@nu_phases

Associate Professor @UCSanDiego, Cosmologist and Particle Theorist

drgreen.physics.ucsd.edu/
Joined February 2019
  • Tweets1.9K
  • Following546
  • Followers10K
343Media
This December, I will be trying another experiment: Each day I will post 1 under-appreciated paper/result (my definition/no rules). I will also be posting my choices on blue sky (dangreen.bsky.social). I am curious to see which site has more interesting feedback/discussion
9
7
5
57
13,453
Reminder: most fields of basic science are driven by experiments and a key limiting factor is $$$ for equipment, materials, etc. AI companies have no incentive to pay for this kind of work (that's why it doesn't happen in industrial labs already).
This quoted post is unavailable.
Try viewing anyway · View on X
1
13
1,148
If I'm wrong, and AI companies decide to fund, eg, a 100 TeV collider, that would be great news for science and scientists. But this kind of work won't help them train models / make money, so it would be entirely philanthropic. The progress is math does not make this more likely.
2
4
290
Why is it important to learn math? So that you can see a huge outlier like this one and realize it needs more explanation before publishing @nytimes * Duke's number is the median of 17 people, i.e. it's determined by 9 salaries (you might even be able to interview all of them)
5
1
13
1,752
Daniel Green retweeted
Marko Papic
@Geo_papicOct 5
Geopolitical Cousins dropped... a bit of a swerve! @JacobShap and I had our "Science Cousin" -- Cousin @nu_phases - come on to talk AI, innovation, importance of science, etc. He tried his best not to blow our brains with fancy schmancy big words. geopolitical-cousins.captiva…

The Warning Label Is The Pitch - Geopolitical Cousins

OpenAI burned $15 million in three days to beat Anthropic to a Millennium Prize problem. Nobody's quite sure whose work they trained on to do it. This w...

geopolitical-cousins.captivate.fm
4
4
3
43
7,262
1. This should have never been a "field of physics". If a paper is so obvious 100 people write it in 24 hours, then it wasn't worth writing it in the first place. 2. Apparently not everyone knows what ambulance chasing means (I had to explain this term to students last week)
Will Kinney
@WKCosmoSep 28
Ambulance chasing will be the first field of physics to be fully automated.
7
5
617
73,336
There is a single common problem facing academia: People that claim to be experts trying to convince the world that learning/knowledge is unnecessary The world will not be a better place if: - Everyone gets a degree and no one knows anything - AI can think but humans cannot
Alec Stapp
@AlecStappSep 16
I thought the tweet below must be omitting some critical context or paraphrasing the text ungenerously. But no, the NYT actually published an op-ed that claims the University of California hasn't sacrificed excellence even though it admits students who struggle with basic math.
2
2
15
1,644
If AI is great at finding mathematical counter-examples, where are all new theories physics that contradict conventional thinking we were promised? Wouldn't AI have already produced a theory that replaces dark matter and dark energy already? Or a new approach to quantum gravity?
20
5
1
39
5,994
Apparently, all these revolutionary ideas are on the internet but haven't been developed because of physics gate-keepers. If that is really true, throwing a few thousand agents at these problems should be generating loads of new self-consistent theories that are allowed by data.
5
1
13
831
Daniel Green retweeted
Raphael Bousso@RBoussoJun 30
College admissions is broken. It can be fixed. Liberals should own this cause, not cede it to the imbeciles who believe Trump lowered drug prices “by 1600%” washingtonpost.com/opinions/…

Opinion | I’m a professor at Berkeley. Bring back this requirement for entry.

The outcomes of California’s admissions experiment are even worse than predicted.

washingtonpost.com
1
1
410
Come to @UCSDPhysics. We will continue to have human grad students.
Harvard Physics@harvardphysicsMar 24
The AI grad student: anthropic.com/research/vibe-…

Vibe physics: The AI grad student

A Harvard physicist guided Claude Opus 4.5 through a full quantum chromodynamics research project in two weeks.

anthropic.com
7
7
2
170
24,274
Want to understand how we got to a point where a large fraction of UCSD students struggled with 7+2=6+[] (1st grade level)? Just look at the proposed solutions You can't make up 12 years of math with a few "evidence based" courses. Eliminating calculus is just accepting failure
3
1
7
1,396
See the opinion piece here: insidehighered.com/opinion/v…

The Bigger Picture Beyond the UCSD Math Report (opinion)

Declining math skills are a real problem nationally, but UC San Diego is an unusual case—and other California colleges offer a range of promising models for rethinking calculus requirements.

insidehighered.com
1
3
703
This is such an absurd claim it's actually pretty funny. A few years ago I went over 31 examples of breakthroughs from the past 40 years (excluding Nobel prizes) xitter.cf/nu_phases/status/15983… But 50+ years really opens up some all time great results:
Sabine Hossenfelder
@skdhJan 2
Replying to @Hassaan_PHY
It's a fact that the foundations of physics have been stuck for 50+ years, everyone with half a brain can see that. The only "counterargument" against this are physicists who complain that writing a lot of papers is sorta progress.
37
30
7
325
151,103
more replies
Inflation (1981): Guth lays the foundation for every theory of structure formation journals.aps.org/prd/abstrac… Later Linde, Albrecht and Steinhardt later improve the model to the point it makes a number of non-trivial predictions (many now verified)

Inflationary universe: A possible solution to the horizon and flatness problems

The standard model of hot big-bang cosmology requires initial conditions which are problematic in two ways: (1) The early universe is assumed to be highly homogeneous, in spite of the fact that...

journals.aps.org
5
22
4,932
Quantum Computing (~'80): The idea that there are things that can be calculated faster on a quantum computer than a classical one has profoundly changed the relationship between the theory of computation and fundamental physics semanticscholar.org/paper/Th… link.springer.com/article/10…
5
1
26
4,524
Science's Breakthrough of the Year 2025 is a failure of science reporting: (1) Installing solar power is neither science nor a breakthrough. (2) Rubin is a runner-up for being built? Scientific progress requires discoveries in the real word, not just meeting construction goals
2
9
854
This kind of grade inflation hurts science. It is too common to declare something a breakthrough just for being popular. Science's breakthroughs have previously been very good and focused on real progress. It's sad to see them fall into the same trap. science.org/content/article/…

Science’s 2025 Breakthrough of the Year: The unstoppable rise of renewable energy

Clean energy infrastructure is being deployed with unmatched scale and speed—and China is leading the way

science.org
3
596
Load more