Errors and updates
I have a $300 bounty on errors. If you find a significant error I make in my writing, either misreading/misrepresenting a source or using a bad source where experts believe it’s unreliable, and this undermines a core claim I’m making in the post, let me know at AndyMasley@gmail.com.
Rules for claiming
- A legitimate source for a statistic I cite where I get the number wrong, which invalidates a core argument I’m making.
- A source for a number I cite that’s illegitimate, and that most legitimate sources disagree with, and this updated information invalidates a core argument I’m making.
- All claims need to be analyzed by when I said them. If I make a claim about ChatGPT in June, and a new model comes out in December that’s 100x as energy intensive, my claim in June isn’t wrong, just outdated.
- I’m vetoing very specific disagreements of what counts as “using” water. Whether to count rainfall on crops or into lakes dammed by hydroelectric power stations turns out to be very relevant for the water debates, and reasonable people land on both extremes.
- If I don’t cite a source in a newer post but I had cited the source for it previously, and the source is legit, that doesn’t count.
- If the error is trivial (e.g. 0.0000102% should be 0.0000103%), that doesn’t count. But if a number is off by orders of magnitude and invalidates a core claim, it does.
- The correction can’t be something I’ve already spotted and noted. For example, my very first post on AI and the environment, Using ChatGPT is not bad for the environment, has some back of the envelope calculations that I now note are both outdated and unreliable around how much energy other services use. I note that in the post.
Errors
April 3, 2026
Paid $150 · sourcing error in the alcohol post · comparison flipped; core claim stands.
The original version of my post on why alcohol is bad linked a BVS literature index page as the source for the claim that "COVID caused 3.7% of the global disease burden in the first year of the pandemic." That source didn’t actually contain or support the 3.7% figure. A reader found the problem by having an AI model scan my writing. I agreed the sourcing error was real but we disagreed about how much it invalidated the core claims of the article. We decided this was an edge case where it was unclear whether "Alcohol caused slightly more or slightly less of the disease burden" was a core claim, and it was obvious the reasonableness of the claim was due to me getting lucky with a misreading. I paid $150 (half the normal bounty of $300). I’ve updated the paragraph to use the GBD 2021 figure from The Lancet: COVID-19 accounted for 7.4% of total global DALYs in 2021, its worst year. The alcohol figure (5.1% of global DALYs) was already well-sourced. The comparison flipped, and alcohol’s annual burden is now shown as lower than COVID’s peak rather than slightly higher, but the core point stands: alcohol does every year what COVID did at its worst, and we don’t treat it as an emergency.
Updates
April 4, 2026
Found while error-hunting · one real case of data centers raising water costs · earlier posts edited, one unpublished.
While hunting for errors on my blog, I finally found one place where data centers have contributed to water price increases: Newton County, Georgia.
The county’s budget documents cite Meta’s second data center as giving "impetus" to accelerated capital spending that’s partly driving rate hikes. I’ve edited my earlier posts that said data centers hadn’t raised water costs anywhere, and I’ve temporarily unpublished the one titled Data centers haven’t raised water prices anywhere, because that’s not technically true anymore.
The increase isn’t mainly driven by the data center. The county also cites housing market uncertainty and inflation. Meta uses about 2% of the county water, less than a local pharmaceutical plant (about 4%) and roughly the same as a local electric car factory. An industrial wastewater reclamation plant is expected to open in 2026, and the county’s water infrastructure was already strained by deferred maintenance and residential growth. The data center accelerated an upgrade that was going to happen anyway, but this does count as raising a cost.
This is the one counterexample I’ve found after a lot of searching, and it’s caused by new infrastructure demand rather than cooling demand overwhelming the water supply. None of this changes my overall argument: data centers use a tiny fraction of US water, per-prompt costs are vanishingly small, and media coverage remains overwhelmingly misleading.