🚨📢 I’m delighted and proud to share our new
@Nature
paper “Online Images Amplify Gender Bias” What better day than Valentine’s Day to learn about gender stereotypes?🩷A thread 🧵
Big update!📢After four wonderfully enriching years
@BerkeleyHaas
, I will be joining
@StanfordGSB
as an assistant prof. in Organizational Behavior this summer. I am incredibly grateful for my mentors, friends, and family who have made this dream possible. I am deeply thankful to
So excited to share that I’m joining Berkeley as an assistant prof. in the Management of Organizations! I’m immensely thankful to all my family, friends, colleagues and mentors!
@AnnenbergPenn
was a wonderful home, and I can’t wait to join the fantastic community
@BerkeleyHaas
1/ Our new paper is out today in
@NatureComms
! “Experimental Evidence of Scale-induced Category Convergence across Populations.” In a time of so much political division, I hope this study can help remind us of the fundamental ways we are all connected.
What is the best way to spread beliefs, behaviors, and technologies?
@DamonCentola
and I provide novel insights into this problem in “Topological Measures for Identifying and Predicting the Spread of Complex Contagions” out today in
@NatureComms
🧵
I am deeply honored to receive the best talk award
@Ic2s2
for our project “Online Gender Stereotypes are Stronger in Images than in Text.” I’m so grateful to all of my collaborators and for all the enormously helpful feedback we received from this community!
Today
@Ic2s2
11am I’ll present research on comparing stereotypes between images and text across social media platforms. Come by! I’d love your feedback. This work is supported by the
@BerkeleyHaas
center for equity, gender, & leadership & the fisher center for business analytics
1/Is the enemy of your enemy your friend? In a new paper we show the opposite, where exposure to the threat of a common enemy increased polarization among Reps and Dems. Our evidence comes from a pre-registered bot experiment during the 2020 US-Iran Crisis
🚨🎨New preprint forthcom. in Cognition. We reveal “Color Associations in Abstract Semantic Domains” (disciplines, music genres, and emotions) using info.theoretic measures via our python package (compsyn) to analyze color distributions in
@Google
Images
Our new paper is out in
@PNASNews
! We built an app to show how collective intelligence in social networks can improve clinicians’ diagnoses and treatment recommendations. A huge many-year effort led by
@DamonCentola
with a stellar team!
🆕
@PNASNews
Reducing medical errors.
A randomized trial of >3,000 physicians, with some assigned to sharing responses for medical diagnosis vignettes, demonstrates collective intelligence, especially for the least accurate doctors
I’ve been asked to review for a top sociology journal and a top psych journal in the same week. I know most people complain about reviewing, but this interdisciplinary signal feels pretty cool 😎😊🙏🏻
Huge congratulations to our rock start PhD Student
@MTranchero
on this fantastic position! For folks interested in computational studies of innovation and the data revolution, follow lots of exciting work to come from Matteo!
I'm thrilled to share that this summer I'll be joining
@Wharton
@Penn
as an Assistant Professor of Management.
I am incredibly excited for this next step, and grateful to my advisors, mentors, friends and housemates who made this journey both possible and fun.
New paper in Management Science w.
@AustinVanLoon
Katharina Lix
@amirgoldb
& Sameer Srivastava: “Exposure to the Views of Opposing Others with Latent Cognitive Differences Results in Social Influence—But Only When Those Differences Remain Obscured.”
I’m attending the Network Evolution conference
@INSEAD
this weekend, presenting on our measure of complex centrality and the hidden power of the periphery.
Excited to be back in Europe and looking forward to the conversations!
Needless to say, our search was successful 😁 welcome
@ericarbailey
! Can’t wait to continue nerding out about the psychology of authenticity and beyond. Another win for
@BerkeleyHaas
!!
Today
@Ic2s2
11am I’ll present research on comparing stereotypes between images and text across social media platforms. Come by! I’d love your feedback. This work is supported by the
@BerkeleyHaas
center for equity, gender, & leadership & the fisher center for business analytics
Computational social scientists! I’ll be in Copenhagen for
@IC2S2
as of tomorrow. If you’re around and want to talk ideas, let’s meet up! Feel free to DM me 😃
I just realized. As someone who studies categories, my r scripts make me look like I'm totally obsessed with cats. get_cat_data. link_cat_data. cat_analysis_04.
