Posts

Link to Bruce Yu's user-driven evaluation of generative web search tools using 5 simple test questions.

[comparing results for [What is generative web search? site:danielsgriffin.com]](/2023/12/14/comparing-results-for-what-is-generative-web-search-site/) For other systems on this theme, see: - [Perplex...

A minor bug report on a search user interface.

What organizations are best situated to apply their own resources and promote developer attention and energy to this problem and opportunity?

I've been thinking about these sorts of questions a lot in relation to our current disruption in web search. See some of my musings in Towards “benchmarking” democratization of good search and Who is...

A speculative tweet thread on terminology choices in documentation, library and information sciences, and information retrieval.

Thinking about what if general-purpose web search were more like searching for directions...

Exploring Google Bard's non-selectable 'Search related topics' links and whether this UX limitation constitutes hostile design.

Loose notes on Google's search result warnings for unreliable or rapidly evolving topics, with Vox and Verge reporting on the feature's 2021 rollout.

Follow-up: very few 2-3 sentence descriptions connect (weblog). What is a "worthwhile task"? [ highlighting added] \@benedictevans The more I look at chatGPT, the more I think that the fact NLP didn’t...

Responds to Benedict Evans' tweet about ChatGPT tasks, arguing the challenge isn't describing tasks briefly but finding the right prompts that connect searchers to their goals. Draws on dissertation research about how data engineers grow queries from code and conversation seeds.

What can and will we do to unseal knowledges?

Here are tweets that make me think of Bailey and Leonardi's work on bridging gaps.

Weblink to Data & Society's 'Trust Issues' event exploring perspectives on community, technology, and trust.

Critiques the '15% of Google searches are new' statistic, examining how Google frames it to suggest impressive capabilities without providing meaningful context.

Looking at a recent paper from Jill Walker Rettberg[^1]—discussing a method of "algorithmic failure", which "uses the mispredictions of machine learning to identify cases that are of interest for qual...

A recommendation.

We have a big chance to really change search for the better. Will we?

A survey of search engine UX micro-interactions, documenting which sites have the search cursor ready on page load versus those that don't, including some surprising failures from academic and library search tools.