Connections preserving what already works
We do not tear down. The capability you already have skips past the bottleneck of redesign and continues forward, intact.
/06vi.
Positional Encoding context, timing, and place
Generic advice fails because it is not positioned to a specific moment. Every recommendation is encoded to where you actually are.
/07vii.
Layer Normalisation stable ground for deployment
Agentic systems land on stable ground or they do not land at all. We normalise the operational layer so the deployment is real, not performative.
/08viii.
The Transformer Block integrated, repeatable, alive
When all of the above are working together, the organisation has a complete unit that can repeatedly transform inputs into outputs. That is the work.
§ 03 The annotated paper

Reading Attention Is All You Need,
through the lens of organisations.

The paper that opened the AI era was a research artefact. We have spent considerable time reading it as something else: a guide to how attention, hierarchy, and parallel processing might be redesigned in the organisations of the next decade. What follows is a small selection of passages from the original paper, with our commentary.

From the Abstract "The dominant sequence transduction models are based on complex recurrent or convolutional neural networks. We propose a new simple network architecture, the Transformer, based solely on attention mechanisms, dispensing with recurrence and convolutions entirely."
Our reading

The first sentence of the paper is a diagnosis. The dominant approach is too complex. The second sentence is a thesis. Simpler architecture, focused on the one thing that mattered (attention), would outperform. Almost every successful organisational redesign we have ever seen follows this structure. Diagnose where the complexity is unnecessary, identify the one mechanism that matters, build around that. Most companies inherit elaborate machinery from a previous era and never ask whether the machinery is still doing the work.

Section 3.2.1 "An attention function can be described as mapping a query and a set of key-value pairs to an output, where the query, keys, values, and output are all vectors. The output is computed as a weighted sum of the values, where the weight assigned to each value is computed by a compatibility function of the query with the corresponding key."
Our reading

This is the technical heart of the paper, and it has a remarkable organisational analogue. An organisation is also a function that maps queries to outputs by weighting which information sources are most compatible with the question being asked. A good organisation routes the right query to the right person with the right context, and weights their input correctly. A bad organisation routes everything to everyone and weights nothing. The transformer's contribution was to make this routing explicit, parallel, and differentiable. The same operations could redesign how decisions get made inside companies.

Section 3.2.2 — Multi-Head Attention "Instead of performing a single attention function with d_model-dimensional keys, values and queries, we found it beneficial to linearly project the queries, keys and values h times with different, learned linear projections."
Our reading

Multi-head attention is the discovery that one perspective is not enough. The model gets better when several different attention mechanisms run in parallel, each looking at the same data through a different learned lens, and the results are concatenated. This is exactly how good consulting works, when it works. A senior team brings several perspectives to the same problem in parallel, integrates them, and produces a synthesis that no single perspective could have reached. Most organisations operate as single-head systems. They have one dominant lens, usually financial, and miss everything the other lenses would have caught.

Section 3.4 — Why Attention "A single attention layer connects all positions with a constant number of sequentially executed operations, whereas a recurrent layer requires O(n) sequential operations. Self-attention layers are faster than recurrent layers when the sequence length is smaller than the representation dimensionality."
Our reading

Recurrent networks process information one step at a time, like a hierarchy passing memos up and down. Self-attention sees the whole context at once. This is the deepest structural insight of the paper for organisational work. Most companies are built on recurrent communication. Information goes up the hierarchy, gets compressed, comes back down, gets misinterpreted. The companies that will thrive in the agentic era will be the ones that flatten this. Where context is shared, where everyone sees the whole sequence, where decisions can be made closer to where the work happens.

§ 04 Notes from the field

A small record of recent work,
told without naming names.

We do not publish case studies. The work belongs to the people who commissioned it. What follows are short, anonymised notes on the kinds of questions we have been asked, and the angles we have taken on them.

Note № 01
Spring MMXXVI

Codifying the operating model of a venture firm into agentic infrastructure.

A venture studio asked us to help redesign itself from the inside out, around the assumption that agents would be first-class participants in its operations from day one. We did not build them an AI tool. We did something more difficult and more useful: we codified the operating model of the firm into a structure that an agent could read, reason about, and act on, with humans coordinating the relationships and judgement that agents cannot. The result is a chief of staff agent that holds context across the entire firm, coordinates project teams, and removes a category of friction the partners had been carrying for years.

Operating model · Chief of staff · Friction removal
Note № 02
Spring MMXXVI

An executive-tier agent for a founder operating across multiple ventures.

A founder running several companies in parallel asked us to design what an executive-tier agent might look like for someone in their position. The technical challenge was straightforward. The harder question was philosophical: which decisions should remain entirely human, which should be agent-assisted, and which could be agent-led with human approval? The work produced a working agent and, more importantly, a written constitution for how the founder would relate to it. The relationship turned out to matter more than the technology.

Founder operations · Agent constitution · Human–agent relationship
Note № 03
Spring MMXXVI

Building agent infrastructure for a non-profit helping ordinary people use AI well.

