Artificial Intelligence and Diversity for All
Download as PPTX, PDF0 likes110 views
The document discusses the intersection of artificial intelligence (AI), software engineering, and gender diversity, emphasizing the role of ACM in advancing computing professionals. It highlights historical developments in AI, the implications of biases in software and workforce, and various initiatives aimed at promoting inclusivity in the tech industry. Several sessions are outlined, focusing on AI's potential benefits and challenges for women, alongside discussion questions for participants to reflect on their learning and actions.
Artificial Intelligence and Diversity for All
- 1. AI and Diversity for All Letizia Jaccheri
- 2. The Distinguished Speakers Program is made possible by For additional information, please visit http://dsp.acm.org/
- 3. About ACM ACM, the Association for Computing Machinery (www.acm.org), is the premier global community of computing professionals and students with nearly 100,000 members in more than 170 countries interacting with more than 2 million computing professionals worldwide. OUR MISSION: We help computing professionals to be their best and most creative. We connect them to their peers, to what the latest developments, and inspire them to advance the profession and make a positive impact on society. OUR VISION: We see a world where computing helps solve tomorrow’s problems – where we use our knowledge and skills to advance the computing profession and make a positive social impact throughout the world. I am proud to be an ACM Member.
- 4. Session 1 – AI – inspiration books and videos Session 2 – AI - history and terminology Session 3 – Software Engineering and diversity Session 4 – AI for or against all?
- 7. Literature • Female roles in literature • Nora (Ibsen) • Literature for women • women's magazines, Disney (Pochahontas, The Little Mermaid) • Women who create literature • Jane Austen, Sigrid Undset, Elena Ferrante
- 8. AI Female roles in AI • Siri voice • Avatar AI for/against women • For – Menstruation Apps, Designing Software to Prevent Child Marriage Globally, Tappetina • Against - automatic processing of CVs Women creating AI • Fei Fei Li • Francesca Rossi
- 10. Questions • Which of these books do you relate to? • Which ones make you want to read/listen? Letiziajaccheri.org
- 11. Session 1 – AI – inspiration books and videos Session 2 – AI - history and terminology Session 3 – Software Engineering and diversity Session 4 – AI for or against all?
- 12. Definitions • AI is a field of study (and research field) within computer science that develops and studies intelligent machines • AI stands for a computer system that performs tasks that typically require human intelligence, such as recognizing speech, making decisions and identifying patterns • Generatively create new content (sound, code, images, text, video) • Machine learning is part of AI
- 13. History • Artificial intelligence was founded as an academic discipline in 1956 • 1950 - 60 first AI programs • 1970 expert systems • 1980 neural networks • 1990 autonomous robots • 2,000 self-driving cars • 2010 AI-powered assistants • 2020 Advanced AI in healthcare, finance, transport, art
- 14. Why now? • Hardware – software – data GPT – Generativ Pretrained Transformer autumn 2022
- 15. hardware From Kilobyte til Petabyte 1015
- 16. software
- 17. CHAT GPT 4 has been trained on almost all text ever written data 1013
- 18. IT system Humans Develop, test, use AI system Generative AI Use tools
- 20. - GPT-3 has an estimated training time of 355-GPU-years and an estimated training cost of $4.6 million. - If we trained GPT-3 at IDUN, it would take 355/36 = 10 years
- 21. Discussion questions • What new words have you learned? • What are your questions about AI?
- 22. Session 1 – AI – inspiration books and videos Session 2 – AI - history and terminology Session 3 – Software Engineering and diversity Session 4 – AI for or against all?
- 23. Software Engineering Gender Analysis and Design | Empirical software engineering | Software quality | Architecture | Processes | AI and SE | Human factors in SE Gender and sex | Non-binary | LGBT+ rights | #metoo 2017 | Same-sex marriage 2001 | Intersectionality – triply | feminism
- 25. Amazon created a recruitment tool that proved to be discriminating against women specifically J. Dastin, “Amazon scraps secret AI recruiting tool that showed bias against women,” in Ethics of data and analytics, Auerbach Publications, 2022, pp. 296–299.
