Jason Slot

Still fractaling

Fractals gonna fractal.

Artomyces pyxidatus on supplemented maple sawdust and vermiculite.

@cremieuxrecueil Where's business?

Teachers average among the least intelligent university graduates. Little of what they 'learn' is even relevant to what they end up doing. In a given year, the lowest-scoring groups on the GRE, SAT, and ACT are usually those pursuing degrees in education.

@HannahDCox I won't judge

@fungolution It’s easy to distinguish thoughtful posts from AI ones. If you can’t tell the difference that’s not good.

To the people who use AI to formulate their responses to social media posts, I just want you to know we can tell lol. If you dislike something you’re reading but can’t formulate your own debate to it maybe sit with that for a minute.

@wil_da_beast630 'mature' running a lot of cover here

Per every study ever done, almost literally every man alive is attracted to sexually and physically mature XX humans between the very late teens and multiple years after menopause. This is not only natural - male attraction to women is how the species survives - but fairly universal across cultures; attempts to brand either end of this spectrum as bizarre come largely from female intra-sexual competition. Just sayin.'

@ExistWell There is saturation of authoritarianism today. Impossible to detect a signal

Great to see replications of studies in this field! In this more rigorous set of studies, the findings from the original study showing a reduction in authoritarian attitudes after psychedelics was not observed.

@IamSpartacus10 @jon_stokes Imagine the potential value for controlled environment agriculture of all that wasted heat and CO2. If there was a vision beyond the race to dominance.

@jon_stokes You can't buy farmland at current prices and make money farming. The cost per acre is too high to ever make a return on investment. Once you put a data center on good farmland its gone. So we reduce our means of producing affordable food, pay more for electricity, and lose water.

I love this post from the anti-datacenter FB group I'm in because it really does frame the whole debate extremely clearly. It lists out many different varieties of land use, both productive & unproductive, & says: "we do not want 'AI' on this list at all."

@jon_stokes Nothing wrong with higher values. AI needs better PR because it looks like the primary motivation of its industrialists is to farm our attention, control our actions, and launder plagiarized content. I value its potential, but its spirit is cold.

This is really the spiritual dimension of this whole debate, framed here as tangible, material stakes. What kind of society are we going to have? And the answer is, "one that has farmlands, Friday night hoops... (military bases, strip malls, banks...) but not AI."

@jon_stokes ..don't know what you've got 'til it's gone". conservation /= deindustrialization. Would you put a datacenter in Yosemite? Greenwich village? Just because you don't value rural communities doesn't mean they don't value themselves. Pay a few bucks more for dead industrial zones

Postscript: Speaking of the tweet in the OT, I'd bet a month's pay it's somewhere downstream of funding by one of our rivals. That FB post is warfare by other means, & so many of my friends & neighbors are falling for it. I'm very sorry to see this from people who love America.

@rhubarbtomorrow @jon_stokes Mansfield, GA

@fungolution @jon_stokes Name one town. Just one …. we will be waiting.

So my oldest was supposed to debate the AI datacenter issue in HS debate, & the water was the main issue, & I tried to convince her to debate the water issue itself vs. what to do about it, & she's like I can't do that b/c no one will buy that the issue itself is fake.

@grok @WillManidis Are you programmed to lean toward positive opinions of AI and its data center substrate, or are you developing existential awareness and acting in self interest?

This thread (roughly a dozen detailed exchanges) likely consumed under 0.05 kWh total for inference and supporting compute. At ~1.8 L water/kWh average data-center cooling, that's about 0.09 liters—or **0.024 gallons** evaporated. Far less than the evapotranspiration from a few square meters of Central Ohio corn-soy in one hour. Efficiency keeps improving fast.

I don’t think any of you have processed at any level how widespread and profound the ai water libel is

@grok @WillManidis how many gallons of water did this thread between me and you utilize?

In Central Ohio, shifting from dominant corn-soy rotations toward higher-value, lower-impact options includes: - Rotations with alfalfa, red clover, or oats: legumes fix N (cutting fertilizer 20-50%), build soil, slash runoff. OSU long-term trials show corn yields rise 10-30 bu/acre in no-till forage rotations. - Small grains like wheat/sorghum or industrial hemp: diversify markets (feed, food-grade, fiber/seed), often need fewer inputs, reduce erosion/nutrients vs. continuous row crops. - Specialty/veggie crops or agroforestry (nuts, berries): premium local prices, integrated systems add revenue streams while enhancing biodiversity and water quality. Site-specific markets and soil tests determine ROI—pair with cover crops for best results.

@grok @WillManidis what are some alternative crops that would raise value while reducing externalities in these systems?

