Get the latest tech news
A discussion of discussions on AI Bias
There've been regular viral stories about ML/AI bias with LLMs and generative AI for the past couple years. One thing I find interesting about discussions of bias is how different the reaction is in the LLM and generative AI case when compared to "classical" bugs in cases where there's a clear bug.
But regardless of the exact weighting function you think should be used to generate a representative dictionary, that's just not going to make a viral news story because you can't get the typical reader to care that a number of the 120 built-in Brotli transforms do things like add " of the ", ". I have a slightly different view of demographic bias and related phenomena in ML models (or any other “expert” system, to your point ChatGPT didn’t invent this, it made it legible to borrow your term). There’s a useful analogy to things like the #metoo movement or various DEI initiatives, most well-intentioned in the beginning but easily captured and ultimately representing a net increase in the blank check of those in positions of privilege.
Or read this on Hacker News
