Vol. I · MMXXVILondon · Glasgow

A house for the intelligence era

Intelligence is getting cheap. Attention isn’t.

The intelligence era will be won by the organisations that know what deserves their attention. We help leadership teams decide where it should go, then build the people, the agents and the work around it. Working systems, in days. Not decks.

Run the Attention Audit Read the thesis
§ 01The shift

What gets cheaper, and what doesn’t.

Every firm now has the same models. A competent first draft, a passable analysis, code that runs: each costs less every quarter. What doesn’t get cheaper is knowing what matters, making a call you can stand behind, and proving the thing works. That’s where the advantage moves.

Getting cheaperCost

  1. A first draft
  2. A competent analysis
  3. Code that runs
  4. A polished deck
  5. An answer to almost anything

Getting scarcerValue

  1. Knowing what matters
  2. Judgement you can stand behind
  3. Proof that it works
  4. Trust between people
  5. A team’s focus

Illustrative, not measured. The direction is the point.

arXiv:1706.03762
Vaswani et al. · June 2017
The paper that opened the era
.

In June 2017, eight researchers at Google published a paper that changed the field.1 They argued it had been building the wrong machinery for years. You didn’t need recurrent networks. You didn’t need convolutions. You needed attention. The architecture they proposed was simpler than what had come before, not more complex. Focus on the right thing beat the elaboration of the wrong things.

Nine years on, every modern AI system descends from that paper.2 The lesson isn’t about machine learning, though. It’s about how progress happens in any field. Most of the time it comes from removing complexity rather than adding it. From attending to what matters and ignoring what doesn’t.

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 is being asked to put agents into its operations, redesign its workflows and rethink the organisation itself. Almost none of them know where to begin, and in the noise the question that matters gets lost.

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

That is what attention means for an organisation. Not the load on your inbox. 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 it badly. They optimise the visible at the expense of the meaningful, and mistake activity for progress.

The agentic era will reward the organisations that can focus their attention. It will punish the ones that can’t. This isn’t a technology problem. It’s a leadership problem, an organisational design problem, a culture problem. The technology is the smaller half of the work.

We named the studio 1706 because we think the parallel holds.3 The paper that opened the AI era wasn’t the most technically complex research published that year. It was the simplest one with the right insight. The firms that win the next decade won’t be the ones with the most technology. They’ll be the ones with the clearest attention.

Our work is to help leaders find that clarity, and then to prove it. We think like advisors and prove it like builders. The strategy doesn’t arrive as a deck. It arrives as a working system, in your organisation, in days.

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.

The deeper material (the annotated paper, the timeline of attention, the reading list) lives in the library, open to everyone. It’s an invitation to think with us. If any of it rings true, you know where to find us.

SignedCraig Hepburn
On behalf of1706 Studio

1Vaswani, Shazeer, Parmar, Uszkoreit, Jones, Gomez, Kaiser and Polosukhin. Attention Is All You Need. The title is the thesis.

2The architecture is the transformer. It is the T in GPT.

3arXiv numbers papers by year and month. 1706 is June 2017. Hence the name, and the full stop.

§ 03The offer

Diagnose. Build. Run. The strategy ships.

Three stages, one team, no handoffs. At every stage there’s something real to look at rather than something to imagine.

i. / 01

Diagnose

the attention audit

A short, senior engagement inside the operating reality. Where attention goes today, where it should go next, and what agents should do first. A first pass runs on this site, in three minutes.

What exists afterwardsA map leadership can act on.
ii. / 02

Build

working systems, in days

We prove the strategy by building it, with the people who’ll run it. Nothing arrives as a recommendation that could have arrived as a demonstration.

What exists afterwardsA working system inside the organisation.
iii. / 03

Run

owned by your people

We stay while the organisation learns to run what we built together, then hand it over properly. Done with people, not to them.

What exists afterwardsA capability the team owns.
Instrument · Diagnose

The Attention Audit →

Ten questions, three minutes, a reading of where your organisation’s attention goes.