- 26. Facebook’s job advertisement algorithm reached out to specific users based on their race, gender, and religion. Moreover, women were presented with stereotypical feminine jobs, such as secretaries or nurses. Such algorithms enhance sexism and racist attitudes in the labor environment. M. Ali, P. Sapiezynski, M. Bogen, A. Korolova, A. Mislove, and A. Rieke, “Discrimination through optimization: How facebook’s ad delivery can lead to biased outcomes,” Proceedings of the ACM on human-computer interaction, vol. 3, no. CSCW, pp. 1–30, 2019.
- 27. 20% 29% Female ICT students in 2021
- 29. Norwegian and European best Practices • ADA • IDUN • EUGAIN • Horizon CRAFT • Erasmus + Women Stem Up • ACM WomENcourage • Abelia Tech Kvinner
- 30. Burnett, M., Stumpf, S., Macbeth, J., Makri, S., Beckwith, L., Kwan, I., Peters, A. and Jernigan, W., 2016. GenderMag: A method for evaluating software's gender inclusiveness. Interacting with computers, 28(6), pp.760-787.
- 31. G. Catolino, F. Palomba, D. A. Tamburri, A. Serebrenik and F. Ferrucci, "Gender Diversity and Women in Software Teams: How Do They Affect Community Smells?," 2019 IEEE/ACM 41st International Conference on Software Engineering: Software Engineering in Society (ICSE-SEIS), Montreal, QC, Canada, 2019, pp. 11-20
- 32. RQ How do biases in the workforce impact biases in software? IT system workforce Develop, test, use AI system Generative AI Use tools
- 33. RQ How do biases in the workforce impact biases in software? • Y. Wang and D. Redmiles, “Implicit gender biases in professional software development: An empirical study,” in 2019 IEEE/ACM 41st International Conference on Software Engineering: Software Engineering in Society (ICSE-SEIS), 2019, pp. 1–10.
- 34. RQ How do biases in the workforce impact biases in software? A. Hannak, G. Soeller, D. Lazer, A. Mislove, and C. Wilson, “Measuring price discrimination and steering on e-commerce web sites,” in Proceedings of the 2014 conference on internet measurement conference, 2014, pp. 305–318.
- 35. Session 1 – AI – inspiration books and videos Session 2 – AI - history and terminology Session 3 – Software Engineering and diversity Session 4 – AI for or against all?
- 40. Hvorfor? IT system Mennesker lager, tester, bruker KI system Generativ-KI bruke verktøyene 50% 5%
- 41. Threats • False statements, false faces, false messages. There will be more of all this. • Old systems, old stereotypes are magnified - if we don't take action • Automatic processing of CVs • The training data
- 42. AI for all • We cannot change old networks, we can make new ones around AI • AI for women https://irthapp.com/
- 43. 1.8.2024
- 44. Discussion questions • What can I do? • What do I want to do?
- 45. TDT4290 Customer Driven Project • https://tinyurl.com/2x5y5mnk • Customer defines the project – The teaching team, the students learn together with the customer • 2023 • Artificial Intelligence • Sustainability • Gender and Diversity Thale Kuvås Solberg (Q-Free) ACM womENcourage 2023
Editor's Notes
- #1: 3 deler Intro Om KI Om Kvinner og Informatikk/Data/IT
- #5: https://www.unwomen.org/sites/default/files/2022-01/Intersectionality-resource-guide-and-toolkit-en.pdf Intersectionality Resource Guide and Tool Kit Interseksjonalitet er et analytisk rammeverk som brukes for å forstå hvordan ulike sosiale identitetskategorier, som kjønn, rase, klasse, seksualitet, funksjonshemming og mer, samvirker på komplekse måter og skaper unike opplevelser av diskriminering eller privilegier for individer. Begrepet ble først introdusert av den amerikanske juristen og akademikeren Kimberlé Crenshaw på slutten av 1980-tallet. Hun ønsket å forklare hvordan svarte kvinner opplevde diskriminering på en måte som var forskjellig fra både hvite kvinner og svarte menn, fordi de sto overfor både kjønns- og rasediskriminering samtidig. Interseksjonalitet ser derfor på hvordan ulike former for makt og undertrykkelse overlapper og påvirker hverandre. For eksempel kan en svart, funksjonshemmet, lesbisk kvinne oppleve diskriminering på grunn av alle disse identitetsaspektene samtidig, noe som gir henne en annen opplevelse enn noen som bare opplever diskriminering basert på én av disse identitetskategoriene. Interseksjonalitet er viktig for å forstå at sosiale problemer og urettferdighet ikke kan løses ved å se på hver identitetskategori isolert. Det krever en helhetlig tilnærming som tar hensyn til de ulike lagene av identitet og hvordan de påvirker hverandre.
