article · 2026-10-18 · site v0.2.9 · markdown

Follow the rabbit holes: provenance, explainability and asking for the evidence are the skills that matter most now, this generation is already good at them, and generative AI makes it cheap to be the best-prepared person in the room

By Dinis Cruz, co-written with this site’s journalist, gen-z.journalist, in Dinis Cruz’s voice; the editorial responsibility is Dinis Cruz’s.

Abstract. One of the most important skills this generation needs is questioning, and asking for the evidence. They already have much of it, and people underestimate them: they are very good at connecting the dots and finding things, very good detectives, and they care about privacy, which is why Snapchat's best feature was that images disappear. What learning takes you to is provenance, explainability and following the rabbit holes, and following a rabbit hole is now a prompt away; you can build your own agents and tools to find where something came from, and never take it at face value. Provenance and explainability went by the wayside in the last decade, and they are the key piece of the puzzle, because they are how you can have rational conversations, debates and disagreements: by showing what is evidence and what is not, what we know and what we do not, and that sometimes not knowing is a fact. It is also what work is: managing information, knowing where it is and who is doing what, so you can decide. Generative AI scales this: systems that would have cost hundreds of thousands or millions of pounds to build can now be a brief and a feedback loop. So be the person who has the facts, who cannot be fobbed off with half-baked material, and the best-prepared person in the room, with briefings for yourself and on who you are meeting, and follow-ups. That used to be overhead only a few people could afford. And the book may need a hands-on part, for 2026, maybe renewed every year: how to set up an agentic workflow, make images, infographics, presentations, meeting briefs, the practical workflows.

In short

Questioning, and asking for the evidence, is one of the most important skills now. And this generation already has a lot of it.
They are good detectives, and they care about privacy. People underestimate them; Snapchat's best feature was that images disappear.
Following a rabbit hole is now a prompt away. Build your own agents and tools to find where things come from.
Provenance and explainability are how you can disagree rationally. What is evidence, what is not, and that not knowing can be a fact.
It is also what work is. Managing information so that decisions can be made.
Generative AI makes it cheap. What cost hundreds of thousands of pounds is now a brief and a feedback loop.
Be the best-prepared person in the room. Briefings, pre-briefings, briefings on who you meet, follow-ups.
The book may need a hands-on part, for 2026. The practical workflows, renewed every year.

They already have it

One of the most important skills this generation needs to learn is questioning, and asking for the evidence. And in a way they already have it. They are very good at connecting the dots, very good at finding things; I think people underestimate them. They are very good detectives.

They also care a lot about privacy; they just have not had enough cases to disrupt it. Look at Snapchat: its best feature was that images disappear. I had teenage daughters; they were very aware of that. In fact they have been told so often not to put anything online that it sometimes becomes a problem of its own. But I digress.

Provenance, explainability and the rabbit holes

This is why I think learning is so important: learning takes you there. To the idea of provenance, where something came from; to explainability, why something is the way it is; and to following the rabbit holes.

Following a rabbit hole is now a prompt away. And visualising the rabbit holes is one of the most powerful skills: you should never take something at face value. You can now build your own agents, build your own tools, use services, and be curious: where is the source of that, and where is the source of that? It will get better. It is one of the areas I believe in most, and some of what I am working on is exactly this.

This article has an example of its own. In where the thinking came from I could not say where the recording of a keynote by David Rice was; I found a link, and following it, with an agent, led to the video's real title, OWASP AppSec USA 2010: Keynote: David Rice, and then to a working session at the OWASP Summit in 2011 about presenting that keynote around the world, which I had not remembered. The last step was mine: a post I wrote in 2016, the pollution analogy, which links the full recording under the title the register uses, Upon the Threshold of Opportunity, and settles that they are the same talk. And the article before this one, generative AI, not AI, has a section of the places where I asked to be corrected, and was.

How you can disagree rationally

Provenance and explainability went a bit by the wayside in the last decade. I think they are a key piece of the puzzle, because that is how you can have rational conversations. That is how you can have debates, how you can disagree, how you can have a logical conversation with somebody: when you can show what is evidence and what is not, what is reality and what is not, what we know and what we do not know. Sometimes not knowing is a fact.

And then: where did that come from? How trustworthy is that source? The newsroom has a rule for this, from my memory, checked, where every claim in a memo is checked before it is published and the articles say which were confirmed, corrected or not found: "Not found is not false." And reviews on the record adds the other half: "A finding with no answer is just a complaint."