 Instrument · Reference

Agent Network ↗

Our public guide to the agent landscape. See how we think before you write to us.

 Instrument · Film

A new way to work →

A short film on our changing relationship with technology, with a starter instruction to try.
§ 04The position

Human first. Agent augmented. Always in that order.

Not human against agent, and not human replaced by agent. The human at the centre, with a network of agents handling the work below the threshold of judgement.

01
What we believe

Judgement is the asset.

The skills that survive automation need relationship, context and accountability. We build organisations around those skills, not against them.

02
What we reject

Headcount as a measure of seriousness.

The firms that win the next decade won’t be the largest. Their people will do work agents can’t, supported by agents doing work people shouldn’t have to.

03
How we work

Strategy and delivery in the same room.

Senior advisory at the top, delivery underneath, one method connecting them. No handoffs. No translation gaps.

04
Why it matters

Most agent deployments fail for organisational reasons.

The technical work and the organisational work happen separately. The model is right and the organisation is wrong. We work both layers as one.

§ 05Notes from the field

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

We don’t publish case studies. The work belongs to the people who commissioned it. These are short, anonymised notes on the questions we’ve been asked and the angles we took.

Note № 01

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

Operating model · Chief of staff · Friction removal
A venture studio asked us to redesign it from the inside out, on the assumption that agents would take part in its operations from day one. We didn’t build them an AI tool. We did something harder and more useful: we codified the operating model of the firm into a structure an agent could read, reason about and act on, with people holding the relationships and judgement agents can’t. The result is a chief of staff agent that holds context across the firm, coordinates project teams, and removes a category of friction the partners had carried for years.
Note № 02

An executive agent for a founder running several ventures at once.

Founder operations · Agent constitution · The human and agent relationship
A founder running several companies in parallel asked what an executive agent might look like for someone in their position. The technical challenge was straightforward. The harder question was philosophical: which decisions stay entirely human, which are agent assisted, and which can 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.
Note № 03

Agent infrastructure for a non profit helping ordinary people use AI well.

Civic infrastructure · Defensive AI · The other 95 per cent
A US organisation helping ordinary people get real value from AI, and defending them against its harms, asked us to build the agentic backbone they needed. Most AI infrastructure is being built for enterprises and the wealthy. Almost no one is building thoughtful tools for the people most affected by the transition and least equipped to navigate it. The brief was technical. The motivation was civic. We took it because who AI ends up serving is the most important question in the field.
Note № 04

What we’re learning, written down before we forget.

Method · Field observations · The honest work
Across every project this season, very little of the difficulty has been technical. The hard part is almost always organisational: people, authority and judgement that have gone unexamined for years. The agentic transition is forcing companies to be honest about how they work. The ones willing to look clearly are finding deployment easier than they expected. The ones still arguing about model selection are finding it impossible. We’re taking notes.
§ 06The writing

Ground Truth. Thinking in public.

Craig publishes Ground Truth, essays on the intelligence era and what we choose to do with it. The writing does in public what the studio does in private: it pays attention.

Four places to start

Craig also speaks on the orchestration layer and the future of work. Invitations to hello@1706.ai.

The library · open to everyone

The thinking is given away.

The deeper rooms of the house are open. Take what’s useful. Nothing is behind a door.

  1. i.The annotated paper
  2. ii.The method in full
  3. iii.The timeline of attention
  4. iv.The reading list
§ 07The partnership

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

Craig Hepburn

CEO · Founding Partner

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

Fergus Lynch

COO · Founding 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 work on outcomes a leadership team can act on.

Connor Hepburn

Next Gen Partner · Agent Infrastructure

Builds the agents the firm runs on. Part of the first generation of operators native to these tools. Has built executive agents and prototyped chief of staff systems, with the engineering instincts of someone who learned to code by needing to.

An open invitation

Most things don’t need more attention. The right things do.

We’re not selling a service. We’re inviting a conversation. Run the audit, see your organisation the way we would, and if something rings true, write.

Run the Attention Audit hello@1706.ai Agent Network