Montly Update: July, 2024
We got to 2000 subscribers! Updates on ongoing efforts and future plans.
Hello, FILWD subscribers! I hope you are having a great summer! Things have slowed down considerably in this part of the world, and I have more time to reflect on the newsletter’s past and future. Here is a summary of what happened this last month and a few reflections on the near future.
We got to 2000!
Yesterday, the newsletter achieved a major milestone: we reached 2000 subscribers.
While this number is nothing intrinsically special and is quite tiny by Internet standards, it has a special value for me. When I started writing FILWD seriously around the fall/winter of last year, I set 2000 subscribers as my main goal for the year. Achieving this goal much earlier means the newsletter has a higher potential than I had estimated. Now, it’s time for me to set new milestones while I keep focusing on producing useful content for my readers. Later in this post, I’ll provide an overview of what I am working on and what you should expect in the near future. In the meantime, I want to thank you for being part of this experiment. It’s amazing to know I can reach 2000 of you directly in your inbox every time I click the “publish” button. I hope you’ll keep finding my content useful and inspiring.
Posts published in June
June started with a post on the 3rd edition of the FILWD reading club:
The reading club is an online gathering where I meet with a group of FILWD readers to discuss a specific paper. The meetings last one hour and are very lively, engaging, and fun.
In this third reading club we covered a paper on “visualization mirages,” visualizations that look legitimate but lead to erroneous takeaways.
In the post I summarize the ideas we discussed and propose reflection in the reading club format and its future.
The second post of the meeting month was devoted to my series on Visualization for Machine learning.
The fifth post of the series introduces the concept of “machine learning explanations” and paves the way for future posts that will focus on visualization techniques for different types of explanations.
The third and final post of the month introduced the idea of “semantic clashes” in visualization.
The idea I am trying to capture here is that sometimes, visual representations suggest certain relationships or concepts that deviate from the meaning of the data, generating confusion or misinterpretation.
I have been circling around this idea of how visual representations suggest certain concepts for a while, and this is just another attempt to capture this idea. Take a look at the examples in the post. They will make the problem immediately apparent.
Ongoing efforts and future plans
If you are a regular reader, you may have noticed that FILWD has a number of projects I keep developing over time.
Reading club
I already mentioned the reading club above. Here, I want to give you a glimpse of what is in my mind for its future. For now, the reading club is on pause. I will probably restart at the end of summer. This is not because I am not happy with it but more because I need to concentrate on other aspects of the newsletter during the summer. When we restart, the major differences will be that I will set dates and times for 5-6 reading clubs at once, and I will advertise the reading club to all my subscribers (until now, it was limited to a subgroup).
Rhetorical Data Visualization Course
After months of struggling to make progress, I have finally been able to keep developing my Rhetorical Data Visualization course, which I plan to make available online. Last week I finished recording module 3 on “data transformation,” and I started designing module 4 on “visual representations.” My goal is to record a total of 5 modules and then package it as an online course people can sign up to. At some point, I will look for people who want to test the course. If you think you might be interested, add a comment to this post, and I’ll send you information on how to preview the course.
Visualization for Machine Learning
I have plans to publish quite a few more posts in the series on data visualization for machine learning. The next 2-3 posts will be a follow-up of my last post on ML explanations. Each post will focus on how to visualize specific kinds of explanations. Once this part is done, I will move on to the harder part: visualizing deep learning and model components. I will need to do more research here because I am not as knowledgeable as I would like to be. However, reading papers and articles will bring new material and ideas to share here, so you can expect some updates on this area as I develop my expertise. As I mentioned in the newsletter before, my plan is to eventually turn this into a book. It’s going to take a little while because my focus for the summer is the online course I mentioned above. But piece by piece, I will keep developing the series, and eventually, it will be done and ready to be turned into a book.
FILWD chats
During the last few months, I published a couple of recorded chats. One with Nick Desbarats and one with Amanda Makulec. I had a ton of fun recording these chats, and I have been meaning to record more. I am still uncertain if I want to make this a more regular element of the newsletter (e.g., by planning to have one chat a month), but I have a few chats I want to record in the pipeline, so stay tuned for more interesting conversations in the upcoming weeks.
ChatGPT experiments
My post on data analysis with ChatGPT gained a lot of traction, and I have learned a lot from it. I have been meaning to record more data analysis sessions and add more reflections about this space, but alas, I have not yet found the determination to record a whole new session. However, I have been collecting more ideas about the future of data analysis with LLMs, and I intend to publish more thoughts about it over the summer. I am also doing research actively in this space, so I will be able to report some results from our lab in the future.
—
That’s all for now. Let me know what you think about my plans, and feel free to send recommendations or requests if you feel like it! Thanks for reading.
Please help me spread the word about FILWD. Let your colleagues and friends know they can find all this material and beyond here. The more we are, the more useful this newsletter will be for everybody! Thanks.
Thanks for allowing me to join the group.
Congratulations! 🥳🥳🥳