Weird
AI is WEIRD - trained to emulate Western, Educated, Rich, and Democratic values. And that's a problem.
A Brief Welcome
Just a small welcome before we get started.
Happy 2026, dear readers of Purposely Digital!
It’s been a while since I’ve written a post. Influenza, a new family member (an older dog!), and a strong desire to respect winter’s call for reflection have been my focuses for the last weeks.
I’m not one for hard-and-fast New Years resolutions. In the past, I tended to take them too literally, push myself too hard, and feel utterly awful when they inevitably did not materialize.
So, this year, I have selected a word - “community.”
For me, 2025 was the year of burnout, of connecting with myself and my needs, and falling in love with a new path of coaching and gentler living. It required a lot of going inward, of cocooning and connecting to myself. I find myself now craving expansion but with more mindfulness and self-compassion for my needs.
I feel a little lost, like I am a toddler learning to walk, as I try to do this. Lots of us learned how to “succeed” in school, but not how to be kind to ourselves, live gently, and connect in authentic ways with others to create community.
That’s why I’m so happy to have you here in this Purposely Digital space exploring how to live intentionally in the digital age and slowly building this community with me. Each comment, DM, or engagement sparks wonderful discussions as we all try to live a life in accordance with our values in this increasingly complex world.
Even if you do not feel called to engage in those ways, I feel the energy of each of your reads and thank you kindly for choosing to use your precious time to be here, in this space, shaping intentional community with me.
Your love and support is helping this toddler walk into 2026 and learn what school did not teach her - gentle, intentional community.
Thank you.
ChatGPT is WEIRD
You may have read or heard researchers calling ChatGPT WEIRD.
What, exactly, does this mean?
It means that the values reflected in the AI system most mirror those of WEIRD countries - Western, Educated, Industrialized, Rich, and Democratic. It is trained on vast amount of data from these parts of the world and reflects the values instilled in the data.
Simply put: WEIRD-in, WERID-out.
It does not reflect the values of the vast majority of the world.
“Values” can be a bit of a fluffy term - abstract and difficult for people to grasp concretely. But they have real, observable implications in AI systems. We can see how these WEIRD values show up in the way AI generates and responds.
Here are three examples:
WEIRD-ness shows up in the way the AI generates text - it’s “cognition” if you will, the way it “thinks,” presenting arguments and answering prompts by mimicking analytical thinking process more common in the West rather than other modes of thinking that are more creative and intuitive.
WEIRD values show up in AI attitudes towards certain ideas and beliefs, adding more weight and legitimacy to those that mimic Western norms over those that do not.
WEIRD cultures are also highly individualistic - and we see that in AI. For example, when asked to describe someone (“They are …”), it will focus on individual traits (“courageous,” “assertive,” “intelligent”) over social roles (“mother,” “husband,” “friend”). In social psychology, this is known as the individualistic vs. collectivist divide in social norms, values, and beliefs across cultures.
As this LinkedIn post I read today put it - this WEIRD phenomenon is not just a matter of bias, but of cultural dominance. (There’s also a nice diagram included in that post showing what cultures ChatGPT most aligns with.)
Is WEIRD AI a problem?
While much attention is on ChatGPT because it is the most popular AI used, many LLMs (Large Language Models) like it struggle with this phenomenon.
Some people don’t view it as a problem, while others do.
I think it is problematic because AI can also unconsciously shape our values and beliefs - sometimes without our explicit endorsement. This poses risks to our autonomy, our ability to be the driver of our own lives, especially for individuals from non-WEIRD cultures.
A year ago, I explored this in a paper with fellow AI ethics researchers Stefan Buijsman and Juan Pablo Bermúdez. An excerpt here:
Over time, an AI system’s existing biases and embedded value sets can be unconsciously inherited by the decision-maker, even when such a shift is in tension with the user’s core values. Together, this threatens their ability to assess such shifts in relation to their practical identity. The lack of awareness that the shifts are occurring means that the decision-maker cannot actively endorse them, and this can result in problematic, cumulative shifts to the agent's beliefs, values, and attitudes over time.
As a coach supporting clients in their goals to live more intentionally in our technological age, I have seen first-hand how tricky navigating an AI system that can unconsciously influence and perhaps not align with our own values can be.
Combating WEIRD AI
Efforts are underway to tackle this problem, including training AI using data from other languages and cultures. However, the WEIRD-ness seems to persist in these systems for reasons that are not fully understood yet - perhaps because the basis or origin of these systems is still inherently WEIRD.
So - in sum - we’re not there yet.
Working with coaching clients, I’ve noticed that taking the time to slow down, reflect on how you want to engage (or not) with these systems in a safe, supportive coaching environment can help individuals feel more empowered to make intentional technological choices.
It’s in no way THE solution - we exist in a complex socio-technical world where powerful forces are shaping the technology around us in ways that are outside of our control. As I wrote in my PhD thesis on smartphone data privacy decision-making - “the cards are stacked against us.” Large systemic change is sorely needed to balance the power dynamic between AI companies and the everyday person.
But that doesn’t mean we have no power. It start with reflection, with space, with slowing down, and intention, instead of rushing into AI and letting it shape you.
You shape your use - and not the other way around.
An Invitation
If you’d like to start creating space for reflection, you can join a free discussion on using AI (or not!) to empower yourself, on January 29th. It will be facilitated by myself for anyone in the world navigating AI and wanting to live more intentionally with it, in accordance with their values. Free and open to all.
Do you think AI use has shaped your beliefs? What values and norms have you observed that don’t align with your own culture or personal beliefs? Please also leave your thoughts in the comments or drop me a DM!
Wishing you a blessed, mindful, and intentional 2026 - with or without AI, as YOU choose.
Other Offerings
Chat with Sarah
Cost: Free
Would you prefer to have a 1:1 chat with me about AI?
Or perhaps you are feeling overwhelmed by your smartphone, social media, another life challenge - or perhaps just by everything? 🤯
As a coach and AI ethics expert, I work with you to move from overwhelm to intentional living - tackling your challenges, whether they be tech-related or not, with a modern, empathetic, and safe approach.
Let’s have a chat. No commitment, no pressure - just a half hour to talk over what’s overwhelming you right now.
Sensitive humans - those who experience the world more intensely - are especially welcome. ❤️