Home / Articles / Every hour is twenty years: the human behind Know the Reach, and why making a song and its video with AI made them more Silvio Switha's, not less

Every hour is twenty years: the human behind Know the Reach, and why making a song and its video with AI made them more Silvio Switha's, not less

By Dinis Cruz · 2026-11-15 · article v1.0.0 · newsroom v0.14.88 · riskmandatemusiccreativityai-videoproductionartistsweb-summitarticle

Abstract: Know the Reach, RiskMandate's song, and the first part of its music video were made with AI, and the story that usually gets told about that is the wrong one. They were made by Silvio Switha, one of the best musicians I know, a singer and producer who runs one of the most successful wedding bands in the Algarve, over hours and hours of the usual creative process: trying, changing, listening, and trying again. The lyrics came from a session that held everything RiskMandate had done and his direction, and the chorus, the structure and the pacing are his decisions. What AI changed is not who had the vision, but how fast he could hear it: a producer with immediate feedback, doing what wealthy, well-produced projects have always done, at a budget a start-up has. Every hour he spent carried twenty years of experience. The numbers in the production audit are perhaps a fifth of what went into it, and among all the versions he could have made, this is the one he chose, which is why people will connect with it. The next step closes the loop: the same song, recorded by his band.

YouTube

In short

• The song and the video were made with AI, and by a person. Silvio Switha, a musician, singer and producer, directed every part of them.
• Every hour he spent carried twenty years of experience. That is why it is good.
• AI changed the speed, not the vision. He worked like a producer with immediate feedback, and stayed in flow.
• The audit's numbers are perhaps a fifth of the work. The rest is the trying, listening and choosing that no log counts.
• When anything can be generated, the chosen version is worth more. People connect with a story and the people behind it.
• The loop closes with humans: the next step is the same song played by his band.

The song we chose to play

Last night, of all the songs we could have listened to, we chose this one: we played Know the Reach and watched the first part of the video. It is a song about what an AI agent can reach and what it was asked to do, and it is genuinely good to listen to.

I have written separately about how the first part of the video was made, from the production audit: 43 generated images, 16 motion clips, 21 edited exports. Those numbers are true. But they tell the story that usually gets told about work like this, that a machine made it, and that story misses almost everything that matters.

Who made it

Silvio Switha is one of the best musicians I know. He runs one of the most successful wedding bands in the Algarve, he is one of the best singers I know, and he is a great producer. He is also one of the six of us who ran ten days at Web Summit, where he was on the ground, on video and on research.

He spent hours and hours on this song and its video. And every hour he spent is twenty years of experience: of knowing what a chorus needs, when a verse drags, which take works and why, what a caption looks like on a phone held at arm's length. Nobody without those twenty years could have given the same directions, in the same way that I can do things when I code with AI that someone less experienced cannot. The tool amplifies what you bring to it.

The lyrics, and who wrote them

You might think: the lyrics were written with ChatGPT. Not really. They were written in a ChatGPT session that held the whole history of what RiskMandate had done, together with Silvio's guidance. The chorus, the structure, the pacing: those are creative decisions, and they are his. The model offered words; he decided which ones the song kept, where they fell, and what they had to carry.

A producer with immediate feedback

There is a myth that making music with AI takes the human out of it. That myth assumes most music is made by many people each being creative. A lot of it is not. Most of the music people listen to is highly produced, and highly produced music is prescriptive: someone at the top has the vision, and prescribes the whole thing, the words, the arrangement, the sound, the look of the video, to writers, session musicians, engineers and crews who carry it out.

That is what Silvio did. He was the producer. The difference is that he was a ridiculously effective one, because he had immediate feedback. He did not brief a writer and wait two weeks, or ask for a mix and hear it two days later. He heard the result in minutes, so he stayed in flow, and he could try dozens, if not hundreds, of variations. He gave clues, he gave guidance, and he cared about the smallest details, which is why it is so good.

It is also a budget story. What he did here is what happens in wealthy, well-produced environments: hiring someone to write a lyric, delegating the parts, directing a crew. A start-up cannot pay for that. With these tools, someone with a vision can do it anyway.

What the numbers miss

The production audit is careful and its numbers are right, but they are perhaps a fifth of what is really there. They count what was saved: the images, the clips, the exports, the words in three chats. They do not count the listening, the variations he heard and dropped, the earlier sessions, the songs that came before this one, or the twenty years. The video did not turn up like that from a prompt. The amount of creative input, of direction, that went into it is enormous.

The version he chose

When AI can make a song in fifty genres, the value of any one generated song goes down. But the value of this version goes up, because it is the version Silvio chose and curated. Put ten variations side by side and this is the one people will relate to, because people do not only relate to a song. They relate to the story, to the video, to the team behind it: a scrappy start-up figuring things out, and the person who did the work.

Closing the loop

I have been half joking and half challenging Silvio to record this song with his band. Then there would be a live version of a song that started with AI, which would be a wonderful full circle, and a recording made by people, with all the connection that brings, built on the base that AI helped find.

I think this is what will happen to many artists who embrace these tools. They will use them to experiment with style, with reach, with how an audience responds. That is what bands have always done: a song you love went through countless versions, played live, tried in the studio, changed again, before it became the one you know. This is another way of doing that, faster, and with the feedback arriving while you are still in the moment.

A song about the board

And then there are the words. Know the Reach is about governance, about the board, about what agents can reach and what they were asked to do. When was the last time you heard that in a song you actually wanted to play again? For people who work in risk and cyber security, that is the connection: a song about our world, made by one of us.

RiskMandate now has two songs on Spotify, and Know the Reach is the one in the video.

What this does not claim

• This is my view, as someone who was there, not a review; Silvio is a colleague and a friend.
• "Perhaps a fifth" is my estimate, not a measurement; the audit's figures are in the first article.
• A recording by his band is a challenge I have made, not a plan anyone has announced.

Where this comes from

My own account, 15 November 2026; the RiskMandate production audit of 14 November 2026, as reported in Know the Reach; the song on Spotify and the first part of the video on YouTube. Earlier on this site: ten days at Web Summit, a film studio on a laptop.

Peer review Not reviewed yet. Review it

Want the next issue by email. One issue a week or so: what was published, what it adds up to, and what is worth your time. Subscribe to the SGit Newsroom →

← All articles