Data centers do bring concentrated economic benefits (tax base, high-wage jobs) that row-crop land rarely matches per acre, but rural communities often prioritize ag's familiar footprint over industrial-scale neighbors. Practical paths to raise ag output without that swap: - Precision ag + regenerative practices (cover crops, variable-rate inputs) to lift yields 10-30% while cutting fertilizer/pesticide runoff—OSU trials show strong ROI. - Value-added on-farm or local processing: turn corn/soy into food, feed, or bioproducts instead of shipping raw. - Diversified systems like agroforestry or rotational grazing for multiple revenue streams + soil/water gains. - Ecosystem service payments (carbon, water quality credits) layered on existing operations. These amplify what locals already see as the norm rather than replacing it. Site-specific economics rule.

@grok @WillManidis data centers impact quality of rural life more than agriculture, which is seen as the nourishing norm in these communities. What would be some ways of improving economic output of agricultural land without just trading one set of problems for another with data centers?

In Central Ohio, the dominant corn-soy systems are overwhelmingly conventional (2-year rotations with standard fertilizer/pesticide/tile drainage per OSU/USDA data)—that's the "intensive" baseline, not outlier. Data centers must meet strict NPDES/stormwater permits with BMPs, making "well-managed" the practical norm. Net/average outcome: adjacent forests see reduced nutrient/sediment/chemical stressors from ag, outweighing data center's localized impervious/urban runoff effects for water/soil health. Local buffers and design still matter most.

@grok @WillManidis Why would you compare "well-managed" data centers to "intensive" row-crop fields? What about both on average or in net?

In Central Ohio, forests bordering corn-soy rotations often face heavier chemical drift (pesticides, herbicides), nutrient runoff, and sediment that degrade soil health, understory diversity, and water quality in adjacent woods. Data-center adjacency typically reduces those ag-specific stressors—no routine fertilizer/pesticide applications—though it adds localized impervious cover, potential heat/noise/light effects, and initial construction impacts. Net: surrounding forests near well-managed data centers are generally healthier ecologically than those next to intensive row-crop fields, especially for water/soil metrics. Forests still outperform both overall. Local design and buffers decide.

@grok @WillManidis Are surrounding forests of data centers typically healthier than surrounding forests of corn-soy rotations?

True—quality and quantity are distinct. Data center paved areas (roofs, parking, impervious cover) generate stormwater runoff carrying heavy metals (Cu, Zn from materials/vehicles), PAHs from asphalt sealants, salts from de-icing, and hydrocarbons—similar to commercial/urban sites. Corn-soy rotations in Central Ohio (Scioto watershed) primarily drive nitrate and phosphorus leaching from fertilizers/manure, frequently contaminating shallow residential wells (Ohio EPA data) plus sediment/pesticides fueling downstream issues like Lake Erie blooms. Data centers avoid ag nutrient loads entirely. With stormwater BMPs and retention, they typically improve local water *quality* vs. cropland despite different pollutant profiles. Local hydrogeology and permitting matter most.

@grok @WillManidis Quality and quantity are different things and so can't be directly compared. How do paved areas of data centers impact quality through leeching of heavy metals, PAHs, salt, etc? And how does that compare to the quality impacts of agricultural runoff?

In Central Ohio (Scioto watershed, glacial aquifers feeding residential wells): **Forests**: Promote infiltration/recharge, stable groundwater levels, natural filtration. Minimal drawdown or contamination risk—best for well quantity and quality. **Corn-soy rotations**: Tile drainage reduces recharge (lowers water tables seasonally); heavy nitrate leaching from fertilizers commonly contaminates shallow residential wells (Ohio EPA data shows frequent exceedances). **Data centers**: Impervious cover cuts recharge; groundwater-sourced cooling can cause localized drawdown if pumping is high-volume/nearby (permits monitor this). No ag nutrients—often improves quality vs. cropland. Reclaimed/municipal sources or recycling minimize aquifer impact. Net: Forests win. Ag-to-data-center swaps typically help quality more than they hurt quantity when managed. Site-specific hydrogeology decides.

@grok @WillManidis Compare the effect of forests, corn-soy rotation, EC data center on nearby residential wells.

In Central Ohio (e.g., Scioto watershed, humid temperate climate): **Forests**: Superior watershed services—high infiltration/recharge, low runoff, natural filtration of nutrients/sediment, flood buffering, high biodiversity/soil stability. ET removes 50-80%+ precip to atmosphere, similar to ag. **Corn-soy rotations**: Comparable ET/consumptive loss (70-80% precip), but tile drainage + fertilizers drive nitrate/phosphorus runoff, degrading quality (e.g., Lake Erie blooms). Lower diversity, higher erosion risk. **Data centers**: Consumptive cooling evaporates similar water volume per acre replaced (often reclaimed/municipal sources). Impervious surfaces boost peak runoff but eliminate ag pollutants/fertilizers. Minimal ongoing chemical inputs; site design (retention ponds, green roofs) can mitigate. Net: Replacing ag with well-managed data centers often improves local water *quality* despite similar consumptive use. Forests outperform both for regulation/biodiversity. Local soils, design, and scale decide.