Advertisement · 728 × 90

Posts by Alark Joshi

Earlier this week Harvard's Seyed Mahdi Hosseini Maasoum and Guy Lichtinger outlined another helpful framework for thinking about this. Where AI enhances individual workers' productivity it widens wage inequality by multiplying the returns to specialist skills and knowledge; where it reduces barriers to entry by eliminating the need for specialist skills it has the opposite effect. Interestingly, they find evidence that both dynamics are already playing out in measured AI usage, and that at the whole economy level they can to an extent cancel one another out.

Earlier this week Harvard's Seyed Mahdi Hosseini Maasoum and Guy Lichtinger outlined another helpful framework for thinking about this. Where AI enhances individual workers' productivity it widens wage inequality by multiplying the returns to specialist skills and knowledge; where it reduces barriers to entry by eliminating the need for specialist skills it has the opposite effect. Interestingly, they find evidence that both dynamics are already playing out in measured AI usage, and that at the whole economy level they can to an extent cancel one another out.

This is good from @jburnmurdoch.ft.com on how AI will impact on the labour market - two different effects pulling in different directions. Which will be more powerful?
www.ft.com/content/becb...

2 weeks ago 52 17 1 4
Preview
Graphics Intern Job Title: Graphics Intern Location: New York Application Deadline: March 26th, 2026. About Springer Nature Springer Nature is one of the leading publishers of research in the world. We publish the la...

🚨Internship opportunity!🚨 Now accepting applications for Scientific American summer/fall graphics internship: springernature.wd3.myworkdayjobs.com/en-US/Spring... Apply by March 26. Please feel free to share, especially within the 📊 dataviz community.

1 month ago 10 11 0 0

As a fan of small multiples, I love this. I have shared it with my students in the data visualization course. 📊

2 months ago 2 1 0 0

Beautifully hand-crafted interactive data visualization about birds by @nadiehbremer.com !! The tooltips and transitions are particularly fantastic. The only person I know who's a huge birder that may enjoy this website is @derekbruff.bsky.social. 🦉

2 months ago 4 0 1 0

Paper details: Saske, Antonia, et al. "A Multidimensional Assessment Method for Visualization Understanding (MdamV)." IEEE TVCG (2026).

2 months ago 3 0 0 0

They also found that the topic of the chart matters, and readers were skeptical of the chart if they had preconceived notions of the topic, such as climate change. More details can be found at arxiv.org/pdf/2410.238....

2 months ago 4 0 1 0

There is research from last year's IEEE VIS that talks about how aesthetically pleasing charts are believed by participants in their study a lot more than basic charts.

2 months ago 2 0 1 0
Map of snowfall totals in the contiguous United States for the weekend of January 23–26. A band of accumulation of at least 8 inches stretched from northern Texas to Cape Cod, Massachusetts.

Map of snowfall totals in the contiguous United States for the weekend of January 23–26. A band of accumulation of at least 8 inches stretched from northern Texas to Cape Cod, Massachusetts.

Map of winter temperature trends in the contiguous United States. In the three-month winter average, it’s warming everywhere (red). In the single month averages there are some small cooling regions (light blue) including the extreme Pacific Northwest in December and the Mississippi Delta in January.

Map of winter temperature trends in the contiguous United States. In the three-month winter average, it’s warming everywhere (red). In the single month averages there are some small cooling regions (light blue) including the extreme Pacific Northwest in December and the Mississippi Delta in January.

I made some maps to help @climate.us explain how we’re still getting winter storms — sometimes big ones! — despite Global Warming. www.climate.us/news-and-fea...

2 months ago 27 6 1 0

This is unbelievably fantastic!! Congratulations, @albertocairo.com and team. What an amazing resource!! 🙏

2 months ago 1 0 0 0

Fantastic write-up about data visualization issues such as clarity, accessibility, using descriptive text, and more with great examples from @ourworldindata.org by Saloni (@scientificdiscovery.dev). Great place to start for data visualization enthusiasts and students. 👏

4 months ago 4 0 1 0
Advertisement

Happy birthday, Craig!! 🎉 Thank you for supporting our students at the University of San Francisco.

4 months ago 1 0 0 0
Preview
Postdoctoral Research Opportunity – Global LLM Values Benchmarking Working Group: Postdoctoral Affairs - Northwestern University

🚨 Postdoc opening -- Come work with us at Northwestern! 🚨 This position is to work on evaluating LLMs for sociopolitical value expression: postdocs.northwestern.edu/announcement... -- Please share!

4 months ago 4 6 0 0
Preview
Fellow, Upshot New York, NY

YOU, or someone talented you know, could be the next @upshot.nytimes.com fellow.

