ProgressReport 10 April 2026
My number one focus has remained on getting AI plugged into my every day routine; I'm using it wherever I can, but it has brought the biggest gains in programming. This has been truly remarkable, so fun. I'm turning ideas into working apps. In the past months I've built out: Notes Webpage, Third Brain, cmaster-cworker, porover, maccounts, barista-console, airbnb lock updating and more.
Habits have all been good. I could use some more meditation. Reading is picking back up after a bit of a slump. I finished the confederacy of dunces (average), have been working through Shakespeare: (Hamlet / Othello), and have been reading a lot of blogs on AI.
I have had a number of nagging injuries which have been piling up: shoulder, finger, knee. These have kept me from BJJ and basketball for a couple of weeks, but now seem to be getting better.
Big news around the house is that Steph has started her job. She has going into the office, but seems a bit frazzled.
I've meet Matthew a couple of times for coffee, he seems really cool. Have had Tom and Deb over to see the new hours and did easter with Lori and Mark. Mark and Barbie will swing by this weekend.
Airbnb has been going steady. We will finally turn a profit in April (even without the draft weekend booked) and have even better months following. I got super host status, we will see how much this moves the needle.
On the 4b side, my focus – apart from serious development and code cleanup with claude – had been uisng the mixed data as 4b background model. Ive setup the mixed-vs-data training "MvD" and see a reasonable closure. The modelling of the b-tagging weights are especially promising; this should open the door to including them in the SvB. I have used this as an oppurtunity to use Snakemake workflows to control the jobs. This is definetly the right direction and I have started working on barista-console which will allow us to monitor all the condor jobs at once.
AE has pushed throught a first version of the low-pT analysis and is not seeing big of an impact as we had hoped. The improvement on the combined limit is about 10% and is even less significant on lambda. We will iterate for a bit to see if there is a bug, if not we need to see if this is worth it in the end.
On the HGC side, there are lots of pending personelle changes; Austin will start end of April; Jasques and Arvind are going to grad school in the fall and will be leaving in the end of the summer. I have leads out on two replacements: Cooper and Shrindini, but am waiting on getting the ads through HR. Modules are being made. We have signed off on the 5% milestone and the MMTS comining into its own! The next big step is to reliably get the dew point below -40-degrees which will allow cold testing. This will likely require a new dry-air system.
I joined the safety trip to Alabama. It was good to see that part of the country, but its not clear that the work that they are doing there will be so relevant to the project. Will follow up more with Jeremey in the next few weeks to see where they might be able to fit in.
The last bit of news is that I will be teaching Physics of Energy next spring. This should be a good chance for me to learn something new. I have started reading to prepare and will develop some modules (eg: on energy in biology, energy for AI) using AI.