I see today via the LHC Page 1 Vistar that a problem at a cooling tower will cause part of the accelerator to need to be warmed up to room temperature, putting the LHC out of business for the next 4 weeks or so.
The LHC has just been coming out of long shutdown the past few months, starting its Run 3. In the past couple weeks it has started to get up towards its full luminosity potential, with over 2400 bunches in the beam. So far during Run 3 the machine has delivered an integrated luminosity of about 10 inverse fb to each of the experiments (ATLAS/CMS). The plan for Run 3 (which is expected to last through 2025) is to accumulate an integrated luminosity of about 300 inverse fb, doubling the 140 inverse fb of Run 2, at a slightly higher beam energy (6.8 TeV vs. 6.5 TeV).
Update: Thanks to Benson Woo for pointing out to me an article today in the Wall Street Journal about plans for possible shutdowns of parts of the CERN accelerator complex this winter in case of energy supply problems due to the ongoing conflict with Russia.
Although 300 fb-1 does not sound like huge improvement over Run 2, there are several triggers (online data filters) that are getting tested for the High-Luminosity LHC era. Of course, Run 3 will also chase down some interesting anomalies we saw in Run 2.
One interesting thing about Run 3 analyses will be how machine learning techniques are incorporated into nearly every facet. The comfort level for ML has really increased. Concerns about things like overtraining and bias are real, but not considered insurmountable.
Snowmass (shall we call it Rainmass?) report
https://www.scientificamerican.com/article/physicists-struggle-to-unite-around-future-plans/