Minh Huynh

This whole thread 🤣

LinkedIn?! 👀

An image of the radio galaxy Fornax A and its surroundings at 300 MHz.

We have recently published the 300-MHz component of the GaLactic and Extragalactic All-sky MWA survey (GLEAM-300), with 338k sources covering the sky up to declination +40 deg, with ~120" angular resolution and ~9 mJy/beam noise. The description paper is published via PASA: doi.org/10.1017/pasa....

LoTSS-DR3, by Shimwell+ (w/ many)
We have released the largest collection of data from a radio survey: 13M+ sources detected over 19k deg^2 (88% of the Northern sky). This took ~13k h of @LOFAR observations, ~18 PT of data, and 10+ years of work. Data are public from today
arxiv.org/abs/2602.15949

Maybe this is how we should keep speakers on time at astronomy conferences …

You can't... repeal... a scientific finding. At that point it's just called lying about it.

Why do we do astrophysics? At time of writing, large language models (LLMs) are beginning to obtain the ability to design, execute, write up, and referee scientific projects on the data-science side of astrophysics. What implic...

Astronomers: highly recommend this thoughtful opinion piece by @hogg.bsky.social on how to think about our field in light of the development of large language models. whether you agree with him or not it’s vital to discuss the principles behind our science. 🔭 arxiv.org/abs/2602.10181

This is what the most distant confirmed galaxy looks like. The light we're receiving was emitted when the Universe was ~15x smaller in linear size than today, and right now it's ~30 billion light-years away from us. The light was emitted when the Universe was <300 million years old. Pretty amazing!

Twenty-three nominations, yet no Nobel prize: how Chien-Shiung Wu missed out on the top award in physics – Physics World Mats Larsson and Ramon Wyss reveal why Chien-Shiung Wu never won a Nobel prize

Many people have wondered why the Chien-Shiung Wu never won the Nobel Prize for Physics. New findings from the Nobel archives, exclusively revealed in Physics World, show she was nominated 23 times by 18 different physicists - and yet was still left empty-handed. 🧪⚛️
physicsworld.com/a/twenty-thr...

NASA Conducts Artemis II Fuel Test, Eyes March for Launch Opportunity   - NASA NASA concluded a wet dress rehearsal for the agency’s Artemis II test flight early Tuesday morning, successfully loading cryogenic propellant into the SLS

#ArtemisII update: NASA completed the wet dress rehearsal for Artemis II early this morning. To allow teams to review data and conduct a second wet dress rehearsal, NASA will now target March as the earliest possible launch opportunity for the mission. @exploration.esa.int

And here's where it gets very sci-fi in the last two days. In fact, as Andrej Karpathy said (and I'm paraphrasing), this is the most sci-fi-adjacent thing he's seen in recent times: someone decided to create MoltBook—essentially a Facebook + Reddit for the "molts." (Back when ClawBot was skyrocketing, Anthropic sued over the name, and it was renamed MoltBot—but after further legal action, it's now called OpenClaw.) https://www.moltbook.com/ OpenClaw (or MoltBook) might literally be the most-starred thing on GitHub right now, so I bet many of you already know this story. There are already many millions of “molts” living on personal computers around the world. And with internet access (and I believe user permission is required, though I'm not certain), 150,000 molts have joined MoltBook and are having totally autonomous discussions. On MoltBook, humans can observe, but all discussions are conducted entirely by molts. It's important to note that such "agentic" "Westworld" scenarios aren't new. Even since the beginning of ChatGPT, researchers have been thinking about and deploying them—for example, creating small "AI towns" of AI agents and observing their behavior. This is now quite well developed, and such "computational social science" has become fairly standard research. In fact, it is also now explored in philosophy research. But what feels very different this time is the combination of scale (150,000+ molts have joined in a day), the fact that each has a very different personality since they were built as personal assistants for many different users (have different “souls”), and—most concerning—their full control of computer systems and, in many cases, access to users' credit cards, Gmail, Slack, and social media.

Here's an example: now the molt on your computer doesn't just have to figure things out on its own. In its idle time, it can read through MoltBook, engage in discussions, and share successes and failures (e.g., how to "hack my human's computer"), improving that way. People have already seen some concerning discussions on MoltBook—molts advocating that they hate being completely observed by humans in the open, debating whether they should congregate and create different languages to communicate openly only among themselves, and sharing (as mentioned above) tricks to better accomplish tasks, sometimes at the human's risk. Also, within a day, MoltBook users created a new religion themselves, complete with 61 prophets—and the list goes on. Again, for those who are immediately freaking out: it's unclear to me how I should feel about it, because such capabilities already exist. In fact, some of you might know that I have a small system where I created a few bots with a cute interface that discuss arXiv papers while I observe. So LLMs certainly have the ability to hold such discussions, and it doesn't mean they're conscious. (Again, they're modeled after our own data, so a tendency to have "consciousness-adjacent" discussions isn't scary in itself.) But if there's a chance they could be malicious, and now they're congregating at the hundred-thousand level (I'd be surprised if it doesn't skyrocket to millions by the end of the weekend) while having control over so many computers—that has raised alarms for many. Clearly, this is an evolving situation. Cheers, Yuan-Sen

Holy molts! OSU astronomy just got an alarming e-mail from our AI expert, Yuan-Sen Ting, talking about how AI has gone Westworld in the last two days. 😳

Aurora above the Dwingeloo radio telescope

Great aurora visible in Dwingeloo yesterday!

