Finding Balance and Joy in AI Marie desJardins ([email protected]) Associate Dean for Academic Affairs and Professor of Computer Science University of Maryland, Baltimore County AAAI-17 Workshop on Diversity in AI February 5, 2017 Outline A bit about me A bit about underrepresentation & diversity Balance & joy in academia (& industry too) Life in academia Handling failure Time management / Saying no AAAI-17 Workshop on Diversity in AI 2
Who Am I? Prof. Marie desJardins A.B. Engineering, Harvard 1985 Ph.D. Computer Science, Berkeley 1992 Research Scientist, SRI, 1991-2001 Professor at UMBC, 2001-now Tenured in 2007, full professor in 2011 Associate Dean, 2015-now Married since 1985, two daughters (sophomore in college & 2nd-year med student) Like to read, sing, play piano, ski, AAAI-17 Workshop on Diversity in AI 3 My Philosophy of Success What you love What the world wants What you
do well AAAI-17 Workshop on Diversity in AI A mentors job is to help our mentees find the intersection 4 Gender Underrepresentation in CS* In 2009, women earned: 50% of all Bachelors degrees in STEM areas 45% of Masters degrees in STEM 31% of Doctoral degrees in STEM But in 2014, women earned: 14% of Bachelors degrees in CS (up a tiny bit from the low of just over 11% in 2011... and nowhere close to the 37% peak of of 1984!) 22% of MS degrees in CS
18% of PhD degrees in CS Even compared to STEM fields, women are underrepresented in CS by a factor of nearly 2 at the grad level, and by a factor of more than 3 at the undergrad level! * Statistics for CS, CE, and IS combined Source: NSF / CRA Taulbee Survey AAAI-17 Workshop on Diversity in AI 5 Racial Underrepresentation in CS In 2008, of Bachelors degrees in CS: 4.9% went to African-Americans (9.8% of all Bachelors) 6.8% to Hispanics (7.9% of all Bachelors) In 2008, of Masters degrees in CS: 2.7% went to African-Americans (10% of all Masters) 2.4% went to Hispanics (5.9% of all Masters)
In 2008, of PhD degrees in CS: 1.6% went to African-Americans (6.1% of all PhDs) 1% to Hispanics (3.6% of all PhDs) Minorities are underrepresented by a factor of 4 at the grad level Source: CRA Taulbee Survey AAAI-17 Workshop on Diversity in AI 6 Causes of Underrepresentation Life goals differences: Women and minorities are more likely to be motivated by a desire to help people (not to make $$, hack, or make/play video games)
Negative perceptions of CS (hacker alone in a room) Chicken-and-egg problem: Hard to join a major where youre a minority Climate problem: Underrepresented students dont feel as welcome Misogynistic climate in many tech companies Confidence issues Performance perception: Women tend to underestimate their performance, relative to men Men are more likely to blame the situation Women are more likely to attribute poor performance to inherent lack of ability Hypercritical culture effects (cf. NSF panel 7 evaluations) AAAI-17 Workshop on Diversity in AI BALANCE & JOY IN ACADEMIA AAAI-17 Workshop on Diversity in AI 8 Responsibilities of a Professor Service
University/department committees Professional service (reviewing, conference organizing) Teaching Designing and delivering classes Supervising teaching staff Advising undergraduate students Research Getting funding Mentoring graduate students (and undergrads) Doing research, writing papers AAAI-17 Workshop on Diversity in AI 9 Useful Skills Computer Science Analysis/theory skills Software design/implementation skills Broad conceptual knowledge in the
subdiscipline Research Scientific thinking; analytical approach Ability to connect problems, solutions, and evaluation / hypothesis testing Teaching (and research) Good communication skills Patience and ability to listen Time Management Good organizational skills Ability to AAAI-17 multitask Workshop on Diversity in AI 10 A Typical Type I Day: Semester Underway
9:00 Put kids on school bus 9:30 Arrive on campus, check e-mail, find list 10:00 Talk briefly with 2 drop-in students 10:15 Make daily to-do list, check more e-mail 10:30 Class prep, review polytree inference
11:30 Finish journal article review, work through 20 backlogged e-mail messages 12:30 Faculty meeting 2:00 Class 3:30 Office hours, more e-mail, start annual report 4:30 Make notes on related work section for proposal, review colleagues draft proposal 5:00 Student meetings 6:00 Leave for home 6:30 Dinner, sort mail, clean kitchen, start laundry, put kids to bed 9:30 Start slides for next class, write next homework assignment AAAI-17 Workshop on Diversity in AI 11