It’s a paid, full-time, one-year job working on the staff of (dare I say) the coolest desk in the newsroom.

But hurry — applications close on Wednesday, so apply now and/or share with friends!

5 months ago 0 4 0 0
Video

🗺️ The true size of the world's countries.

👉 The widely used Mercator projection distorts their size.

5 months ago 74 23 0 2
Post image

#IEEEVIS - we’re hiring @ City St George’s, University of London.

Does this sound like you?
bit.ly/3WEYlmm

It's a permanent (tenured) position:
www.jobs.ac.uk/job/DPC906/l...

Join us @ giCentre & be part of our amazing new ...
Data Visualization Doctoral Training Centre:
bit.ly/diverseCDT25...

5 months ago 5 5 0 0
LinkedIn This link will take you to a page that’s not on LinkedIn

Bob Laramee will be presenting our paper "Examining a PCP Intervention through the Anderson and Krathwohl Taxonomy Lens" at the upcoming EduVis workshop at IEEE VIS on Monday, November 3.

More details including the PDF of the paper can be found at www.cs.usfca.edu/~apjoshi/pap...

5 months ago 3 0 0 0
LinkedIn This link will take you to a page that’s not on LinkedIn

Teaching data visualization? Our research reveals that students need more support with procedural knowledge (tracing polylines, mapping colors to polylines) when learning Parallel Coordinates Plots. The Anderson & Krathwohl taxonomy helped us pinpoint exactly where interventions should focus.

5 months ago 3 0 1 0
Graph with title: Older people are addicted to their digital gadgets. But the graph itself only has data about device ownership %

Graph with title: Older people are addicted to their digital gadgets. But the graph itself only has data about device ownership %

This graph that the @economist.com is boosting on Facebook is irresponsible.

Titles create powerful frames in the way people interpret data (see: Kong et al 2019, www.hidykong.com/papers/Kong-...). This one tells a story that doesn't exist in the charted data.

5 months ago 9 1 1 1
Advertisement
Preview
Data Visualizations as Propaganda: Tracing Lineages, Provenance, and Political Framings in Online Anti-Immigrant Discourse | Proceedings of the ACM on Human-Computer Interaction Along with other visual content, data visualizations are increasingly used within online discourse, including political communication. Though often considered to be ''objective'', data visualizations ...

Really excited to share this new paper on "Data Visualizations as Propaganda", co-led by PhD students Priya Dhawka and Nina Lutz, which just won a Best Paper award at the CSCW conference: dl.acm.org/doi/10.1145/...

[Short thread]

5 months ago 225 79 5 7

The CS dept at USF is hiring an Assistant/Associate Professor. The job is perfect for an individual who enjoys teaching small classes (25-30), mentoring students on research projects, & working in a truly diverse department in an urban setting. Apply at usfca.wd5.myworkdayjobs.com/en-US/USF_Fu....

6 months ago 1 0 0 0
Preview
Home | ReVISit reVISit: Reproducible and Powerful Visualization User Studies

Revisit by @laneh.bsky.social and team is an excellent option. It requires some configuration and setup, but it works beautifully and all your data stays on your server. (revisit.dev)

6 months ago 8 3 0 0

AI is not a mere tool for helping teachers. When you remove the human element of teaching and learning, you are not being more efficient nor effective. The ability to learn is what makes us human, and humans teaching humans is a fundamentally human act that is necessary for meaningful engagement.

6 months ago 14 4 0 0
What Americans die from and the causes of death the US media reports on.

4 stacked bar charts. showing in short that while heart diseases and cancer constitutes 55% of the causes of death, they receive about 7% of the media coverages. Homicide is under 1% but receives between 42% and 52%. Terrorisme barely registers in the causes of death, but gets between 11% and 18%.

The first stacked bar is causes of death in the US in 2023
Heart diseases 29%
Cancer 26%
Accidents 9.5%
Stroke 6.9%
Lower respiratory diseases (6.2%)
Alzheimer's disease (4.8%)
Diabetes (4.0%)
Kidney failure (2.4%)
Liver disease (2.2%)
Suicide (2.1%)
COVID-19 (2.1%)
Influenza/Pneumonia (1.9%)
Drug overdose (1.8%)
Homicide (<1%)
Terrorism (<0.001%)

Media coverage of these causes of death in 2023 in...
New York Times
Heart disease (2.8%)
Cancer (4.1%)
Accidents (9.7%)
Suicide (3.8%)
COVID-19 (5.3%)
Drug overdose (7.5%)
Homicide (42%)
Terrorism (18%)