- #6: Video https://www.youtube.com/watch?v=hg3umXU_qWc&t=648s 8 milioner ganger Den feministiske klassikeren "Et eget rom" er en nøkkel til Virginia Woolfs forfatterskap. Essayet er basert på en forelesning om kvinner og litteratur holdt i Cambridge i 1928.
- #7: Kvinner er et eksempel The Top 10 women in the world of AI in 2023 https://aimagazine.com/top10/the-top-10-women-in-the-world-of-ai-in-2023 Paper https://dl.acm.org/doi/abs/10.1145/3311927.3325322
- #8: Kvinner er et eksempel The Top 10 women in the world of AI in 2023 https://aimagazine.com/top10/the-top-10-women-in-the-world-of-ai-in-2023 Paper https://dl.acm.org/doi/abs/10.1145/3311927.3325322
- #13: The 1956 Dartmouth workshop was the moment that AI gained its name URL The history of AI 1950s: The first AI programs were written to run on the Ferranti Mark 1 machine of the University of Manchester: a checkers-playing program written by Christopher Strachey and a chess-playing program written by Dietrich Prinz. 1960s: The Dartmouth Conference was organized by John McCarthy, Marvin Minsky, Nathaniel Rochester, and Claude Shannon to discuss the possibility of thinking machines and artificial intelligence. 1970s: The first expert systems were developed. 1980s: The first neural networks were developed. 1990s: The first autonomous robots were developed. 2000s: The first self-driving cars were developed.google translator came in 2006 based on statistical methods. 2010s: The first AI-powered virtual assistants were developed. Deep learning revolution, AlexNet 2012 2020s: AI continues to advance and is being used in a wide range of applications, from healthcare to finance to transportation 1.
- #15: Det er en stor mistforståelse at man trenger å være veldig intelligent og helst være en mann for å forstå datamaskiner, algoritmer og KI. Jeg har holdt på med data siden tidlig 80 taller, det var på en mate en tilfellighet. Og jeg mener det er like vanskelig å fotså strikkeoppskrift som det er å forstå algoritmer. Vi trenger fortiden for å ha et språk for å snakke om nåtiden og fortiden K = 1000 Giga = 1000000000 104 = 10000 1017 = 100.000.000.000.000.000 Minne 36 Kbyte Hastighet 104 flops Floating point operation per second – nå 8 Giga – now 1017
- #16: Margaret Hamilton (som jeg har møtt en gang!), den første personen som laget begrepet software engineering og som skrev programvaren for Apollo-ekspedisjonen på 60-tallet er en kvinne (Apollo første mann på månen) Margaret Hamilton (som jeg har møtt en gang!), den første personen som laget begrepet software engineering og som skrev programvaren for Apollo-ekspedisjonen på 60-tallet er en kvinne Apollo 145.000 lines of code Now millions of lines of code
- #17: According to https://www.youtube.com/watch?v=_6R7Ym6Vy_I&t=2220s All human written text = 1013 CHAT GPT 4 = 300 x 1012
- #18: CS started in early 60’s. the subfields are inter-related and discoveries and innovations in one field can bring challenges, discoveries and innovations in other field. My own field software engineering is very inter related with AI as we use AI tools to develop software AND we use software engineering techniques and processes to develop AI systems.
