#pyconau @vivekkatial.bsky.social A case study on building our first LLM feature – how to balance speed + quality http://youtu.be/N7vV23muhMY

Using research and feedback to build safe and ethical AI. @lpeate.bsky.social and @vivekkatial.bsky.social #yowtls25

Honeycomb did alert reporting 10x faster with Multitudes - chart showing disrupted hours trending down.

Quote: If our team wants to look at the on-call data on a weekly basis instead of a quarterly basis, it’s easy now. - Fred Hebert, Staff Site Reliability Engineer

@honeycomb.io needed to track and report the human impact of on-call alerts.

Together, we co-designed a solution that cut their reporting time by 90% and provided transparent reporting on the impact of alerts.

Read here: www.multitudes.com/success-stor...

Exploring Retrospective Meeting Practices and the Use of Data in Agile Teams Retrospectives are vital for software development teams to continuously enhance their processes and teamwork. Despite the increasing availability of objective data generated throughout the project and...

Proud to share our original research, now available on arXiv! Co-authors, @uvic.ca PhD candidate Alessandra M Paz Milani, @margaretstorey.bsky.social, @vivekkatial.bsky.social, and @lpeate.bsky.social explore how engineering teams take action on data in retros. Read more👇

arxiv.org/abs/2502.03570

OpenAI has to be the most insufferable company in the world. They can steal from the whole world & guzzle all resources. But no one can give them a taste of their own medicine even a little bit.

How long till they use this to say give us even more resources & this is why we can't release anything.

How to overcome AI’s biased perceptions of gender, Thu, Jan 30, 2025, 11:00 AM | Meetup **How to overcome AI’s biased perceptions of gender** In a world where AI increasingly shapes how we see and represent ourselves, understanding how these systems interpret

We're hosting an important discussion on AI's gender bias challenges!

Join @jrosenbaum.com.au for the Multitudes #TechLeaderChat exploring how AI systems oversimplify & misinterpret gender identity.

I watched one of her talks at PyCon AU and it was ️‍🔥 ️‍🔥

www.meetup.com/tech-leader-...

4 Years, 80 Projects A celebration of our last four years

We share some reflections on this post: www.gooddatainstitute.com/post/4-years...

Now, we've supported 60+ charities on 80 projects on 4 continents. GDI has been one of the most rewarding projects I've worked on in my life and we're super proud of all the work.

4 years ago Tom Perfrement kicked off Good Data Institute (GDI) to support NFPs build data capabilities with a mission to make NFPs industry leaders in their use of data analytics, not laggards.

Creating a Data Culture (SSIR) How nonprofit organizations can do a better job with their data.

In 2018 this study from Stanford showed that while 75% of nonprofits collect data, only 6% use it effectively (ssir.org/articles/ent...)

Oo!! will there be an NLP / LLM equivalent paper ?? 👀👀

Screenshot of Table of Contents (Part 1) Contents 1 Introduction 217 2 Positionality 221 3 Overview of Risks and Harms Associated with Computer Vision Systems and Proposed Mitigation Strategies 223 3.1 Representational Harms . . . . . . . . . . . . . . . . . . . 223 3.2 Quality-of-Service and Allocative Harms . . . . . . . . . . 229 3.3 Interpersonal Harms . . . . . . . . . . . . . . . . . . . . . 237 3.4 Societal Harms: System Destabilization and Exacerbating Inequalities . . . . . . . . . . . . . . . . . . . . . . . . . . 245 4 Frameworks and Principles for Computer Vision Researchers 266 4.1 Guidelines for Responsible Data and Model Development . 267 4.2 Measurement Modeling . . . . . . . . . . . . . . . . . . . 271 4.3 Reflexivity . . . . . . . . . . . . . . . . . . . . . . . . . . 273 5 Reorientations of Computer Vision Research 276 5.1 Grounded in Historical Context and Considering Power Dynamics . . . . . . . . . . . . . . . . . . . . . . . 276 5.2 Small, Task Specific . . . . . . . . . . . . . . . . . . . . . 279 5.3 Community-Rooted . . . . . . . . . . . . . . . . . . . . . 280

Screenshot of Table of Contents (Part 2) 6 Systemic Change 285 6.1 Collective Action and Whistleblowing . . . . . . . . . . . . 285 6.2 Refusal/The Right not to Build Something . . . . . . . . . 287 6.3 Independent Funding Outside of Military and Multinational Corporations . . . . . . . . . . . . . . . . . . . . . . . . . 289 7 Conclusion 291 References 293

Dear computer vision researchers, students & practitioners🔇🔇🔇

Remi Denton & I have written what I consider to be a comprehensive paper on the harms of computer vision systems reported to date & how people have proposed addressing them, from different angles.

PDF: cdn.sanity.io/files/wc2kmx...

Our community for Tech Leader Chats has hit 800 members! 🎉

We’re thrilled to hit this milestone, as our group continues to grow and be a place where engineering, product, and data leaders come to learn, grow, and connect.

Keen to listen to this!

Funding research for economic return sounds good – but that’s not how science really works The Marsden Fund was set up to support pure research. Diverting half of it to fund applied research undermines New Zealand’s potential to generate ideas that underpin commercial success.

The Marsden Fund was set up to support pure research. Diverting half of it to fund applied research undermines New Zealand’s potential to generate ideas that underpin commercial success.

@mmitchell.bsky.social if you're trying to report on intersectional accuracy without access to self-reported AND consented sensitive variables - any practices you recommend for still meaningfully reporting on potential intersectional biases? "perceived" measures would just reinforce existing biases

@allenholub.bsky.social “10X programmers actually slow down software delivery” — eventually leading to a steady state where all knowledge is concentrated in one person and they become a key person risk and bottleneck #yow24

Thanks @posit.co for hosting it!!

Empowering Indigenous Communities with Data - Just Reinvest NSW Just Reinvest NSW (JR NSW) is a not-for-profit organisation dedicated towards improving justice outcomes for Aboriginal people.

We did a cool collab with Just Reinvest NSW.

A shiny dashboard that maps Indigenous incarceration rates and justice spending across NSW communities.

Check it out here: www.gooddatainstitute.com/post/empower...

👋 excited to join Bluesky! Keen to see chats on data / ML for social impact.

Here’s hoping this platform doesn’t turn into X/Twitter 🤣

Thank you!! but lol @killedbygoogle.com

Also @timnitgebru.bsky.social - do you have good examples of businesses using Model Cards and making them public?

Welcome!!! Big fan from GDI :)