Johanna Teske

@johannateske.bsky.social

She/her/hers. Astronomer, runner, feminist, vegan, Nats fan, Peloton fan.

I cannot rest from discourse. I will see that film again tomorrow. I will drink life to the lees. Again I will enjoy greatly, will suffer greatly, both with those that love me, and in Pleasantville NY. I am become a meme. Film showing: burnsfilmcenter.org/booking/the-...

The Odyssey : Jacob Burns Film Center

The Odyssey, Matt Damon, Tom Holland, Christopher Nolan, Zendaya, Robert Pattinson, Odysseus, Greek, Classic, Antiquity

burnsfilmcenter.org

Nature did a short piece on our preprint about how LLMs will affect the practice of science; my more detailed thread is below.

Scientists using LLMs will ‘do more, less well’, modelling study predicts

Research suggests that the combination of incentives to publish and the use of large language models will lead to more papers, but they will be less refined.

nature.com

Carl T. Bergstrom@carlbergstrom.com · 2w ago

1. We—@eduede.bsky.social, @mjcrockett.bsky.social, Kevin Gross, and I—have a new preprint on the arXiv today, based on ideas that emerged during an @sfiscience.bsky.social workshop in November 2024: The unintended consequences of large language models as a labor-augmenting technology in science.

The countdown is on for the debut of the Nancy Grace Roman Space Telescope, which is named for NASA’s first chief astronomer. The Roman telescope is slated to launch on Aug. 30 aboard a SpaceX Falcon Heavy rocket from NASA’s Kennedy Space Center in Florida, per NASA. #NASARoman

Here's What The ‘Most Advanced Telescope Ever' Is Going to Reveal

The Nancy Grace Roman Telescope is expected to launch Aug. 30 and help unlock mysteries around dark matter, exoplanets and more.

today.com

I really enjoyed this collaboration: it goes to show that especially in the sciences, where data remains costly, Bayesian approaches like BO, BAL, and BED can make a difference. Also, I guess this is officially my first astrophysics paper, very exciting! 🔭🪐🌌

Forming Worlds Lab@formingworlds.space · last wk.

Harrison worked closely with our group all through his PhD, and still does. His new paper asks what you can learn about an exoplanet by fitting its whole life instead of a snapshot of today. 🔭 https://arxiv.org/abs/2607.25845 #Exoplanets #PlanetaryScience

New paper on arXiv today! On the efficacy of using novel Bayesian machine learning to dispatch 'expensive' forward models of planetary evolution. This connects with #PLATO ageing and can help solve degeneracies in current static-retrieval approaches which treat #exoplanets as snapshots.

Constraining the lives and times of exoplanets through evolutionary Bayesian retrievals

Static retrieval frameworks are leading tools for interpreting exoplanet observations, yet time-independent modelling leaves them prone to degeneracy and unable to resolve exoplanets' histories. The c...

arxiv.org

I had the honor of giving a keynote at the International Conference on Machine Learning last week. I addressed the widespread anxiety about how we should adapt as AI capabilities increase. I was thrilled by the talk’s reception, so I have made my slides available www.cs.princeton.edu/~arvindn/tal...

What will be left for us to work on?

ICML 2026 invited keynote — slides and edited transcript, presented click-by-click as delivered. Arvind Narayanan, Princeton University.

cs.princeton.edu

I'm consumed be "externalities" that private companies quietly distribute across humanity for profit while preaching narratives about "progress". Add in "Shifting Baseline Syndrome" and most people don't notice what's happening until it is too late. Some underlying verifiable facts 👇 ✨🔭🧪

Cartoon illustrating satellites crossing the night sky circa 2030, seen from a wilderness area. A child asks, "What is that point of light that doesn't move?"

Two cool things I want to highlight: i) no coronagraphs were used in the making of this discovery. Just the pure raw power of adaptive optics! ii) Look at that Airy ring around b 🤩 that shows we were working at the diffraction limit of the telescope which enabled us to get its best resolution!

beta Pic d had a good run at hide-and-seek, but now it’s found and its a great system for studying in even more detail. It tells us that we are far from done with even the most well-studied systems and that more planets are very likely hiding out there still. Count from 10 everyone, let’s get going!

Both teams discovered the planet serendipitously, and agreed not to share their work until publication so not to bias their interpretations, and got highly complementary results, one studying the atmosphere in detail, the other studying the planet orbit and its disk-shaping effects.

Aidan Gibbs at his team in the US had also seen d while looking at b with JWST! They got a spectrum which shows water, methane, and carbon monoxide in the planet atmosphere. Our observations also suggest there is carbon dioxide too.