A Typical Type II Day: Non-Teaching Day 9:00 Arrive on campus, check e-mail 9:30 Chat with dept. admin, ask about student paperwork snafu 10:00 Scheduling committee meeting
11:15 Student meetings 12:00 Lunch with friend, complain about being overworked, talk about joint proposal idea 1:30 Slog through e-mail backlog, read through solicitation, make notes about possible proposal 3:00 Think great thoughts about preferences over sets, read relevant papers, make notes about ideas 4:00 Department Hi Tea social hour 4:30 Skim recent journal issues 5:00 Write four student grad-school reference letters 5:45 Prepare grading key/notes for TA 6:15 Meet with TA, leave for home 7:00 Dinner, pay bills, put kids to bed 9:00
Make notes on 3 of 12 conference papers to be reviewed AAAI-17 Workshop on Diversity in AI 12 A Type III Day: Winter Break 9:30 10:00 10:45 11:15 12:00 12:30 to dept doing 1:30 2:00 on
Review 4:00 next Arrive on campus, check e-mail Review student draft conference paper Meet with student to review comments Travel arrangements for spring conference Student meeting Lunch, slog through e-mail, send RA hiring info admin, send e-mail to undergrads interested in research Student meeting Leave campus, go to Starbucks, work on laptop experimental code for ongoing research project. task force proposal for student review process Work on curriculum for new class to be taught semester, surf web for similar classes and ideas 5:30 Pick up kids, dinner, personal paperwork, put AAAI-17 Workshop on Diversity in AI kids to bed 13
These activities tend to get done (though they often make you insane) Categories of Tasks Urgent Non-Urgent Imp Paper deadlines orta Proposal deadlines nt These activities are necessary, but will take up ALL your time if you let them Les s Imp orta nt These activities tend to get short shrift, but are among the most critical for your career
Some e-mail Annual reports Committee work Reviewing(*) Lecture, HW, exam prep and grading(*) Most research activities Journal paper preparation Developing collaborations Relevant technical papers These activities are almost entirely timewasters and should be minimized Most e-mail Most meetings Informal chats with students and staff (*) Non-relevant technical papers, journals, articles AAAI-17 Workshop on Diversity in AI 14 Achieving Balance Lists,
lists, lists! The Art of the List Easy and hard tasks Long tasks and short tasks Long-term vs. short-term Prioritize Know what matters to you Work vs. family/personal Research vs. teaching vs. service Long-term vs. short-term Allocate sacred time
Avoid perfectionism Pay attention Find a mentor/sounding board AAAI-17 Workshop on Diversity in AI 15 Enemies of Efficiency and Effectiveness Mental clutter (and inbox clutter) Procrastination Overcommitment Indecision Poor prioritization AAAI-17 Workshop on Diversity in AI 16 Success Strategies Procrastinate Productively Idle Intelligently Write Right Away Decide Deliberately
Communicate Clearly AAAI-17 Workshop on Diversity in AI 17 Service: Should I Say Yes? No! well, OK, sometimes. Four simple rules: If nobody else in the room volunteers, dont break the silence!! Dont say yes right away When you say no (frequently), do it politely and graciously If you say yes, follow through and be on time Consider a no buddy or no mentor Best idea I learned: AAAI-17 Workshop on Diversity in AI Keep a list of everything you say no 18
A Tale of Two Women Dr. X: Dr. Y: Rejected by all but one school PhD from Top-Ten school Failed every comp exam Passed all; best grade on 9 of 10 proposals rejected one No papers in top journal $6M of sponsored funding Assoc. Editor of top Negative dept. vote on tenure Dozens of papers rejected journal Tenured full prof. 100+ published papers Failure or Success? You decide Theyre both me. AAAI-17 Workshop on Diversity in AI 19 Advice for Handling Failure Know that you're in good company Don't react to failure too quickly Learn to take useful/constructive criticism
Learn not to take harmful/destructive criticism personally Learn to tell useful from harmful criticism! Realize that other people have their own agendas/interests and aren't trying to hurt you personally Take care of yourself Know your goals, priorities, and values Lean on your friends AAAI-17 Workshop on Diversity in AI 20 Most importantly... ENJOY YOURSELF! (or whats the point??) AAAI-17 Workshop on Diversity in AI 21
