Pushing Boundaries: In Life, Science and Social Justice

Presented for the Loughborough Pride in STEM Research Showcase. I discuss my personal life and evolution as a queer person, my research in Machine Learning for Astronomy, and the role that Social Capital plays in Allyship and Activism.

Transcript

1. Pushing Boundaries: In Life, Science and Social Justice Dr Ashley Spindler (she/hers, @drashleynova) Dr James Geach, Dr Alex Hocking, Dr Yi Sun, Dr Neil Davey
2. The Agenda • Me, My Life and My Research • Machine Learning In Astronomy • Unsupervised Learning with SOMs • Donkey Kong, Good Allyship and Social Capital
3. Who Am I? • Astronomer and Data Scientist • MPhys and MSc at Surrey University • PhD at The Open University • Fellow at University of Hertfordshire • Online Content Creator • Twitch Streamer • Online Model • Author • Science Communicator, Activist and Trustee of Pride in STEM
4. Who Am I? I’m Queer AF
5. Machine Learning in Astronomy Deep Space, Deep Seas and Deep Learning
6. What is Machine Learning? Machines imitating and adapting human like behaviour
7. What is Machine Learning? ML finds natural patterns, gets insights and predicts the unknown
8. Unsupervised Learning Two Classes of Learning Supervised Learning
9. Supervised Learning Regression • Solving problems with continuous results • Weather prediction • Calculating redshifts from photometry • Predicting stock prices Finding patterns from input data and it’s outputs Classification Finding discrete categories for data • Galaxy morphological types • Natural Language Programming • Facial Recognition
10. Supervised Learning Example – Gravitational Lenses • Gravity bends light, acting like a lens to magnify distant galaxies • Finding arcs is something humans can do very easily, not so much for computers • Traditional arc finding techniques can have lots of false positives from edge on galaxies • Convolutional Neural Networks can identify lenses in very large data sets, such as the future Euclid and LSST data
11. Supervised Learning Example – Gravitational Lenses • CNNs are powerful tools that use many layers in a neural network • While powerful, they require large training datasets, and in astronomy that can mean relying on simulated data • Neural Networks with many layers are often black boxes, it can be difficult to visualise and analyse what is happening in the inner layers, and how the algorithm is reaching it’s output
12. Unsupervised Learning Clustering • Finding groupings of closely related data • Genetic expression clustering (Iris flowers) • Medical Imaging – differentiating different types of tissue is PET scans • Social networking – identifying groups of users Finding new patterns from just input data
13. Self Organising Maps Dimensionality reduction using a 2D grid of nodes
14. Finding Galaxies in the Hubble Ultra Deep Field Original Image Node Labelled Image 4x4 SOM 100 Epochs 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
15. Finding Galaxies in the Hubble Ultra Deep Field Original Image Node Labelled Image 4x4 SOM 100 Epochs 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
16. Finding Galaxies in the Hubble Ultra Deep Field
17. Industry • Agricultural Imaging • SONAR • Autonomous Vehicles • Medical Imaging ASTROSENSE: From Astronomy to Remote Sensing Astronomy • Classifying Galaxy Morphology • Segregating Molecular Clouds • Finding Galaxy Clusters • Measuring Photometric Redshifts
18. How to be an Ally… Featuring Donkey Kong
19. I always get asked “how do I be a good ally?” I answer with things like: Use your Platform Give Queer People a Voice Be Vocal in your Support
20. I always get asked “how do I be a good ally?” Being a good ally is about using your SOCIAL CAPITAL to advance LGBT/PoC/Disability/ect. causes
21. What is Social Capital? Social Capital is the Resources available to a person because of the Networks and Relationships they have built, and the Value of those resources in the impact they have on Forming and Maintaining relationships
22. What is Social Capital? People with High Social Capital are able to influence their networks and build new networks with new people. Attention is the Currency of social capital in the Marketplace of Ideas. You can spend your social capital to gain attention, and vice versa.
23. Social Capital is a Trap Much like financial capital, the platforms we use (twitter, facebook, youtube) push us to accumulate and hoard social capital. More social capital means more attention, more attention means more power and the ability to generate financial capital. People use social capital to protect themselves from criticism.
24. Hbomberguy, Donkey Kong and Using Social Capital for Good • Harris Brewis, aka Hbomberguy, set out to raise $3000 for Mermaids, a charity that supports trans kids and their families, while playing Donkey Kong 64. • Thanks to Hbomberguy’s social capital, and the tireless work of internet sloth Casey Explosion, dozens of famous trans and cis people lent their voices to the cause and raised $340,000 • Voices included Chelsea Manning, Paris Lees, Owl, Natalie Wynn, Lily Madigan, Laura Kate Dale, Alexandria Ocasio- Cortez and Donkey Kong’s VA Grant Kirkhope
25. Hbomberguy, Donkey Kong and Using Social Capital For Good • Social Capital can be a force for good if it is shared and democratized. • The stream was a massive success because it allowed marginalized trans voices the opportunity to use a much larger platform than they normally could. • Allies used their social capital to push attention towards the issue of trans rights, and raise money for a charity facing extreme backlash by right wing media.
26. Nothing you do is perfect – But it doesn’t have to be • Despite it’s success, Hbomberguy’s stream was not without flaws. In the chaos of bringing on voices, many people fell through the gaps, including non-binary and trans PoC. • It is vitally important that we don’t allow those with high social capital to avoid criticism, but we also shouldn’t shut people out for mistakes. • What is most important is that we learn from those mistakes.
27. The End Game? When we think about Social Capital, platforms, allyship and LGBT+ activism, we have to ask: What Do We Want? How Can We Be Different? How Can We Measure It?
28. The End Game? How can we restructure Academia and STEM in general to avoid the Capitalistic Traps of Social Capital? How is Publishing, Funding and Collaboration affected by the mechanisms of Social Capital?
29. The End Game? We have a Responsibility to pay keen attention to the communities we want to create. - Angie Speaks, 2019
30. Any Questions?