What The Flock?; UnhiredGPT; Lexically Passive Weasels
Apologies for no Drop, yesterday, and no offensive R code drops for Advent of Code the past two days. We had a family medical emergency, and I’m just now catching my breath from it and some Elastic-caused tech disasters (remember folks: friends do not let friends use Elasticsearch).
So, we’ve got three wildly different topics for today. One I’ve been trying to figure out how to cover for the past week or so, one that involves a really daft work situation, and another that’ll help us all commuincate more effectively in 2025, provided you like Bash.
TL;DR
(This is an AI-generated summary of today’s Drop using Ollama + llama 3.2 and a custom prompt + VSCodium extension.)
- BlogFlock reimagines RSS feeds with communal curation features, aiming to foster collaborative content discovery while remaining open-source and ad-free (https://blogflock.com/)
- Generative AI is significantly impacting hiring processes, particularly in resume screening, leading to decreased reliability of application documents and potential disadvantages for non-traditional candidates (https://newsletter.pragmaticengineer.com/p/how-genai-changes-tech-hiring)
- Matt Might’s shell scripts help detect passive voice, weasel words, and lexical illusions in technical writing to improve clarity and precision (https://matt.might.net/articles/shell-scripts-for-passive-voice-weasel-words-duplicates/)
What The Flock?
As noted, I’ve been trying to figure out how to cover this for a while, and I’m still not sure I’ve figured out a truly useful “why” of it. But…
BlogFlock (GH) is a social platform that “reimagines the traditional RSS feed by infusing it with a communal touch” (ick). Developed by Raphael Kabo, it lets us curate and share lists of our fav blogs/feeds. In theory, this is intended to help foster a collaborative environment for content discovery. Unlike conventional feed readers that focus solely on individual consumption, BlogFlock emphasizes collective curation, with a feature that lets us follow lists created by others with another goal of helping folks get more diversity in their information consumption.
This platform supports a wide range of RSS or Atom feeds, including blogs, Mastodon feeds, newspapers, Reddit subreddits, and even podcasts. By making RSS feeds public and shareable, BlogFlock aims to create a more curated and welcoming web experience, countering the impersonal nature of large tech platforms. Kabo is also committed to it remaining open source, ad-free, and independent (it follows the principles of the IndieWeb movement).
Curated and communal feed reading has been explored in various forms over the years, though none have achieved truly massive adoption. The key “why” behind such experiments has often been a desire to break free from algorithmically controlled feeds, foster decentralized web ecosystems, or create more intentional browsing habits.
Google Reader (R.I.P. 🪦 2013) is the towering figure in the history of RSS-based aggregation. It wasn’t communal by design, but its shared items feature let us share and comment on posts. This feature created scores of “micro-communities”. When Google Reader was discontinued, it left a vacuum that prompted the creation of services like Feedly, Inoreader, and The Old Reader, which brought some shared or curated features but rarely emphasized collaboration as BlogFlock does.
Pinboard was created as an alternative to social bookmarking sites like Delicious. It, like, BlogFlock, prioritizes curation over algorithms. While its focus is more on bookmarking than feed aggregation, its communal aspect (shared tags and public bookmarks) parallels the way BlogFlock users can share lists.
The IndieWeb movement has spawned other projects such as Monocle and Aperture, which focus on decentralized feed reading using Microsub. These goal of these tools is to give us more control over our reading experience. Unlike BlogFlock, they focus heavily on self-hosting and technical customization, making them less accessible to normal folks.
There are, of course, more; and the enduring appeal of services like BlogFlock is rooted in the frustration with algorithm-driven content curation. Platforms like Facebook and X have prioritized engagement metrics over meaningful discovery, often leading to information silos, echo chambers, ?or worse. BlogFlock and its peers emerge as antidotes, aiming to:
- reclaim agency over content consumption
- foster serendipity through shared human curation
- support the decentralized ethos of the open web
- encourage niche and long-form content discovery
While I have an account there and have followed some lists others have curated, I’m not sure I have the wherewithal to make it a regular part of my daily info gathering. It’s “one more thing” to invest time into, and I don’t think I’ll get more out of than I do with Inoreader and some custom scrapers I have set up.
If you do end up giving it a go, def let me know how you’re using it.
UnhiredGPT
If you follow me on Bluesky, you likely noticed two kvetch posts about candidates using ChatGPT (et al.) during live interview Zooms. At least half the resumes we’ve seen also been heavily generated by one or more of our present-day wannabe AI overlords, and we’ve seen plenty of examples of exercise submissions that have been clearly generated almost completely by AI.
In response to said kvetching, friend-of-the Drop Mikhail Popov replied with this great post about how generative AI is changing the hiring processes across virtually all industries, but especially in tech.
If you are involved in any part of the hiring process, you 100% should read The Fine Article™. I’m going to (briefly) focus on one aspect: resumes.
Generative AI lets candidates build resumes and cover letters that “precisely” align with job descriptions, running roughshod over HR team keyword filters. This makes it easier for underqualified applicants to advance through initial screening stages, increasing the burden on everyone in the hirhing process to identify good candidates.
This proliferation of AI-generated resumes has (unsurprisingly) thus led to a decline in the reliability of these documents as indicators of candidate quality. As a result, some companies are placing greater emphasis on educational background (ugh) and previous experience, which has the real and present danger of disadvantaging candidates from non-traditional pathways (truly some of my best hires).
I’m not suggesting folks refrain from using “AI” as they hack out a resume creation. I am, however, asserting that job seekers need to realize just how important it is that their resumes accurately and authentically represent their unique qualifications and experiences. We hiring managers and potential co-workers going to find out one way or another, and it’s better to be honest about it than to be caught in a lie.
Again, the whole article is well worth your time, and covers a wide array of sub-topics around AI in the workplace/hiring processes.
Lexically Passive Weasels
In this post, Matt Might discusses the challenges of teaching Ph.D. students to write with precision and clarity. In it, he points out some common issues such as the overuse of passive voice, reliance on weasel words (phrases that sound meaningful but lack substance), and the presence of lexical illusions (unintended repeated words).
He developed a series of shell scripts that automagically detect these patterns in text, integrating them into LaTeX document build systems (you don’t need to use LaTeX to grok the article or use the tools; they should work fine in Typst workflows). The goal is to cajole writers to consciously choose their words and constructions, thereby enhancing the quality of their technical writing.
Given the large adoption of AI in acdemia, perhaps Matt might (heh) consider baking some of the concepts from the Bash scripts into a custom ChatGPT prompt.
FIN
We all will need to get much, much better at sensitive comms, and Signal is one of the only ways to do that in modern times. You should absolutely use that if you are doing any kind of community organizing (etc.). Ping me on Mastodon or Bluesky with a “🦇?” request (public or faux-private) and I’ll provide a one-time use link to connect us on Signal.
Remember, you can follow and interact with the full text of The Daily Drop’s free posts on:
- 🐘 Mastodon via
@dailydrop.hrbrmstr.dev@dailydrop.hrbrmstr.dev - 🦋 Bluesky via
https://bsky.app/profile/dailydrop.hrbrmstr.dev.web.brid.gy
Also, refer to:
to see how to access a regularly updated database of all the Drops with extracted links, and full-text search capability. ☮️
hrbrmstr's Daily Drop 
Leave a comment