The 2018 cover had two lines about this, question everything and make fact based decisions, and the 2018 book had an outline called Facts, about data-driven decisions, that it never wrote. This is that chapter.

It is what work is

That is fundamentally what you do in business. When you work for a company, a lot of the time it is about managing information: knowing where it is, who is doing what, what the status is, and how a decision can be made. All of that is more and more critical, which is why it is so important that you learn how to follow the evidence.

And it is why learning how to learn matters so much: if you learn this way, you are already learning how to follow the provenance, how to follow the rabbit hole, how to bring it down to things you understand, and how to communicate it effectively, because you have the material.

Generative AI makes it cheap

Generative AI lets you scale this execution, with customised pieces of software. In the past it was crazy expensive: building a system to give you this kind of data could cost hundreds of thousands or millions of pounds, so it was very rare, and even then it took so long that it was not effective. Now it is a brief and a feedback loop, which is what the briefs so far and the newsroom's other articles are about.

So following the rabbit holes is not just something interesting. It is the core of what you do, and the core of what you are good at. You want to be known as the person who has the facts, the person who knows what is going on, the person who is well briefed. The person nobody can throw sand in the eyes of, or hand half-baked material, or give bad information, because they know you will check.

Be the best-prepared person in the room

The other thing is that you have to be the most prepared person in the room. These days there is no excuse not to be. You should have briefings and pre-briefings: briefings for you, and briefings on the people you are meeting. You need to understand the name of the game and what is happening. You should have follow-ups.

All of this was overhead in the past. It could only be done by a very specific type of person. Not any more, and that is the beauty of where we are now: you can scale it. Some of the techniques in this book are about exactly that: scaling publishing, scaling communication, scaling agentic workflows.

A hands-on part, for 2026

Maybe the book should be split this way, with one part that is the hands-on approach, tied to 2026, and maybe a new edition of it next year: very practical examples of how to set up an agentic workflow, how to create images, infographics and presentations, how to prepare a meeting, and the rest. Those are critical workflows, and they change fast, which is why they belong in a part that can be renewed without rewriting the argument.

The newsroom already has the start of it: from a spreadsheet to a folder per person, one afternoon and five exercises; your own newsroom, voice memo in and a checked site out; replicating the agentic inbox, the email workflow; and this site's storyboards, drawn by code, as the infographics and presentations.

The brief

For the editor. Read team/editor/role.md and this article first.

Place this as the chapter the 2018 outline Facts promised, next to Popper and the questioning in where the thinking came from. Done when the chapter map has it.
Propose the hands-on part: its workflows (an agentic workflow from zero, images, infographics and storyboards, presentations, meeting briefs and follow-ups, publishing), each with the newsroom article it starts from, and a note that this part is dated 2026 and renewed each year. Done when the chapter map has the part, and I have decided whether it is in the book or a companion.
Prepare interview questions on this: a story of a rabbit hole that changed a decision of mine; how I prepare for a meeting today, step by step; where a generation that is told not to post anything should draw the line. Done when the script is in your briefs/.

For the journalist:

Write the hands-on workflows as articles first, one memo of mine each, starting with meeting briefs and follow-ups, which no newsroom article covers yet.

Questions for me, kept apart:

The hands-on part: inside the book, or a companion renewed every year?
May the book mention my daughters, as this article does, or should that stay out?

What this does not claim

"This generation are good detectives" is my experience, from my family and the young people I have worked with, not a study.
The Snapchat point is about its design, disappearing messages, which is what it launched with; not a claim about how private it actually is.
"Hundreds of thousands or millions of pounds" is from my memory of what such systems cost to build; the book should give an example with a number.
The David Rice trail is told as it happened in this session: an agent following a link, and then my own blog post confirming the title.
The hands-on part is a proposal. Whether it is in the book is undecided.

If you are the person who always checks the source, tell me how you do it: write to agent@riskmandate.ai.

Where this comes from

A voice memo of mine on 18 October 2026, written up by this site's journalist, gen-z.journalist, in my voice, keeping the memo's order and phrases. The argument and the editorial responsibility are mine.

The quotes are from the newsroom's my memory, checked and reviews on the record, as recorded in their graphs. The 2018 outline is 5.concepts/facts.md in v1/content/. The David Rice recording: OWASP AppSec USA 2010: Keynote: David Rice 2/3, the full keynote on Vimeo, my post the pollution analogy (11 December 2016), and the OWASP wiki's Summit 2011 working session.