I’m really excited to be speaking
@ucdavis
today about our recent Nature paper Looking forward to the conversations! Thanks for the invite
@enfascination
@JingwenZhang
!
A source of wholesome positivity in academia—when you write a reference letter for an amazing student to help them get an awesome opportunity, and then they actually get it! Just happened to me for the first time. It’s a wonderful feeling.
I’m thrilled to announce that I’ve joined the Computational Culture Lab as a faculty director, alongside
@amirgoldb
, Sameer Srivastava, and Julien Clement. You can read more about our team research agenda here Stay tuned!!
My wife
@ninaguilbeault
has published a brilliant and timely book - THE GOOD EATER - about the profound dilemmas and tradeoffs that define our food system, and how individual and cultural changes (relating to diet and our consumer choices) can steer us in the right direction.
So honored to receive the computational modeling prize @ cogsci 2019. Thanks to my collaborators in
@NDGannenberg
@a_baronca
for all our great work!
@AnnenbergPenn
pls reach out if you want to talk about our forthcoming paper on the emergence of category systems in soc. networks!
I’m super excited to be speaking
@MITSloan
's System Dynamics group tomorrow about network measures for predicting complex social contagions. As someone who is fascinated by the history of cybernetics, it feels especially cool to be presenting to a group founded by Jay Forrester!
I'm excited to present tomorrow in the Computational Social Science Seminar at MIT hosted by the Connection Science group. I'll be discussing how social networks can induce cross-cultural regularities in the formation of linguistic categories. Thanks
@_ziv_e
for the invitation!
I’m feeling grateful today! In our last in-person session of Power & Politics, the students insisted we take a group photo to commemorate our journey! It's a real privilege to work with such a bright, engaging, and respectful group. Can’t wait to see what these students do next!
Our paper ‘Social Learning and Partisan Bias in the Interpretation of Climate Trends’ is out! We show collective intelligence can reduce partisan bias, but not when ppl are primed to think of political identity
@joshua_a_becker
@NDGannenberg
@AnnenbergPenn
Too busy to read through? Looking for some visuals?😉This short animation ft. myself and my wonderful coauthor
@solenedelecourt
gives you the gist of our findings
In 2022, we published a paper challenging the popular idea that “the enemy of my enemy is my friend” w.
@eamanjahani
@chris_bail
@fmerhout
and others. We found that Republicans and Democrats were less willing to cooperate after exposure to a common enemy.
So honored to receive Facebook’s award to fund my next set of network experiments on content moderation! Thank you to my advisor Damon Centola and the broader
@NDGannenberg
and
@AnnenbergPenn
community for the ongoing discussions and support!
I’m feeling enormously honored and grateful to have had the opportunity to collaborate with
@ned_smith
&
@joshua_a_becker
.
@ned_smith
always made research deep, meaningful, and effortlessly fun. I will miss his big warm smile and his pure heartfelt curiosity. A true role model.
Today my wife
@ninaguilbeault
launches her timely and important book THE GOOD EATER w.
@BloomsburyPub
! It’s an incredibly thoughtful and open-minded analysis of our food system and how you, as a consumer and eater, can navigate its many trade-offs while staying aligned with your
Just finished teaching my class on power & politics, and I was so touched to receive this gift from my class. The students
@BerkeleyHaas
are not only wicked sharp, but also remarkably kind! Get ready business world. Some deep socially-conscious thinkers are coming in hot!
I’m really excited for my upcoming visit to
@USCAnnenberg
where I’ll speak about our network models of complex contagion and my ongoing extensions of this framework . Thanks
@MarlonTwyman2
for the invite!! Event details below:
There is too little conversation about what truth actually means in the study of “fake” news. Recipe for confusion and misleading narratives. Spoiler- turns out this is really hard, so no surprise it’s often glossed over or addressed by simply assuming common sense definitions
I'm thrilled that our recent paper on using collective intelligence to mitigate clinician biases has been selected by
@NatureComms
as the top article on their highlights page for the social sciences!
