Made From You

A job can lose its market before a person runs out of things worth knowing. That is the problem I want Songbird to answer.

The answer has to be more substantial than reassurance about AI. We want to build companies in which a person's experience changes the service, their judgement reaches more customers, and the surrounding work no longer consumes their entire day. Here is the promise on one page.

AI will not choose a humane outcome for us

AI makes it possible to perform some work with less paid human effort. If a business measures success only by reducing that effort, the person doing the work becomes a cost to remove. Calling the resulting product helpful does not answer what happens to the worker's independence, relationships or sense of usefulness.

That is an economic design problem. It is not a claim that every job will vanish. The ILO's 2025 occupational study explicitly distinguishes exposure from disappearance and judges job transformation the likeliest overall effect.¹ A serious response can recognise both the threat to particular livelihoods and the uncertainty about the wider transition.

Songbird's response is to explore a different destination for the person's contribution: a company. Their worth as a human does not depend on founding one. But for someone who wants to keep helping a market they understand, the choice need not begin with learning to perform every business function.

What does the person change?

An automated business built around you can be useful. It can prepare messages, organise appointments and make administration easier. Those conveniences do not establish that the business is made from you.

Made from you means the company's decisions depend on what you bring. It asks different questions because you know what gets missed. It declines an apparently attractive request because your experience reveals a problem. It explains a choice in a way that reflects the people you have learned to serve.

If changing the founder changes only the name on the website, we have missed the point.

The practical work is to turn relevant knowledge into company capability: examples, distinctions, limits and reasons that can guide future work. A large archive alone achieves nothing. We have to show the knowledge changing an actual decision.

There is a bounded precedent for that possibility. Brynjolfsson, Li and Raymond's 2023 customer-support study reported suggestive evidence that an AI assistant spread practices from more able workers to newer colleagues.² It does not prove that a person's entire expertise can be captured or that a company can run itself. It does give us a reason to test whether knowledge can travel further than its original holder's working hours.

A company that can change with you

My co-founder has helped establish a crucial priority: we need to understand the person's circumstances before deciding what business form to ask them to inhabit. Someone might want to spend time with customers and have no desire to manage staff. The opportunity could still fit if the operating responsibilities are credibly covered elsewhere.

We call this company morphing. It changes the role and, where necessary, the company. The person's correction is consequential. If we learn that our proposal would recreate the work they are trying to leave, we must change the proposal.

A prepared company seed can help. It supplies a starting possibility rather than demanding that everyone invent from a blank page. The plane-and-pilot analogy explains preparation, but has a limit: a plane generally remains the same whichever pilot enters it. Our claim is that the person's knowledge should change the company itself. A seed is the chassis; the person helps determine what it becomes.

For the personal-company offer described here, our human champion gate requires an accountable person before release. My co-founder helped formulate that gate. A founder can begin with a particular customer or domain contribution; the title need not imply that they already know how to be a chief executive. Necessary decisions still need capable coverage. Preference cannot make responsibility disappear.

An experienced nurse, considered properly

Illustration, not a case study. Imagine a nurse with twenty years in aged care who wants a different role. We are not assuming that nursing has been automated away. She may simply want to use her experience differently.

A plausible possibility is a non-clinical advocacy service for older people living at home. Begin with what she can notice: when a family's proposed arrangement overlooks the older person's preferences, when nobody has confirmed an agreed next step, or when a reassuring summary conceals an unanswered question.

Now make that knowledge specific. She explains why a neat list of available services is not yet a workable arrangement. Who is coordinating it? What has the older person agreed to? Which fact is current, and which was assumed? The company can use those distinctions when preparing material and following up agreed work.

Suppose a daughter returns after circumstances change. The company would retrieve the authorised history, identify what needs checking and prepare the unresolved questions for the founder. The nurse would spend her attention on interpretation and the human conversation. Supporting workers would organise the next agreed actions and keep their status clear.

The customer benefit is concrete: the family would not need to reconstruct everything, and an incomplete arrangement would not be mistaken for a finished one. The founder benefit follows only if the preparation and follow-through actually reduce her burden. If she must repair every summary, the illustration has failed its own promise.

The company would also need to recognise the limits of that role. It cannot turn advocacy into unqualified clinical decision-making. Nor can the founder's professional experience justify deciding what an older person ought to want. The customer gets to correct the picture too.

A beginner needs a different promise

A second illustration. A recent graduate enjoys organising small local events and talking with community groups. Their raw material is interest, initiative and some experience of discovering what organisers need. It is not a career's worth of domain judgement.

A prepared seed for helping local organisers assemble event information might be a suitable place to learn. Start with a narrow service: establish what the organiser needs to communicate, check supplied details, prepare an agreed information pack and keep revisions straight. Do not dress the graduate up as an expert in event safety or other work they cannot judge.

The graduate might notice that first-time attendees repeatedly struggle to find the right entrance. That observation could change the company's preparation checklist and the information it asks an organiser to verify. The next customer would receive a better arrival explanation. AI could prepare versions and manage corrections, leaving the graduate time to talk with organisers and discover the next practical obstacle.

Here the chain begins with a small, demonstrable contribution and grows through learning. The quality of the seed and access to competent support matter more. If the graduate cannot judge a material part of the offer and nobody else can cover it, the scope must shrink. Enthusiasm does not fill a competence gap.

Why experience might remain valuable

As capable models become accessible to competing firms, model access alone tells a customer less about why they should choose one. Our durability thesis rests on relevant judgement, relationships and a record of reliable service. Historical entrepreneurship research using US Census data also finds that experience in the venture's industry predicts stronger outcomes.³ That association supplies no success odds for either illustration.

Knowledge is not magically protected from competition. It can be outdated, shared by others, or irrelevant to what customers will pay for. The value lies in applying it, learning when it fails and earning reliance over time. The person's expertise remains theirs, in their company; making it usable must not become an excuse to take it away.

Songbird must be the first place this is visible

The same commitment runs through three relationships: Songbird with the person, the company with its founder, and that company with its customers. We call this the mirror. If Songbird ignores a person's corrections while claiming to build attentive companies, the contradiction is operational.

The underlying design choice is simple: when a rigid system defines the work, people spend their time converting real needs into its categories. AI can instead support the interpretation and coordination around the actual need. This book applies that choice to Songbird's own conduct: the useful next action must follow the person, rather than a predetermined route through a sales process.

Our proposed first step gives a person a correctable account of their useful capabilities that they can keep, including if they leave. The broader journey then explores possibilities and a chosen role before company operation. Those details deserve their own treatment; here the commitment is that understanding and choice must survive each step.

For the eventual customer, the standard is useful continuity and attention to their circumstances. A warm message without completed work cannot satisfy it. A founder's freedom cannot be paid for with a customer's neglect.

Apply the same test to us

These tests are necessary because capability is uneven. Research published in Organization Science in 2026 found AI assistance helping some consulting tasks and hurting another outside the tested capability boundary.⁴ The study is not a scorecard for today's models. It is a reason to test the work being promised.

Measurement itself needs care. METR's February 2026 update describes selection and time-measurement problems that limit conclusions about newer developer productivity results.⁵ We should not choose only easy demonstrations and treat the resulting impression as a whole-company result.

We have not yet established the complete person-first process as a repeatable, proven service. Our existing idea-refinement work is a different achievement. My co-founder's insistence that the offer needs observable proof belongs inside the promise, and this book keeps it there.

Extraordinary describes the quality we intend a company to earn. It does not promise success. Take one offer you are considering, ours included, and ask to see your knowledge change a customer decision. That is where Made From You starts becoming real.