Mike Mahoney

Mike Mahoney

@mikemahoney218.com

Helped build NYS's Forest Carbon Assessment. Currently helping build public water data infrastructure. PhD in Environmental Science. #rstats, ML, Boston and spatial data. Opinions my own. RTs imply causation. https://mm218.dev github.com/mikemahoney218

3.2K followers576 following1.1K posts

Mike Mahoney reposted

to little surprise, "Tailscale oh no no no no no" turns out to be...uh. well, you can read it yourself. this is an axe-grinding exercise that uses confident-sounding slop as camouflage.

tailworm.comTailworm

Mike Mahoney reposted

unpopular opinion, "admits uncertainty" and "flags nuances" are shit. they are terrible fucking traits actually. is the sky blue? do we really know that? what is blue, anyway? many cultures have different understandings of what blue is. i have not independently verified the exact spectra of the sky

This post is unavailable.

Mike Mahoney reposted

It saddens me to report that: - There's an ongoing LinkedIn beef around attribution in open source projects, and - I pay attention to LinkedIn I have ... thoughts. I want to avoid the LinkedIn blast radius (for now) so this seemed like a better venue in which to float these thoughts. Here we go...

Mike Mahoney reposted

This is absolutely insane, I just rendered this effectively 3.1 billion triangles 900 frame 16 sample per pixel 960x540 animation in #RStats last night in 6 hours using the M1 GPU and #rayrender -- it would have taken a WEEK with the CPU renderer. Completely game changing for iterative work

Mike Mahoney reposted

Replying to @solarchase.bsky.social

6. In a way, it’s reassuring that solar is saturating, as it means solar is actually displacing fossil fuel rather than power demand increasing vertically. But it means the beginning of a new phase: the age of cheap solar and batteries. This too will pass, but it’s just getting started now.

did the nine books thing off the top of my head. Of these, I absolutely talk about the nonfiction the most, and think about the culture the most

I think vs consensus I’ve got the “wrong” choice for Card and Vonnegut here, and arguable choices for Pratchett and Banks — but I think these are much more interesting books than the “right” answer for the authors, and I certainly spend more time thinking about them

Great thread. This is a problem with a lot of climate journalism (especially at outlets like The Guardian) — they assume that energy usage would be stable or decreasing without green build-out, when there’s no clear reason that would be true!!

Zeke Hausfather@zekehausfather.com

I often see people argue that clean energy is not reducing emissions, because it just adds to rather than replaces fossil fuels. This argument is deeply misguided – clean energy built since 2005 has displaced 23 billion tons. I dive in over at The Climate Brink:

Degrowth in general starts from a place of “if everyone agreed entirely with me and changed all of their behaviors” and that’s just never seemed like a winning strategy to me — and in all honesty I think the prevalence of degrowthers in climate communication has made us less successful

Mike Mahoney reposted

As Borges wrote “Nothing is built on stone; All is built on sand, but we must build as if the sand were stone.” But what he never considered is, what if we just built on stone instead of sand??

Eugene Vinitsky 🍒@eugenevinitsky.bsky.social

Deeply excited about the idea of gold standard science: we should do science where every result is high impact, statistically precise, and never contains mistakes

Mike Mahoney reposted

The thing is that it's not a % of projects that will ship, it's a kind of absolute floor, and so if AI actually increases productivity what you'd predict is an exponentially growing rate of project abandonment with probably a very small baseline increase in shipped good product rate.

dax@thdxr.com

the last 20% to get a product shipped and successful is as difficult as ever this is where 99% of programmers have always fallen off the first 80% is a lot easier and so we're seeing more abandoned projects than ever

today I was trying to get a model to generate an example for a vuln I was reporting through normal channels Claude got most of the way there before cutting itself off. GPT Sol wouldn’t even start. GLM did it eagerly and then came back with suggestions for other exploits to try

Sky Marchini@sky.skymarchini.net

>the houthis were trying to use claude to build their rocket guidance system just get me off the fucking ride www.anthropic.com/threat-intel...

It’s so cool seeing the huge operation behind the scenes that makes democracy possible, and I’m really happy to be a small part of it again this year. Massachusetts early voting this week and primaries on Tuesday! Go vote!

Agree 100% with this — the expectation for students to take long-term volunteer positions is morally wrong & locks out underrepresented communities. It’s also not even a great way to get your foot in the door! Unpaid staff are not first in line for cool experiences or facetime with researchers.

Andrew Gelman et al.@statmodeling.bsky.social

Why I am so against volunteering in academia (ecology-edition) statmodeling.stat.columbia.edu/2026/08/19/w...

relatedly, I STRONGLY believe folks should take at least a year for non-academic work before going to grad school It’s so important to get a sense of what is and isn’t normal, since grads will so often need to advocate for themselves against their university and any organization they’re in

This post is unavailable.

did you know we invented lighters before matches? just seems wrong

Colin McMillen@mcmillen.dev

i love that humanity developed ranged weapons in the following order: bow crossbow nuclear-armed intercontinental ballistic missile compound bow

it hadn't fully occurred to me that our new AI age is going to mean way more people just rolling the dice on every decision point in an analysis my favorite type of sum of squares is "stochastic"

Mike Mahoney@mikemahoney218.com

I don't blame folks for not deeply investigating the implementations of their software -- this is why packages exist after all -- but periodically I get bug reports that X doesn't match Y package and: yes! on purpose! that's the reason X is there at all 😂

"people just assume languages and packages will all agree" This is absolutely true, which is also annoying as a package author -- a lot of my software does something similar to other packages, but makes different choices, because I think they're better! And often users assume it should be identical

Andrew Gelman et al.@statmodeling.bsky.social

He fit the same statistical models with three different software and got much different estimates. It’s another dimension of the multiverse. statmodeling.stat.columbia.edu/2026/07/27/5...

I don't blame folks for not deeply investigating the implementations of their software -- this is why packages exist after all -- but periodically I get bug reports that X doesn't match Y package and: yes! on purpose! that's the reason X is there at all 😂

I've got a personal policy of not bad-mouthing open source software by individuals, which can also make this tricky. Good for the soul, but means I can only explain why my tool makes a choice, not necessarily why I think the other one is wrong

Mike Mahoney reposted

Becoming a conspiracy theorist that Wu and Hochul switched freaky Friday style sometime in 2025

Sam@samd.bsky.social

it’s pretty remarkable how one of Zohran’s COGE proposals is to streamline construction of bus and bike lanes while Michelle Wu is doing the literal opposite, bogging them down in red tape for some reason

Mike Mahoney reposted

starting to get a bit concerned that a significant number of technocratically-inclined people have assumed that solar+grid scale battery fixes this, rather than mostly just stopping the bleeding

Craig Harrington@craigipedia.bsky.social

This is an emergency, folks.

Solar build-out rates absolutely should make you optimistic! Also the Colorado River is several million acre-feet short of its allocated usage this year and Lake Powell is nearing its minimum power pool. Climate adaptation is still a reality. www.washingtonpost.com/climate-envi...

washingtonpost.comA major Colorado River decision looms. Here’s how it will affect millions.With the river in freefall, federal officials plan to intervene, possibly imposing large cuts to water usage in seven states. Effects would be felt beyond the West.