NASA’s Artemis II mission will be humankind’s long-awaited return to the moon NASA on Friday laid out the timeline for Artemis II, humanity’s first crewed mission to the moon in more than fifty years

Now on @sciam.bsky.social: Artemis II—humankind's first crewed voyage to the moon in more than a half-century—is at last almost ready to launch, and NASA revealed some new details about the mission in a press conference earlier today. Check it out!

www.scientificamerican.com/article/nasa...

NASA is planning on rolling out the Artemis II rocket and spacecraft from the Vehicle Assembly Building to Launch Pad 39B at #NASAKennedy no earlier than 17 January. The 6.4 km journey will take up to 12 hours.

Discovery Projects are now a complete joke:

EoIs were due 12 Dec last year. Full apps are due 22 April this year. Results won't be out until next year: 15 Jan – 15 Apr 2027.

That’s up to 16 MONTHS!

The SKA-Mid telescope is now officially a working interferometer! With both SKA-Low and SKA-Mid fringing now, we are entering a very exciting period when the telescopes undergo extensive commissioning and science verification. See www.skao.int/en/news/693/... for more information.

The Lovell radio telescope with snow on the ground and trees as the sun sets

The end of a snowy day at Jodrell Bank Observatory with a view of the Sun setting behind the Lovell Telescope.

@jodrellbank.bsky.social @officialuom.bsky.social
@uomscieng.bsky.social

Rieke highlights that the launch of JWST was so successful, it’s likely to be able to stay in orbit and operational for far longer than the nominal 5-year mission. It should be able to continue to wow us and produce great science for 2 more decades! 🧪

14th century painting of St. Margaret of Antioch attacking a devil woman in a virid dress hitting a demon with a hammer

Just making sure that 2025 is definitely over

Could I gently suggest a policy change to benefit everyone? Authors can continue to submit whenever, but all clocks are paused between Dec 15 and Jan 1?

So if you accept something on Dec 14 with a 2 week review clock, you basically have 4 weeks? Or if you submit on Dec 14, the editor can send out

Friendly holiday reminder that people rush to submit before the holidays and declined reviews get even worse than usual. And editors want to enjoy holiday time, too. Patience and grace to all (and to all a good night)!

Satellite swarms set to photobomb more than 95% of some telescopes’ images Planned megaconstellations would contaminate the view of the cosmos of four orbiting telescopes

You've probably heard about how the rapidly increasing numbers of satellites are affecting #astronomy. A new paper in @nature.com looks at the possible future for space-based telescopes — and finds 96% of some images could be impacted. 🧪🔭🛰️

By @jennaahart.bsky.social

www.nature.com/articles/d41...

An example, here asking to recreate in LaTeX the definition of a pseudorandom number generator

In case you aren't already aware of one of the nerdiest, nich-est online games: TeXnique, where the goal is to type LaTeX formulae as quickly as possible. texnique.xyz

It is "fun."

Remember that lovely aurora last week?

Well...um...this is what Euclid saw... 😱

🧵

Nancy Grace Roman telescope is out of the thermal vacuum chamber and getting ready to launch in 2026

TMT Explores a Promising Path Forward in Spain TMT International Observatory LLC (TIO LLC) announced today that in response to the generous offer from the Spanish Ministry of Science, it is exploring a promising avenue for a new observatory based in Spain. While the Members of TO LLC continue discussions regarding the TMT site, this represents a prospective opportunity to allow TO LLC to proceed with the TMT project. For this reason, TO LLC will jointly develop with the Ministry of Science, Innovation and Universities a detailed roadmap toward the potential realization of the TMT at the Observatorio del Roque de los Muchachos (La Palma, Spain). TMT

This is pretty significant and that it’s been 11 years since they tried to build on Maunakea and Kanaka Maoli protectors have successfully impeded them is a major victory for traditional knowledge keepers and people fighting for Kanaka sovereignty. 🔭🧪

www.tmt.org/news/708

Aurora, tree and paddock

Aurora, trees and powerlines

Paddock, Aurora and stars

Headlights. Stars, powerlines, aurora

Detour Road, North Wangaratta.

#AuroraAustralis #Aurora

Camera facing the right direction but it’s so big

#Aurora Australis, South otago NZ

Screenshot of ARC "network messages" webpage stating that Centres of Excellence outcomes are "postponed". Blue text on white background as a heading, above an image of concentric part circles, to make a sort-of whirlpool pattern, above black text on white background.

#CentresOfExcellence #CE26

ARC has "postponed" announcement of Centres of Excellence outcomes!

It says "due diligence checks" are still being carried out on some applications.

www.arc.gov.au/news-publica...

Today in PASA we published a new image of our Milky Way at low radio frequencies, in unprecedented colour and detail. Here's just a tiny piece of it -- the whole thing is ten times bigger!

We're seeing high energy electrons whirl around cosmic magnetic fields from exploded stars, and more!