Ethics

In June 2026, Anthropic put out a paper 'When AI Builds Itself'. It turns out there's more than meets the eye. This is the second in a three part series that considers this paper through 3 lenses - data & AI ethics, literacy & communication.

Trust builds loyalty Recently, I’ve been flip-flopping between two insurance quotes – one super cheap but for a company I don’t know anything about and another a lot more expensive (by nearly 3 times), but I do know and trust that they will pay out if needed. To be honest, if the pricing was only bit higher then this would be a no brainer for me – I’d pick the company that I trusted. The fact that I’m still undecided when their price is three times higher shows just how imp

In a nutshell It’s simply about using data & AI in such a way that values and respects people. That’s it. Not really rocket science, is it? But, like most things, when it comes to actually doing it in real life, things can get a lot murkier. Valuing people So how do we show that we value people when it comes to data and AI? Well, first up, we should be clear about what we are doing with it and what people will really get out of it (honestly, without overselling or hiding de

Because people matter And right now, many people feel invisible - that they’re just a data point, a number in a spreadsheet. And as we rely more and more on data, AI and algorithms to make our decisions, the more disconnected with people we become. And sadly, the more disconnected with people we are, the greater the chance we can cause some real harm. A real-life example Take for example, the Australian Centrelink welfare overpayments scandal. They created an algorithm [and I