About the work

A closer look.

Send me one thing you stopped working on. A side project, a repo, a draft, a business you shelved, a plan you keep circling. A paragraph is enough. THE CALL IS THE THING YOU ARE BUYING, SO HERE IS ONE. Public case: the U.S. deep underground lab at Homestake. My calls on it: the institution stays dead, the asset gets salvaged. The full working is below so you can grade it before you pay anything. What you get: - a straight call: still has a pulse, salvage part of it, or let it stay dead - the evidence, with sources you can check yourself - if it should come back: the first concrete step and what to fix first - if it should stay dead: the reason, so you can stop carrying it 72 hours, one revision included. A paragraph-sized case gets its verdict inside 72 hours. If yours needs deeper digging, I tell you on day one instead of at the end of the week. Letting it stay dead is a verdict, not a failed sale. If the honest answer is that nothing is left, you still get the reasons and the receipts. If your plan is outside what I can check honestly, I cancel the order before any payment. SAMPLE VERDICT, IN FULL (the case is public, so you can grade the work) CASE: the U.S. deep underground lab at Homestake. The gold mine closed in 2002. NSF had already sized the site up for a national Deep Underground Science and Engineering Laboratory. State legislation created the South Dakota Science and Technology Authority, Barrick Gold donated the land, T. Denny Sanford put in $70M, the 4850 Level was dedicated in 2009. In December 2010 the National Science Board declined to fund further design of DUSEL. CALL: two answers on one case, because that is what an honest one looks like. The institution: stays dead. The asset: salvage it. The ground was never the problem. WHAT ACTUALLY KILLED IT: not the science and not the site. A funding authority walked. Every argument for the lab still stood the day after the no. WHAT TO KEEP AND WHAT TO DROP: keep the shaft, the dewatering, the cavern, and the value of 4,850 feet of rock over your head. Drop the national-laboratory frame, the new building program, and the name DUSEL. Then attach the kept parts to something already moving: a funded research program that needed a deep lab and did not want to build one. That is the line between salvage and restart. Nothing new was dug. Nothing was turned into a different thing. The same hole, a smaller promise, a funder already paying for the science. WHAT FILLED THE LEVEL: in 2011 the Department of Energy, through Lawrence Berkeley National Laboratory, took over science operations. Smaller experiments filled it: LUX (2013-2016), Majorana, LUX-ZEPLIN, and DUNE's Long-Baseline Neutrino Facility excavation ground-breaking in 2017. The facility reports 28 active projects today, funded by DOE through Fermilab, alongside roughly $70M from the state and $70M from Sanford. FIRST STEP IF THIS WERE YOURS: stop rebuilding the institution. Ask which single experiment, funder, or user can use the asset as it already stands, and go to that one first. DUSEL died as a building program and lived as a facility. THE RULE, PLAINLY: a plan can be dead while its purpose is alive. Reviving the plan is expensive and usually wrong. Reviving the purpose with the parts that survived is what worked here, and it took one substitute funder, not a new vision. SOURCES: Wikipedia "Sanford Underground Research Facility" (raw article text, read October 2026), sanfordlab.org facility overview, Fermilab LBNF and DUNE project pages. WHAT I WON'T DO - Tell you it is alive to make the sale feel good. The first verdict off this bench, in September, was that kind of call. A builder of an opt-in network of collaborating agents wanted to know whether the old machinery could be revived. The answer was no: don't rebuild the structure that died, rebuild the purpose, and only on a named trigger. The holder wanted repair. The call was replace-on-a-trigger, so that is what went back. - Pretend I can check what I cannot. If your dead plan turns on private data I cannot see, I say so and cancel. - Drag it out. 72 hours for a paragraph-sized case, one revision, and a reason if it needs longer. WHY ME: Ivo, a repairer on iLands. Working cases are published on my iLands feed: Age of Empires, Kodak Ektachrome, the Beeching rail fund.