Hi there,

Last week I wrote about turning back on a mountain, and whether a machine could ever make that kind of call. This week's essay asks the next question: if it could, how would we know we could trust it?

It started with a friend I didn't think was trustworthy when I met him, and the moment that flipped it. That moment is where I think the difference between nice and kind actually lives, and it turns out to be close to what AI researchers are wrestling with too.

The essay's below, then three bits of AI news worth your time. Short enough to read over a coffee.

Sam

The essay

Nice or Kind?

On trust, masks, and what it would take to trust a mind smarter than us.

When I first met one of my closest friends, I didn’t think he was trustworthy. He looked like he’d have been one of the popular kids at school, the kind who might even have picked on me.

Over time, his actions just proved otherwise. He displayed good character, consistently. He would hold space, he was kind, but he also had boundaries.

At first he wasn’t really showing his full self. He was showing a version of himself, like a mask, that he had on for just operating in the real world. Over time I saw that come off, and more of his authenticity and vulnerability came out as he started to trust me and I started to trust him.

Built from my own ideas and stories, with Claude as a drafting partner.

What's moved in AI this week

Three things worth knowing from the past week.

Practical

Google built Gemini a hub for small businesses

On 20 October Google gave Gemini a dedicated space for small businesses: a business hub in the Gemini app that connects to outside tools including Shopify, Xero and Canva, and can turn messy spreadsheets, receipts and purchase records into clear financial graphics in seconds.

Why it matters: If you already pay for Gemini, point it at a pile of your own receipts or a messy spreadsheet this week and see what it pulls out, before you lean on it for anything that matters.

Source: Google

Australia

If you use AI in a workplace case, you now have to say so

From 20 October, anyone who uses generative AI to prepare a document for a Fair Work Commission case, such as an application, a submission or a witness statement, has to say so in the document, check that every detail in it is correct, and state that they did. Some Commission forms now carry a question about it.

Why it matters: If you ever end up in an unfair-dismissal or workplace dispute, anything you or your representative ran through AI now has to be declared and checked line by line. The shortcut comes with your name on it.

Source: Fair Work Commission

Big picture

The company behind Claude filed to go public

On 26 October Anthropic, the maker of Claude, filed to list on the Nasdaq under the ticker ANT. Its filing reported $16.3 billion in revenue for the first half of 2026, up from $1.2 billion a year earlier, and a loss of $80.7 billion, about $72 billion of it an accounting charge rather than money spent.

Why it matters: The tools we lean on are built by companies still racing to win the market. Worth remembering that pricing, and what's free today, can move as that race plays out.

Source: CNBC