Silvia Canelón

@silviacanelon.com

1107 Followers 541 Following

Data scientist @ UPenn Independent data journalist https://silviacanelon.com she/her/ella 🇻🇪 🏳️‍🌈 Philadelphia #RStats #RLadies #ddj Newsletter: https://silviacanelon.com/contact

My hand holding a hardcover copy of The AI Con: How to Fight Big Tech’s Hype and Create the Future We Want, written by Emily M. Bender and Alex Hanna
Screenshot of a slide about aspects to consider in different AI development stages. Details in following text. Data stage: loss of privacy/consent, surveillance capitalism, labor exploitation, erosion of IP; company & government actors; individuals and workers as impacted people Compute stage: energy/water use, mineral extraction, emissions, geopolitics; company, government, and investor actors; communities, planet, and citizens as impacted people AI models stage: embedded abuse, automation, embedded bias, errors/hallucinations; company, talent, and investor actors; individuals, communities, and businesses as impacted individuals. Applications stage: surveillance, job loss/degradation/discrimination, mis/disinformation; individual, business, and government actors; communities, consumers, democracies, and workers as impacted people
Don't let go!: The challenges and benefits of continuing your data coverage Description: A moderated panel on the challenges and benefits of following a story, and getting multiple stories out of the same or related data sets -- sometimes over the course of months or years. We’ll go through recent examples and discuss how we structured our continuing coverage, how we were able to return to and extend the original data, how we made it work with our newsroom’s editorial strategy, and how we tried to get the most out of every piece. Speakers: Rosie Cima - Big Local News, Amy Fan - Independent journalist, Weihua Li - Bloomberg News, Tom McGinty - Wall Street Journal
View displaying the selection of 3 different accessibility sessions mentioned earlier in the thread, and the Download Calendar button that now appears on the top right of the page. Clicking it would download a .ics file for use in a calendar of your choosing.
Shiny app table view displaying Saturday morning sessions that won't be recorded. Sessions include topics on chatbots, VisiData, sports myths, finding narratives in a pile of documents, mapping with R or Python, visualization R packages, networking for international journalists, education data, and Airtable
Shiny table results after selecting Hands-on from the dropdown menu in the Session Type field. Results include sessions on ChatGPT, SvelteKit, Google Sheets, IRS data, immigration data, Python, disaster data, DNS searching, and georeferencing
Looking up Karen K. Ho in the Shiny app shows two results in the table. One session titled Building a Culture of Accessibility and Why It Matters and another titled Navigating Your Mental Health in Journalism.
Shiny app showing results from a search for keyword "accessibility" which include three sessions: Getting the most out of Datawrapper, Building a culture of accessibility and why it matters, and Using open-source tools to investigate public transit
Shiny app table of NICAR 2025 sessions, displaying fields including session type, time, length, skill level, location, and whether the recorded. The interactive options to filter the sessions include a keyword search bar, a radio button for the day of the conference, a slider for the time of day, a search bar for speakers, session type, and skill level, and lastly, a radio button for whether the session will be recorded. Official conference schedule is posted online at https://schedules.ire.org/nicar-2025/index.html