Washington Post
Heart disease (2.9%)
Cancer (4.7%)
Accidents (5.9%)
Suicide (3.3%)
COVID-19 (7.9%)
Drug overdose (9.5%)
Homicide (46%)
Terrorism (12%)

Fox News
Heart disease (2.3%)
Cancer (3.8%)
Accidents (6.1%)
Suicide (4.1%)
COVID-19 (6.0%)
Drug overdose (9.8%)
Homicide (52%)
Terrorism (11%)

Note: Based on the share of causes of death in the US and the share of mentions for each of the causes in the New York Times, the Washington Post and Fox News. All values are normalized to 100%, so the shares are relative to all deaths caused by the 12 most common causes + drug overdoses, homicides and terrorism. These causes account for more than 75% of deaths in the US.
A "media mention" is a published article in one of the outlets which mentions the cause (e.g. "influenza) or related keywords (e.g. "flu") least twice.
Data sources: Media mentions from Media Cloud (2025); deaths data from the US CDC (2025) and Global Terrorism Index.

Fox News

What Americans die from and the causes of death the US media reports on. 4 stacked bar charts. showing in short that while heart diseases and cancer constitutes 55% of the causes of death, they receive about 7% of the media coverages. Homicide is under 1% but receives between 42% and 52%. Terrorisme barely registers in the causes of death, but gets between 11% and 18%. The first stacked bar is causes of death in the US in 2023 Heart diseases 29% Cancer 26% Accidents 9.5% Stroke 6.9% Lower respiratory diseases (6.2%) Alzheimer's disease (4.8%) Diabetes (4.0%) Kidney failure (2.4%) Liver disease (2.2%) Suicide (2.1%) COVID-19 (2.1%) Influenza/Pneumonia (1.9%) Drug overdose (1.8%) Homicide (<1%) Terrorism (<0.001%) Media coverage of these causes of death in 2023 in... New York Times Heart disease (2.8%) Cancer (4.1%) Accidents (9.7%) Suicide (3.8%) COVID-19 (5.3%) Drug overdose (7.5%) Homicide (42%) Terrorism (18%) Washington Post Heart disease (2.9%) Cancer (4.7%) Accidents (5.9%) Suicide (3.3%) COVID-19 (7.9%) Drug overdose (9.5%) Homicide (46%) Terrorism (12%) Fox News Heart disease (2.3%) Cancer (3.8%) Accidents (6.1%) Suicide (4.1%) COVID-19 (6.0%) Drug overdose (9.8%) Homicide (52%) Terrorism (11%) Note: Based on the share of causes of death in the US and the share of mentions for each of the causes in the New York Times, the Washington Post and Fox News. All values are normalized to 100%, so the shares are relative to all deaths caused by the 12 most common causes + drug overdoses, homicides and terrorism. These causes account for more than 75% of deaths in the US. A "media mention" is a published article in one of the outlets which mentions the cause (e.g. "influenza) or related keywords (e.g. "flu") least twice. Data sources: Media mentions from Media Cloud (2025); deaths data from the US CDC (2025) and Global Terrorism Index. Fox News

This is really one of the best charts by @ourworldindata.org 📊

Amazing how much research and work goes into creating a chart like this. And it's such a good insight into society.

6 months ago 46 16 1 3
Preview
!!!* INTERNSHIP OPPORTUNITY *!!! Hey everyone… great news! please share with your network of budding young students.   We have now opened applications for our internships in Story, Ideation, Concept…... !!!* INTERNSHIP OPPORTUNITY *!!! Hey everyone… great news! please share with your network of budding young students.   We have now opened applications for our internships in Story, Ideation, Concept ...

2026 internships available in my team at The LEGO Group!

Internships in
Story,
Ideation,
Concept Model Design (ie, building with bricks),
Prototyping (3D printing etc),
Hardware (soldering and weird sensors!)
Physical (springs and spinny things)
Digital (code for doing stuff with bricks)

6 months ago 45 45 1 5
Preview
Measles vaccines save millions of lives each year Measles once killed millions every year. Vaccines changed this, preventing disease, long-term immune damage, and deadly outbreaks.

Your regular reminder that measles vaccines are the most lifesaving childhood vaccine in the schedule, and save over a million lives globally per year.

ourworldindata.org/measles-vacc...

7 months ago 234 93 3 5
Post image

Disease prevalence in US states before & after vaccine introduction 🧪

From Edward Tufte & graphics.wsj.com/infectious-d...

