Water is once again right in the middle of a hot California dispute!

Kudos to Erik Benson for a great article on the market dynamics that are impacting both agriculture and AI data centers around water rights, land values, and the relative impacts on state economies that are or are considering the impact if both are increased or decreased. This is not a hypothetical exercise. Data centers are indeed looking for land, and some of it is or has been used for agriculture. Erik does a great job of diving into the details. I recommend giving the full article a read. It’s worth it because of the time he took to frame things up and reality check a few assumptions many of us that cross both ag and AI ecosystems end up buying into without always wrapping the whole story and context around it.

I think this topic deserves a full response (but that will take me a while!) For now, here are my quick thoughts to provide a little more context around the water conversation:

1) The SGMA allocations around groundwater are getting real this year as allocations from Groundwater Sustainability Agencies start passing out groundwater allocations that will require farmers to fallow acreage in high-risk groundwater basins. In some cases, depending on the type of water rights and basin classification risks, farmers are already making decisions to move from permanent crops (tree crops like nuts and stone fruits or vine crops like wine grapes and table grapes) to rotational crops to mitigate the risk of a fallow allocation from SGMA.

California could have avoided the looming crisis that is SGMA by building more surface water storage and conveyance (Sites Reservoir and many others like it, supported by canals and aqua-ducts) or desalination plants. The solution for a supply shortage is often building more storage. See Peru­—$24B in government commitments for 22 water storage projects— as the opposite approach to California.

2) The water wars were already happening before AI data centers. The water costs and supply challenges were already pushing agricultural acreage toward the highest and best use. Combine rising water costs and supply challenges with the overall regulatory landscape of California agriculture and the never-ending labor cost increases and farmers needed to continually move to better and better economics well before AI became a buzzword.

Crops will now compete with AI data centers just like they competed against other agricultural uses the past couple of decades. In that sense, everything old is new again, and where AI provides the highest and best use it will take some of the agricultural land and convert it.

Link to Erik’s article – (19) Dirty Water | LinkedIn