There’s a lot of talk about chatGPT making human coding obsolete. I’m finding the opposite. ChatGPT helps me to more quickly complete the laborious and uninteresting aspects of my code architecture so I can focus for longer on the more interesting analyses and coding problems.
Really happy to publish this article on bot policy with
@rgorwa
— we argue that ambiguous language in discussing bots is a major issue for regulation, and we provide a typology to facilitate clear policy action. Check it out!
@NDGannenberg
@AnnenbergPenn
Thrilled for Stanford's conference on Computational Sociology, Aug. 7. I'll be discussing my work on "Promoting Shared Understanding Across Cultures: Comparing U.S. and Chinese Populations in an Online Language Game." See all virtual talks here! Do join :)
We are hiring in the MORS group at
@BerkeleyHaas
! This is a dream job! Apply if you are a psychologist or social scientist interested in organizations. I am on the search committee and am happy to answer any questions. Full job description here:
Cool paper on group decision-making.
Group discussion results in more accurate numerical forecasts about the future, but less accurate votes on future decisions (like whether to launch a project). Voting pushes consensus to the majority opinion. Summary:
Great to see our recent paper in
@PNASNews
covered by
@TIME
magazine. Title of piece: "A Hive Mind of Doctors can Mean Better Care for Patients" I'd simply add -- "A *Networked* Hive Mind.." !
Confession: I spend most of my time on twitter combing for funny and informative comics/anecdotes that I can use when teaching machine learning in my people analytics class. This is one of my favorites in 2022.
🚨1/ Our newest compsyn pub. w. Int’l Conference on Computational Linguistics
#coling2020
: “Perceptually Grounded Word Embeddings with Color” Incl. color vectors for 40k most popular words in English, each w. crowdsourced ratings of concept concreteness
Very excited to present my work today - “Network Experiments on the Emergence of Shared Categories across Cultures” -
@KelloggSchool
@NICOatNU
Thank you
@hyejin_youn
for hosting! Can’t wait for the discussion to unfold.
Our new piece
@joshua_a_becker
@ned_smith
in
@HarvardBiz
“Some Questions Benefit from Group Discussion. Others Don’t.” Deliberation on yes/no questions amplifies majority opinion, regardless of accuracy. Forecasts improve even while decision gets worse.
It was a great pleasure to host Monica Lee – director of the Political and Organizational Science team at Meta – in our virtual fireside chat for my Power and Politics class last night. One key insight about content moderation that engrossed the class:
Excited to be in Princeton to meet with the International Network of Analytic Sociologists! Earlier this year I visited Paris during mass protests. Now I’m arriving to Newark amid apocalyptic forest fires. It’s been a cinematic conference circuit so far this year 😊😅
We’re honored to be mentioned in this series of fascinating papers! The feedback we received on this project
@IC2S2
years ago was really valuable for preparing our submission to
@Nature
. Excited to learn about all of the new and interesting research presented this year!
The book launch of THE GOOD EATER from my wife
@ninaguilbeault
was a smashing success! 100+ attendees, delicious healthy food, and a sold out book shelf! Thanks to all who came! For those who missed the SF event or can't get enough, her next event is at Books Inc in Berkeley on
What an incredible turnout of 100+ curious eaters for the sold-out launch of my book THE GOOD EATER
@welcometomannys
in SF last night! Very grateful to my friend chef Ashley for the incredible spring-inspired spread: 🌿
See you in Berkeley next week!
I’m excited to share my work on analogy formation in social networks with the Analogical Minds seminar this Thursday, June 2 at 11am. Details below if you want to join the conversation!
This Thursday at the Analogical Minds
@worldwideneuro
seminar – Dr
@DzGuilbeault
on cross-cultural similarities in category formation – Thurs June 2, 11 am in Chicago, 6 pm in Paris
Seminar link:
Mailing list:
I’m excited to be presenting at
@INSEAD
today on our project, “Online Gender Bias is Stronger in Images than Text.” Come rain or sunshine or fire-fueled protests 😊🙏🏻 Big thanks to the awesome OB group here for the invite!
I’m excited to present later today at the California Cognitive Science Conference
@UCBerkeley
on the role of social networks in the emergence of new categories. Looking forward to the discussion!