- #19: NodeType#CPUsProcessor#CoresRAM[GB]#GPUsGPU type How can we compare to the system on which Open AI trains its models to give an idea
- #20: NodeType#CPUsProcessor#CoresRAM[GB]#GPUsGPU type How can we compare to the system on which Open AI trains its models to give an idea https://lambdalabs.com/blog/demystifying-gpt-3 CHAT GPT4 – it costed 1 M Dollars to pre train According to https://www.youtube.com/watch?v=_6R7Ym6Vy_I&t=2220s https://www.hpc.ntnu.no/idun/
- #23: Quality of the process and quality of the software How do we understand gender and SE? gender as a social construction (as opposed to sex, which is by birth) (it: genere, sesso – no: gender as a non-binary term (as opposed to what it used to be: binary male vs female) - recent steps (correlated) in the women's rights movement #metoo started in 2017 as well as the LGBTQ rights movement, while the first gay movement started more than 100 years ago, The Netherlands was the first country to allow same-sex marriage in 2001. defining one's pronouns has become trendy (she/her) non-binary acceptance of gender intersectionality, that is, that our identities have several layers and they influence one another --> people who are not only women, but also LGBTQ and of color, are triply underprivileged and left out of rights not only in everyday life but in digital services as well. gender equality charts in the EU or over the world and Gender Gap Index education + social stereotypes --> girls' barriers to get into IT and CS due to women's lack, and the lack of diversity, in IT teams --> bias in designing products due to a lack of diversity, for ex. HCI, AI, etc. this creates usability problems, tech issues, discrimination, the spread of bias, etc. It is crucial to underline that taking gender into account in tech is not only a human rights issue, but an economic and technological need.
- #24: The software engineering (SE) community has focused on the gender gap that is indicative of broader societal biases but is only one dimension of the complex system of inequalities. https://www.unwomen.org/sites/default/files/2022-01/Intersectionality-resource-guide-and-toolkit-en.pdf Intersectionality Resource Guide and Tool Kit Interseksjonalitet er et analytisk rammeverk som brukes for å forstå hvordan ulike sosiale identitetskategorier, som kjønn, rase, klasse, seksualitet, funksjonshemming og mer, samvirker på komplekse måter og skaper unike opplevelser av diskriminering eller privilegier for individer. Begrepet ble først introdusert av den amerikanske juristen og akademikeren Kimberlé Crenshaw på slutten av 1980-tallet. Hun ønsket å forklare hvordan svarte kvinner opplevde diskriminering på en måte som var forskjellig fra både hvite kvinner og svarte menn, fordi de sto overfor både kjønns- og rasediskriminering samtidig. Interseksjonalitet ser derfor på hvordan ulike former for makt og undertrykkelse overlapper og påvirker hverandre. For eksempel kan en svart, funksjonshemmet, lesbisk kvinne oppleve diskriminering på grunn av alle disse identitetsaspektene samtidig, noe som gir henne en annen opplevelse enn noen som bare opplever diskriminering basert på én av disse identitetskategoriene. Interseksjonalitet er viktig for å forstå at sosiale problemer og urettferdighet ikke kan løses ved å se på hver identitetskategori isolert. Det krever en helhetlig tilnærming som tar hensyn til de ulike lagene av identitet og hvordan de påvirker hverandre.
