Are You Machine-Readable?

In the identity century the question is no longer whether people can find you, but whether machines can recognise you.

Earlier this year I rebuilt my website with a very particular objective. It was not about design, and not really even about traffic. I wanted to make my work easier for machines to read.

For most of the internet era visibility meant being searchable. Someone typed a query into Google and the task was to appear somewhere in the list of links. Marketing teams optimised pages, bought keywords and tried to climb the rankings. The assumption behind all of this was simple: people would navigate the web themselves.

That assumption is beginning to change. Increasingly people ask AI systems their questions rather than searching for links. Instead of navigating the web themselves they receive a synthesised answer assembled from many sources across the internet (including some that don't even exist).

Those answers are constructed from material machines can interpret: articles, interviews, forums, documentation, commentary, podcasts and public discussion. Patterns of explanation and repetition become the informational layer from which responses are generated. Once you see this clearly the question is no longer only whether people can find you. The question becomes whether machines can understand you.

For several years my work has explored this question from the perspective of identity. In The Future of You I argued that individuals increasingly need to exist not only as people but as identities that can be recognised and interpreted by digital systems. My ideas on Database Selves explored how, in a networked world, recognition often depends on whether a system can identify a person through structured attributes (database selves) rather than through narrative alone (analogue selves)

https://www.youtube.com/watch?v=5p57pmWUtQ4

As AI becomes embedded across everyday life that idea becomes more concrete. In many situations we are already required to exist as identities that machines can recognise. In that sense we are all becoming, gradually, machine-readable persons.

Rebuilding my website forced me to confront this reality in a practical way. For years my work existed online in many different forms. I had written a book about digital identity, delivered keynote talks on geopolitics and technology, lectured at London Business School, run a podcast, written essays and carried out advisory work with global brands and other organisations. From a human perspective these activities clearly belonged to the same body of work. But from a machine’s perspective they often appeared unrelated.

Search engines did not always, for example, recognise that the author of The Future of You, the keynote speaker presenting The New World Re-Order, the lecturer working with London Business School and the host of a podcast about identity and technology were the same person. The content existed, but it was not organised in a way that machines could easily interpret - or integrate.

So I rebuilt the site around a clearer structure. Instead of treating it as a marketing surface I treated it as a body of work with a coherent architecture. Concepts were better defined. Essays, talks, podcasts and research were connected to the themes they belonged to. Pages were structured around the ideas rather than the format of the content and links popped up much more.

Technically this meant introducing structured data and schema markup so that machines could recognise entities and relationships. Pages began to identify the author, the topic, the type of work and its connection to other pieces. A keynote could be linked to the research behind it, a podcast episode to the concepts it explored, an essay to the broader framework it contributed to. Instead of scattered content there was now a graph of ideas.

Once I had done this for myself the next question became obvious. If individuals increasingly need to be legible to systems in order to function online, what happens to brands?

For decades marketing focused on persuasion. Campaigns, messaging and media placements were designed to influence perception. Visibility could be bought through distribution, with the goal of attention.

Now, large language models are not ranking advertising; they are synthesising knowledge. They surface material that appears across the open web and that is referenced, discussed and explained in multiple places.

In that environment the brands most likely to appear in AI-generated answers will not necessarily be the most advertised. They will be the most understandable and the most referenceable. They will be the brands whose expertise, products and ideas exist clearly within the public knowledge layer of the internet.

Seen this way, becoming machine-readable is not simply a technical exercise. It is a response to a new information environment in which systems increasingly interpret the world on behalf of people.

Search engines, recommendation algorithms, platforms and now AI answer engines do not simply distribute information. They interpret it, connect it, summarise and infer from it. The challenge is no longer only about a brand's - or a person's - visibility. It is about identity itself.

We can say that in an AI-mediated world it is no longer a question of simply announcing who you are.

It is a question of whether the systems interpreting the world can even recognise you as you.