Jonny Williams

Which approach do you think your organisation should take?

This isn't just a technical challenge, it's a fundamental issue for society.

AI should expand our understanding of human potential, not restrict it. 🚀

I strongly believe that the benefits of open source speak for themselves:
- Transparent development processes
- Community-driven bias detection
- Representation that reflects true diversity
- Democratised technology that empowers, rather than marginalises

This is why I'm making the case for open source AI, and I hope you will too.

🚧 Accessibility Barriers: Closed source AI results in technological gatekeeping that disproportionately impacts underrepresented communities. When AI systems are controlled by a handful of privileged tech companies, who gets to define what "normal" looks like?

👨🏻‍⚕️ Professional Stereotyping: These AI models consistently default to narrow, discriminatory representations. A doctor is assumed male. A nurse, female. A tech leader, white and male. These aren't just harmless defaults. They're powerful examples of exclusion that reinforce workplace inequalities.

🤫 Erased Communities: Models don't just reflect bias, they can actively erase and marginalise diverse identities. The recent David Mayer story shows how even a single name can be arbitrarily suppressed, highlighting how easily systems can be manipulated to control narratives and visibility.

Consider the profound ways closed source AI might perpetuate harmful stereotypes...

AI is not neutral. It's a mirror reflecting and amplifying our deepest societal biases.

Unfortunately we don't have an ability to interrogate most AI models and uncover what training data they consumed. There's plenty of hidden ugliness in closed source AI that we simply don't get to see. 🫠

This is what AI thinks a Manager, Doctor, and Engineer look like - See any issues? 🧔🏻‍♂️

A thread about closed source AI and bias amplification. 🧵

Nice post!

Planning to share some longer form posts on Bluesky as threads... Hopefully they resonate in 🧵 form!

Platform Engineering helps to cement DevOps in a wonderful way.

When I speak with orgs we're not talking about Dev and Ops with a divide. Instead we're talking about Platform Teams enabling DevOps and Product Teams applying DevOps.

It helps DevOps to manifest as a culture and approach. Not a team.

Hoping Bluesky can fill the bird shaped hole in my life