Comment on A new model proves the galaxy could form without dark matter, but makes our universe twice as old. Succesfully simulated it with only a few test left. It solves a lot of new reserch discoveries.
You are viewing one comment and its replies. View all comments
Comments
At a glance, when I was studying these topics a decade ago, the model would have been discarded right away.
CCC+TL works assuming some constants slowly change over time. The issue is, based on QFT (the most successful theory in terms of predictive power accuracy, more than GR), a slow changing fundamental constant is automatically linked to a low mass particle, the type that are very easy to produce in a particle accelerator.
With this model, they therefore have the burden to explain why we don’t detect particles related to these changing constants.
It’s super interesting, I’ll try to read more about it!
TBF they discard fuckton of data from colliders.
I remember one of the recent-ish spacetime episodes mentioning that they probably “detected” the thing they would need to advance some new theories, but their algorithms for deciding what’s garbage data and what’s useful discarded the data ¯\_(ツ)_/¯
Edit: hence their new goal at CERN with upcoming upgrades, to increase the amount of collision information captured and retained, instead of just increasing collision energies, which new technologies allow… If only it wasn’t for idiotic ai datacentres hoarding said hardware technology and making it costly for scientists.
So a really solid genuine reason to hate ai.
Isn’t CERN now using mistral to help them sort the data out of the noise which is helping discover new particles? The partnership was just announced one month ago so it probably hasn’t resulted in anything yet.
I don’t like generative AI, but this is something it’s really good for. The whole basis of the systems is statistical predictions. It’s really good at taking in a bunch of data and piecing together things that fit. This is also something humans are not good at.
AI is a tool. It’s usually used to just replace jobs, and it’s shit. Using it for things like this is actually very good though.
This looks like a partnership between CERN’s IT Department and Mistral. The physics work happens elsewhere 🙂
Having more data is always good for statistics. Processing with dedicated algorithms is then hamstrung by the cost of memory, disks, and GPUs 😑
Sign in to comment. Sign in