7 months ago 1825 894 40 66
Chinese writing, with last phrase highlight.
Something like, 

第七条 互联网应用程序分发平台在应用程序上架或者上线审核时,应当要求互联网应用程序服务提供者说明是否提供人工智能生 务。互联网应用程序服务提供者提供人工智能生成合成服务的,互联网应用程序分发平台应当核验其生成合成内容标识相关材料。
第八条 服务提供者应当在用户服务协议中明确说明生成合成内容标识的方法、样式等规范内容,并提示用户仔细阅读并理解相关 理要求。
第九条 用户申请服务提供者提供没有添加显式标识的生成合成内容的,服务提供者可以在通过用户协议明确用户的标识义务和 后,提供不含显式标识的生成合成内容,并依法留存提供对象信息等相关日志不少于六个月。
第十条 用户使用网络信息内容传播服务发布生成合成内容的,应当主动声明并使用服务提供者提供的标识功能进行标识。
任何组织和个人不得恶意删除、篡改、伪造、隐匿本办法规定的生成合成内容标识,不得为他人实施上述恶意行为提供工具或者 得通过不正当标识手段损害他人合法权益。
第十一条 服务提供者开展标识活动的,还应当符合相关法律、行政法规、部门规章和强制性国家标准的要求。
第十二条 服务提供者在履行算法备案、安全评估等手续时,应当按照本办法提供生成合成内容标识相关材料,并加强标识信息共 范打击相关违法犯罪活动提供支持和帮助。
第十三条 违反本办法规定的,由网信、电信、公安和广播电视等有关主管部门依据职责,按照有关法律、行政法规、部门规章的 处理。
第十四条 本办法自2025年9月1日起施行。

Chinese writing, with last phrase highlight. Something like, 第七条 互联网应用程序分发平台在应用程序上架或者上线审核时,应当要求互联网应用程序服务提供者说明是否提供人工智能生 务。互联网应用程序服务提供者提供人工智能生成合成服务的,互联网应用程序分发平台应当核验其生成合成内容标识相关材料。 第八条 服务提供者应当在用户服务协议中明确说明生成合成内容标识的方法、样式等规范内容,并提示用户仔细阅读并理解相关 理要求。 第九条 用户申请服务提供者提供没有添加显式标识的生成合成内容的,服务提供者可以在通过用户协议明确用户的标识义务和 后,提供不含显式标识的生成合成内容,并依法留存提供对象信息等相关日志不少于六个月。 第十条 用户使用网络信息内容传播服务发布生成合成内容的,应当主动声明并使用服务提供者提供的标识功能进行标识。 任何组织和个人不得恶意删除、篡改、伪造、隐匿本办法规定的生成合成内容标识,不得为他人实施上述恶意行为提供工具或者 得通过不正当标识手段损害他人合法权益。 第十一条 服务提供者开展标识活动的,还应当符合相关法律、行政法规、部门规章和强制性国家标准的要求。 第十二条 服务提供者在履行算法备案、安全评估等手续时,应当按照本办法提供生成合成内容标识相关材料,并加强标识信息共 范打击相关违法犯罪活动提供支持和帮助。 第十三条 违反本办法规定的,由网信、电信、公安和广播电视等有关主管部门依据职责,按照有关法律、行政法规、部门规章的 处理。 第十四条 本办法自2025年9月1日起施行。

🤖📰 Effective YESTERDAY: China has mandated a digital watermark for all AI-generated content.
www.cac.gov.cn/2025-03/14/c...
Translating in 🧵.

7 months ago 137 53 1 9
Advertisement
Preview
Amazon cloud chief says replacing junior employees with AI is 'one of the dumbest things I've ever heard' Amazon cloud chief Matt Garman says replacing junior staff with AI is short-sighted, calling it "one of the dumbest things" he's ever heard.

Finally, someone is saying this out loud!!! “Amazon cloud chief says replacing junior employees with AI is 'one of the dumbest things I've ever heard' How's that going to work when you go 10 years in the future & you have no one that has built up/learned anything?"

finance.yahoo.com/news/amazon-...

8 months ago 7 0 0 0

Additionally, with all the companies hiring Senior Software Engineers, I wonder where the next generation of "Senior" software engineers are going to come from. Maybe with AI, we won't need senior software engineers too in a few years 🙄

8 months ago 3 0 0 0
Preview
75 years of innovation: How F1 has evolved since 1950 and where it's headed To celebrate F1's 75th anniversary, ESPN charts the sport's evolution since 1950 through its most iconic cars, and details where it's headed in 2026.

This data-driven story - "How F1 has evolved since 1950" from @espn.com Creative Studio uses horizontal, vertical, and even diagonal scrollytelling. The hand-drawn illustrations & visuals are spectacular! 🤩www.espn.com/espn/feature/story/_/id/...

8 months ago 1 0 0 0