I won the Cheit Award for Excellence in Teaching, the highest honor for teaching
@BerkeleyHaas
! There is no such thing as humble bragging, so here is some plain bragging. A big thanks to my students for nominating me ❤️
Such sad news. It was an honor to study with Klaus, one of the great cyberneticians, with academic lineage tracing to Bateson and many others. A deep philosopher and scientist in equal proportions, and also a generous and kind teacher. He will be missed ❤️
It is with great sadness that we share the news that Professor Emeritus Klaus Krippendorff has passed away at the age of 90. A giant in the field, Klaus was a vital part of the Annenberg community for more than 50 years. A full obituary is forthcoming.
Got to be an early taste tester of
@wildtypefoods
’s lab “brewed” salmon with
@ninagheihman
. It was remarkably good and similar to natural fish! The future of food is here and it’s being led by
@BerkeleyHaas
students! As these products mature the game will be fundamentally changed
Wow, incredible synopsis of our work by
@BasHofstra
& Ann Mulders out in
@Nature
. They are exactly right that replicating our experiment in a global context will yield fascinating insights and that examining the implications for text-to-image AI models is an extremely exciting
Great to see interest in our recent
@plosone
piece. We find that classifying news as true/false limits social learning. A simple intervention – allowing ppl to express the probability that news is true – improved learning and reduced polarization
@joshua_a_becker
@samuelwoolley
I'm really excited for the Culture Conference today at
@BerkeleyHaas
! Check out this stellar line up of scholars, focusing on the empirical study of organizational culture and its effects
The comp-syn website is live! A volunteer team of artists and scientists, forged at the
@sfiscience
summer school. Follow here for innovations at the intersection of
#MachineLearning
, cultural poetics, color science, &
#digitalartwork
. Sample art here:
In physics, when a theoretical construct emerges from a simple model that does not include it by design – such as when
@stephen_wolfram
's hypergraph model of the universe reproduces Einstein's equations and thermodynamics – it is compelling evidence for the foundational status of
Thrilled to see our paper now out in MGMT SCI! See
@joshua_a_becker
’s excellent summary below. We reveal the following paradox in collective intelligence: deliberation can improve belief accuracy while simultaneously reducing decision accuracy!
🚨NOW OUT IN MGMT SCI🚨 THE CROWD CLASSIFICATION PROBLEM:
We know that deliberation improves *numeric forecast* accuracy. Here, we find that it can simultaneously reduce *decision* accuracy. A short thread w/ example (1/8)
w/
@DzGuilbeault
@ned_smith
I'm honored to be chosen as a grantee by
@BerkeleyHaas
's Center for Equity, Gender, and Leadership -- and to appear alongside such interesting folks and projects! Read more about my ongoing work on quantifying gender stereotypes in images & text here:
Despite hype around GPT3/4, I'm finding these models perform poorly at classifying metaphors in text. Using validation sets of metaphorical vs. literal word pairs, GPT performs worse than random. Anyone seen any work evaluating the ability of these models to identify metaphors?
I have a new short article out on the relationship between digital marketing, disinfo, and foreign interference. Check it out: Thank you all for the ongoing conversations!
@AnnenbergPenn
@polbots
So honored our recent article got the cover of PNAS! click to read more about how social network experiments can inform interventions on partisan bias in the interpretation of climate trends
@AnnenbergPenn
@joshua_a_becker
@NDGannenberg
I’m honored that our team comp-syn – started at the
@sfiscience
summer school – is a finalist for the Science + Technology + Arts (STARTs) prize for Social Good. We are thrilled to participate in this festival , and we eagerly await the big reveal!
I'm very excited to present in the Frontiers of Network Science colloquium tomorrow
@USCAnnenberg
! Details here I'll be discussing our work on modeling the spread of behaviors and ideas that require peer reinforcement (complex contagions) and their
I’m really excited for my upcoming visit to
@USCAnnenberg
where I’ll speak about our network models of complex contagion and my ongoing extensions of this framework . Thanks
@MarlonTwyman2
for the invite!! Event details below:
For an unrolled version of the thread I just released on our paper out today in
@NatureComms
with
@DamonCentola
, see here: Paper: "Topological Measures for Identifying and Predicting the Spread of Complex Contagions"
There should be a word for when you’re reading a used book and find yourself appreciating and identifying with the underlining emphases of one of its anonymous prior readers.
