Dispatches From The Internets

ChatGPT: New AI system, old bias?

The story of Karen Hunter asking ChatGPT to discuss whether Bessie Smith influenced Mahalia Jackson is pretty telling. As ChatGPT was not fed information that related to the connections between the musical careers of these two influential Black women, it could not shed any light on how they were connected. It could only offer snippets of their biographies, which it gleaned from Wikipedia (or similar).

Despite their seemingly magical “knowledge,” Large Language Models (LLMs) are only able to respond with things they know (or think they know). They aren’t able to create novel connections between subjects in the way that people can.

That’s why it’s absolutely critical that LLMs be trained on content from a variety of sources, perspectives, etc. A LLM is only as good as the data it’s fed. To create truly powerful, creative, and exhaustive LLMs, we need to train them on content created by people whose voices aren’t often centered and subjects that extend far into the long tail.

Disability, Bias, and AI

This is a foundational paper concerning AI and its potential to help and harm people with disabilities. There are a lot of choice quotes in here, but this one really sums up the importance of an intersectional approach to AI:

Integrating disability into the AI bias conversation helps illuminate the tension between AI systems’ reliance on data as the primary means of representing the world, and the fluidity of identity and lived experience. Especially given that the boundaries of disability (not unlike those of race and gender) have continually shifted in relation to unstable and culturally specific notions of “ability,” something that has been constructed and reconstructed in relationship to the needs of industrial capitalism, and the shifting nature of work.

Microsoft’s new Inclusive Design toolkit is designed for the brain

The Microsoft Inclusive Design Toolkit has gotten an update that incorporates cognition, which is awesome! It puts forward three new principles:

  1. Understand the user’s motivation, and the goals and tasks they are trying to complete.
  2. Discern the cognitive load required to reduce that mismatch.
  3. Co-create the final product with a diverse community of people across the spectrum.

It’s so great to see this seminal resource continuing to evolve.

The Great Gaslighting of the JavaScript Era

I feel this piece deep in my bones. If you are a PM, designer, or developer building stuff for the web, you owe it to yourself to give it a read.

I could quote the ever-living crap out of it, but I’ll just drop this one choice excerpt and let you take the rest in, in context:

[W]e’re not asking you to abandon your favorite frontend library on a whim and become a Rails developer, or a Phoenix developer, or a whatever. We’re simply asking you to acknowledge that for years you’ve completely hogged and dominated the #WebDev conversation, ignored our repeated attempts to point out the potential flaws, foot guns, and fallacies with the JS/SPA approach, and in some cases even ridiculed us for our choice of technology stack/language/etc.

CSS image()

The CSS4 image() function is really cool! It enables us to inject portions (fragments) of images, change image direction (flip), provide solid color fallbacks & more.

This is a great writeup from Kevin Powell.

Build things that work, even when parts of it break

Progressive enhancement doesn’t have to be more work

As of this year, I’ve officially been beating the drum of progressive enhancement for decades. With an “s.” And it’s still a philosophy that is foundational to building resilient, accessible projects on the web. Full stop.

Chris offers a great intro/reminder here. And when you want to dig in more, you should read my book.