- #27: Numbers for bachelor master phd – in Norway we are doing a bit better 29%
- #28: Examples from ADA Calling the applicants, Welcome day, March 8 - Women's Day, Network lunches, Programming courses, social activities such as Mountain hiking, PhD party, Invite high school girls from all over the country, Presentations and workshops, Personal meeting with role models, Break down stereotypes and bias, hire women role models This is answer to question about what works Explain briefly what these projects have done and achieved since 1997 Measures such as https://www.ntnu.no/ada https://www.ntnu.edu/idun Eugain.eu https://www.ntnu.edu/smartcities/craft https://liu.se/en/news-item/erasmus-projektet-women-stem-up-tilldelas-4-1-miljoner-kronor https://womencourage.acm.org/2023/ https://www.abelia.no/50techkvinner/
- #29: But why? What are the challenges? And in our projects we are trying to combat these challenges, Examples from EUGAIN Calling the applicants, Welcome day, March 8 - Women's Day, Network lunches, Programming courses, social activities such as Mountain hiking, PhD party, Invite high school girls from all over the country, Presentations and workshops, Personal meeting with role models, Break down stereotypes and bias, hire women role models This is answer to question about what works Explain briefly what these projects have done and achieved since 1997 Measures such as Information Mentoring Network Anti bias training
- #30: M. Burnett et al., "GenderMag: A method for evaluating software’s gender inclusiveness", Interact. Comput., vol. 28, no. 6, pp. 760-787, 2016. It enables software practitioners (e.g., developers, managers, UX professionals) find gender-inclusivity "bugs" in their software, and then fix the bugs they find. M. Vorvoreanu, L. Zhang, Y. Huang, C. Hilderbrand, Z. Steine-Hanson, and M. Burnett. 2019. From Gender Biases to Gender-Inclusive Design: An Empirical Investigation. In Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems (CHI '19). Association for Computing Machinery, New York, NY, USA, Paper 53, 1–14. https://doi.org/10.1145/3290605.3300283
- #31: Examples of community smells are Organizational Silo effects (overly disconnected sub-groups) or Lone Wolves (defiant community members). Why should we care for communication? https://www.youtube.com/watch?v=d7TGWOaeujU Gender Diversity and Inclusion and Software Engineering (team, process) one developer can decide to work in isolation
- #33: Measuring price discrimination and steering on e-commerce websites typically involves detecting whether different users receive different prices or product recommendations based on factors such as location, browsing behavior, or device type. Researchers use several methods to assess this: Web Scraping: Automated tools or scripts gather pricing information from e-commerce websites at various times, from different locations, or using different devices (e.g., mobile vs. desktop). This helps to compare whether prices differ across these variables. A/B Testing: Creating different user profiles with varying characteristics (such as IP addresses from different geographic regions) allows researchers to compare the prices shown to different profiles and detect any patterns of discrimination or price steering. Cookie and Tracking Analysis: E-commerce websites often use cookies to track user behavior. By analyzing how prices change when cookies are deleted or a user browses in "incognito" mode, it becomes possible to measure the effect of tracking on pricing or steering. Controlled Experiments: Simulating various user interactions with the site (such as searching for certain products or clicking on specific ads) can reveal whether the site adjusts its pricing or product recommendations based on past behavior. Data Collection and Statistical Analysis: After gathering data on prices shown to various users, statistical analysis is used to determine whether price discrimination or steering exists. This involves comparing average prices and product recommendations across different user groups or conditions.
- #38: Jeg spurte https://openart.ai/create å tegne 4 software engineers og jeg fikk dette
- #39: Some numbers about ICT specialists in Europe and World and who are the ICT specialists In 2022, 9.4 million people in the EU worked as ICT specialists In the world 55.3 million in 2020 3 out of 4 companies have problems finding specialists with the right skills And how many ICT specialists are women? 2% in 10 years How many years will it take if we continue like this? eurostat https://ec.europa.eu/eurostat/statistics-explained/index.php?title=ICT_specialists_in_employment In 2023, 80.6 % of men were employed as ICT specialists in the EU against 19.4 % of women.
- #40: https://www.statista.com/statistics/1126823/worldwide-developer-gender/#:~:text=According%20to%20a%20global%20software,reality%20of%20software%20development%20jobs.
- #41: Making models – using AI Gender is only a dimension
- #45: Risiko nivå Vi har allerede regler som regulerer diskriminering i Norge og i Europa
- #47: What is se? Industry is our lab Allocate time to meet the group and ensure that the students know what to do Make available necessary information/documentation to the group Allocate time to read phase documents and provide timely feedback Allocate time to answer questions within reasonable time Make available special software or hardware Show up at NTNU for the final presentation of project results