A US-based organisation working to help ordinary people get genuine value from AI, and to defend them against its harms, asked us to build the agentic backbone they would need. The work matters because most AI infrastructure is being built for enterprises and the wealthy. Almost no one is building thoughtful tools for the people who will be most affected by the transition and least equipped to navigate it. The brief was technical. The motivation was civic. We took it because the question of who AI ends up serving is the most important question in the field.

Civic infrastructure · Defensive AI · The other 95 per cent
Note № 04
Spring MMXXVI

What we are learning, written down before we forget.

A note about the work itself rather than a specific engagement. We have been struck, across every project this season, by how little of the difficulty is technical. The hard part is almost always organisational, almost always about people, almost always about questions of authority and judgement that have been carried unexamined for years. The agentic transition is forcing companies to be honest about how they actually work. The companies willing to look at this clearly are finding the deployments easier than they expected. The ones still arguing about model selection are finding the deployments impossible. We are taking notes.

Methodology · Field observations · The honest work
§ 05 The reading list

What we read, and why we read it.

A short list, updated quarterly. Not a syllabus on artificial intelligence. A syllabus on people, work, attention, and the longer arc of how technology either lifts or diminishes us.

2009 Iain McGilchrist
The Master and His Emissary

Two ways the brain attends to the world. The mode you choose determines what you can see. Foundational text on why attention is the substrate of every other capacity.

1958 Hannah Arendt
The Human Condition

The distinction between labour, work, and action. Reads as if it were written for the agentic era. The most important book about what humans are for that we know of.

2002 Carlota Perez
Technological Revolutions and Financial Capital

The pattern that connects every technology revolution to the social one that follows. Helps locate where we actually are in the AI cycle.

1973 Peter Drucker
Management: Tasks, Responsibilities, Practices

The original text on knowledge work and what it asks of organisations. Drucker saw most of this fifty years ago and most companies still have not caught up.

2017 Vaswani et al.
Attention Is All You Need

The paper the firm is named for. Eight authors, eight pages, the architecture that opened the modern era of artificial intelligence. We annotate it above.

1972 Stafford Beer
Brain of the Firm

The cybernetic view of organisations. How information flows, how feedback loops actually work, how a company is more like a living system than a machine.

1999 Stewart Brand
The Clock of the Long Now

Thinking in centuries rather than quarters. The intellectual posture every senior leader needs and almost none cultivate.

2018 James Bridle
New Dark Age

The cost of mistaking computation for understanding. A counterweight to the techno-optimism that dominates AI discourse.

1964 Marshall McLuhan
Understanding Media

The medium is the message. Every new technology rewrites the human relationships around it. AI is the most consequential medium since print, and we are not yet asking the right questions about it.

2023 Karen Hao
Empire of AI

The political economy of the AI industry, told without flattery. Required reading for anyone deploying these systems into their organisation.

§ 06 The position

Human first. Agent augmented.
Always in that order.

The future of work is not human against agent. It is not human replaced by agent. It is the human at the centre, augmented by a network of agents that handle the work below the threshold of judgement.

What we believe

Judgement is the asset.

The skills that survive automation are the ones that require relationship, context, and accountability. The firm builds organisations around those skills, not against them.

What we reject

Headcount as a metric of seriousness.

The companies that win the next decade are not the ones with the largest headcount. They are the ones whose people do work agents cannot, supported by agents that do work people should not have to.

How we work

Strategy and delivery, in the same room.

Senior advisory at the top, operational delivery underneath, with the methodology that connects them owned by the firm. No handoffs. No translation gaps. The strategy ships.

Why it matters

Most agentic deployments fail.

They fail because the technical work and the organisational work happen separately. The model is right, the org is wrong. We work both layers as one.

§ 07 The partnership

Built by people who have done this before,
and people who are doing it next.

Craig Hepburn

CEOFounding Partner

Twenty years building at the frontier of digital. Microsoft, Nokia, UEFA, Art Basel. Perplexity Fellow. Publishes Ground Truth on the orchestration layer thesis. Brings the strategy, the technical depth, and the senior advisory practice.

Fergus Lynch

COOFounding Partner

Operational leadership at scale across Microsoft, Nokia, and Minecraft Studios. Brings the delivery model, the financial discipline, the network of builders, and the business framing that keeps the firm focused on outcomes the buyer can act on.

The next generation

Connor Hepburn

Next Gen PartnerAgent Infrastructure

Builds the agents the firm runs on. Connor is part of the first generation of operators native to these tools, and the fluency shows in the work. Has built executive-tier agents, prototyped chief of staff systems, and brings the engineering instincts of someone who learned to code by needing to. The firm exists partly to demonstrate what happens when builders of his generation are given real responsibility from the start.

An open invitation

Most things do not need more attention. The right things do.

We are not selling a service. We are inviting a conversation. If anything you have read here has landed, write.

hello@1706.ai