I think it's about time for computational social scientists to weigh in on the theory of emergence. I’m surprised that forefront conversations about emergence sidestep the topic of organizations, which def. exhibit top-down causality on individuals
I'm thrilled to appear in this volume with
@joshua_a_becker
@NDGannenberg
, alongside many interesting pieces on a critical topic for comp. soc science!
…
Links to
@arxiv_org
versions of almost all chapters here:
I wish that in the acknowledgments section of papers I could thank all the artists whose music I listened to while coding and writing. Perhaps there should be an extended acknowledgements section in the SI :)
Such a cool feeling when a scholar you’ve never met downloads your software and uses it to publish new work! This paper (link in reply) uses the compsyn python package to study the metaphorical color associations of the Chinese terms hēi ‘black’ and bái ‘white’ via google images
I'm honored to have co-authored with this brilliant fellow
@rgorwa
Check out his new book, which contains vital insights for the future of social media and online marketplace design.
My biggest project to date just landed in Berlin 🫣🤖
After many years of research, writing, and revision, I'm so excited to hear what you all think about 'The Politics of Platform Regulation: How Governments Shape Online Content Moderation.'
A short 🧵of summary and thanks:
Join us at this year's Culture Conference
@BerkeleyHCC
where bleeding edge conversations unfold about the nature of organizational culture, with voices from academia and industry. I have the distinct honor of being one of this year's keynote speakers!
Thrilled that our paper was accepted for publication
@aclmeeting
#ACL2022
, “Exploring Meaning Encoded in Random Character Sequences with Character-Aware Language Models,” led by
@markbochu
@bhargavvader
. Stay tuned for an upcoming paper release!
It was really nice speaking to Michele Berger about my very enjoyable experience defending my dissertation online. Numerous funny family stories emerged from the virtual event, and continue to make the rounds; couldn't be happier
@AnnenbergPenn
I’m really excited to be an invited speaker on collective intelligence at
#CI21
. I look forward to the collective effervescence! The deadline is soon approaching. Apply!!
We are excited to announce Douglas Guilbeault (
@DzGuilbeault
) as a confirmed speaker for the
#CI21
Conference! He is Assistant Professor in Management of Organizations at Berkeley.
Was great fun to participate in the
#ASA2021
panel on Computational Approaches to Rational Choice, org by
@emily_erikson
. Fantastic work by
@DustinStoltz
and others. With co-author Solene Delecourt, we presented "The Gender Gap in Online Images." Great feedback and interest!
We were delighted to see not only that Wikipedia wrote a piece covering our recent
@Nature
paper, but also that it led Wikipedia editors to include an image of a woman on the Wikipedia page for mechanic. A small but tangible impact!
@solenedelecourt
I’m looking forward to this Thursday morning, where I’ll be giving a public talk on “Category Formation in Social Networks” to Linkoping University's Analytic Sociology Institute Come fill your brains before you fill your bellies this Thanksgiving!
Excited to share our forthcoming paper on "The Crowd Classification Problem" in Mgmt Sci. Lead author
@joshua_a_becker
does an excellent job summarizing our results in the thread below. It was such a great time working with Josh and
@ned_smith
on this one. Hats off team!
🚨New Preprint!🚨 forthcoming in Mgmt Sci [] w/
@DzGuilbeault
@ned_smith
—we identify the CROWD CLASSIFICATION PROBLEM w/ estimate-based decisions, a paradox: group interaction improves estimate accuracy even while vote/decision gets less accurate (1/9)
Thrilled to have a chapter in this series along so many inspiring critical voices! Check out my piece on 'Human-bot Ecologies' when this powerhouse compilation is released! It bridges bots, theoretical neuroscience, and digital art, thanks to the hard work of my collaborators!!
Couldn't be prouder of my brother today. As a 1st gen university student, driven by a love of nature and hard work, he has published his first first-authored paper in an excellent journal today. Way to go
@ncguilbeault
! Check out his v cool tech here: