Do Androids Dream of Exploding Stars?

University of Delaware

Department of Physics and Astronomy

Biden School of Public Policy and Administration

Data Science Institute

federica b. bianco

she/her

Why do we study stellar explosions?

we are made of stars

The nitrogen in our DNA, the calcium in our teeth, the iron in our blood, the carbon in our apple pies were made in the interiors of collapsing stars.

We are made of starstuff.

― Carl Sagan, Cosmos

largest explosion on earth 10,000,000 erg

typical supernova....

Why do we study stellar explosions?

a unique opportunity to study extreme energy events

A new, transformational observatory is about to start building a legacy for humanity

before human-made satellites forever change it

10 stars explode in the universe every second

Until the 1900s we would see 1 in a century

Until the 1980s we would see 1 in a decade

Until the 2010s we would see 1 in a month

Starting in 2026, we will see 1,000 every night !

Building a legacy:

The Vera C. Rubin Observatory LSST

the Vera C. Rubin observatory will collect

20Tb of data every night. That is equivalent to

8,000 high definition movies

4,000 hours of tiktok videos

every night for 10 years

Building a legacy:

The Vera C. Rubin Observatory LSST

the Vera C. Rubin observatory will collect

20Tb of data every night. That is equivalent to

8,000 high definition movies

4,000 hours of tiktok videos

every night for 10 years

Building a legacy:

The Vera C. Rubin Observatory LSST

the Vera C. Rubin observatory will collect

20Tb of data every night.

A legacy dataset that belongs to all people in the USA giving access to never before seen corners of the Universe to all

A new, transformational observatory is about to start building a legacy for humanity

before human-made satellites forever change it

We developed new technology

We built the largest camera ever built for the

Vear C. Rubin Observatory

to look farther and wider into the sky than ever before

Eye to the sky…on-sky engineering tests have begun at @nsfgov–@energy Rubin Observatory using the world’s largest digital camera!🔭

April 17

skyviewer.app

In its first year of operation LSST will collect more data than the entire collection of astrophysical data that exists today....

cataloging more objects than there are people on earth

there are not enough people in the world to inspect each image that LSST would collect at full resolution!

How do we study stellar explosions?

with this much data we need Artificial Intelligence

Telescope scheduling is automated

1,000 images per night

processed in 60 seconds to detect all changing and moving objects

image processing is automated

Rubin Observatory Data Management Team

federica bianco - fbianco@udel.edu

Discovery Engine

10M alerts/night

Community Brokers

target observation managers

the astronomy discovery chain

Pitt-Google

Broker

BABAMUL

So have AI taken our jobs??

Does it mean that AI

has taken our jobs?

physics

Input (observables)

Output

(observable)

f(x)

Input

f(x) = mx + b

m: slope

b: intercept

Machine Learning

Prediction

f(x)

Input

f(x) = mx + b

m: slope

b: intercept

parameters

learn

goal: find the right m and b that turn x into y

Machine Learning

Prediction

f(x)

https://symposia.obs.carnegiescience.edu/series/symposium2/ms/freedman.ps.gz

Input

prediction

physics

non-linear modification to a linear function

f(x) = \sigma( mx + b)

f(x)

p(class)

pixel values tensor

f(x)

Plan survey → Get data → Scientific Idea → build AI → process → Interpret results

Science has always been a sociotechnical proces

We need examples to train the AI

Why do we study the night sky?

NASA - Hubble Legacy Field Zoom-Out

Two things fill the mind with ever new and increasing admiration and awe, the more often and steadily we reflect upon them: the starry heavens above me and the moral law within me.

I see them before me and connect them immediately with the consciousness of my existence.

-Emmanuel Kant

Say their names

https://baptistnews.com/article/we-do-not-know-all-the-names/

Training a single large model like a ResNet-50 on a standard GPU (e.g., NVIDIA V100) for a few epochs on ImageNet (1.2M images) can emit roughly ~100-150 kg CO₂or the CO₂ equivalent of a short flight.

But how many models are you training in development?

ISN'T FREE

a 2 kg computer requires 800 kg of raw materials. As well, the microchips that power AI need rare earth elements, which are often mined in environmentally destructive ways.

data centres produce electronic waste, which often contains hazardous substances, like mercury and lead.

data centres use water during construction and, once operational, to cool electrical components. Globally, AI-related infrastructure may soon consume six times more water than Denmark, a country of 6 million when a quarter of humanity already lacks access to clean water and sanitation.

to power their complex electronics, data centres that host AI technology need a lot of energy, which in most places still comes from the burning of fossil fuels, producing planet-warming greenhouse gases.

Say their names

Renee Good, 1/7/26, Minneapolis, MN

https://baptistnews.com/article/we-do-not-know-all-the-names/

Say their names

Renee Good, 1/7/26, Minneapolis, MN
Parady La, 46, a Cambodian national, died on January 9 at the Federal Detention Center in Philadelphia
Luis Beltran Yanez-Cruz, 68, from Honduras, died on January 6 in Indio, Ca\
Luis Gustavo Nunez Caceres, 42, from Honduras, died on January 5 in Houston
Geraldo Lunas Campos, 55, a Cuban national, died on January 3 at Camp East Montana in Texas, where he was placed in isolation after becoming disruptive; his death is under investigation.

https://baptistnews.com/article/we-do-not-know-all-the-names/

832 days of Palestine genocide since October 7, 2023

1,418 days since the start of Russia’s full-scale invasion of Ukraine on February 24, 2022

over 100 days of FRC confirmed famine sincein August 2024. The crisis has since spread, with famine declared in Zamzam IDP camp (North Darfur), El Fasher (North Darfur) and Kadugli (South Kordofan), and at least 20 additional areas at risk.