botwerks

notes: here comes the sun

sulrich@botwerks.org (steve ulrich) — Sun, 30 Aug 2026 16:01:29 -0500

it’s not a doomburger, honest!

this was a really solid run at rallying the troops and driving a positive spin on the economic inertia behind renewable energy. i’ll be honest, it’s tough to read this as the trump administration is asserting control over venezuela’s oil reserves and aggressively paying companies to shut down wind projects that have solid funding and jobs behind them. that said, there’s no flaws in the economic logic of the argument that’s outlined here. the US as it prostrates itself to the petro cos is going to fall behind in many key metrics. the economic logic of renewable energy is, at this point, beyond unassailable. only morons, the 0wn3d or the truly disingenuous think that burning more fossil fuels is the right move. (and 2 / 3 of the previous list know the situation) as the US continues to fuck things up in the persian gulf, more folks will realize that they need to accelerate their country’s transition to maximal energy independence.

if you’ve been following energy developments with any conscious attention over the past 5-ish years, this book won’t be shocking to you. sure, the shape of the curve is a little surprising, but on balance, solar has gotten incredibly great and cheap. batter tech is ramping up at impressive rates as well. this book doesn’t really dig into batteries all that much but the cost and energy curves are comparable to solar and kind of shocking. (ha!)

there’s a fair amount to unpack in this from a policy engagement perspective and it’s struck me that we’ve been woefully under engaged as a populace in the PUCs. i’m not sure how to deal with left-wing NIMBY-ism and a lack of imagination here. but it’s there in spades. personally, i’m just waking up to the fact that there are a lot of retired folks with lots, and i mean lots, of time on their hands and they’re underutilized.

i bought this as the paperback re-issue and that seems about right from a personal value perspective. i would recommend this as a read to folks who are looking for a solid take on the status of renewable energy these days.

there’s plenty to be inspired about.

sources of happiness...

sulrich@botwerks.org (steve ulrich) — Thu, 27 Aug 2026 12:08:07 -0500

this is for future me. these are things that i’ve taken up that just make me happier. if future me is unhappy, check this list.

when i get in the truck don’t listen to podcasts - listen to music. listen to interesting and fun music. hell, listen to the shit you’ve been listening to since you were a pusshead teenager. podcasts will only raise the cortisol levels. i might learn something, but i’ll arrive at my destination irritated and music serves to tickle your heart.
read books, paper books. buy them used and don’t be precious about them. carry them around the house so you can read while cooking, etc. you used to do this pre-internet and the phone replaced it. FFS dude, you read infinite jest on the bus heading downtown in your early 20s. it might have even got you laid a couple of times.
- read fiction - not too much, but enough to amuse yourself and think about things differently. don’t listen to snobs and don’t be one.
- read poetry - you won’t get a lot of it; that’s ok.
put the books app prominently on your phone and use it when waiting. even mastodon can be a time suck, use things on screens in moderation or when you’re actively at a desk.
- corollary - don’t consume media which doesn’t have an end. this is the beauty of forums and mastodon. it’s not an infinite scroll, when you reach the end, you’ve reached the end. go do something else.
scribble notes on paper to help you lock things in your noggin. obsidian’s great and all, but when you take a beat and scribble something down, you’re exercising something very different. you can feel this, this is weirdly satisfying and i would contend it’s better for you.
get outside every damn day. having been mobility constrained for a couple of months, remember the giddy feeling you had when you just got outside the house and don’t take that for granted.

this is not a death house

sulrich@botwerks.org (steve ulrich) — Wed, 19 Aug 2026 14:23:21 -0500

Designing the Dream House of an 87-Year-Old Tech Visionary

here’s a heartwarming write up and video short about a house/condo that stewart brand and his wife built on their land in petaluma.

snow crash 30+ years later

sulrich@botwerks.org (steve ulrich) — Thu, 13 Aug 2026 20:55:45 -0500

i pulled snow crash off the shelf a couple of weeks ago. i’d been meaning to re-read this but … other things.

i have to say, this has not aged well. it’s still an amusing and fun read, but it’s definitely grounded in the milieu of the late 80s early 90s. this novel had left a real mark on me as a form of technological world building. including serious dalliances with VRML and multi-player games. my 20s were composed of way too many late nights getting the limited rendering capabilities of the dimension X toolkit going on primitive 3D accelerated hardware.

that said, chunks of this book haven’t aged well (the creepy relationship between raven and YT is #1 with a bullet.) and others were remarkably prescient. the franchise-ification of pretty much everything incredibly wealthy monopolists who are more powerful than governments and chock full of crazy ideas and have hordes of acolytes. there’s still a lot of good stuff in here.

the hoopla of the facebook 2 meta transition and the ready availability of VR headsets and hardware made it easy to forget in hindsight that the big thrust of the book was not the metaverse. it was more or less a useful sideshow and a model for how networked interaction might take place.

the far more interesting bits in the book were the discussions regarding language as a virus and transmission mechanism. the sumerian/babylonion discussions were bonkers and probably the most interesting diversions of the many that were in the book. the exploitation of language and media to assume power is really what’s at the core of this and this little nugget really hit the mark.

Basically, anyone who reads the National Enquirer or watches pro wrestling on TV is easy to convert. And with Snow Crash as a promoter, it’s even easier to get converts.

“Rife’s key realization was that there’s no difference between modern culture and Sumerian. We have a huge workforce that is illiterate or alliterate and relies on TV - which is sort of an oral tradition. And we have a small, extremely literate power elite the - people who go into the Metaverse, basically-who understand that information is power, and who control society because they have this semimystical ability to speak magic computer languages.0

you can see this playing out in the acquisition of TV networks and movie studios and then there’s that whole twitter thing. increasingly, it’s not just information that’s critical, it’s attention. this is the coin of the realm.

probably less than an hour after i finished reading snow crash, a link to the following rolled in front of me. it’s a threads thread on biblical plagiarism. it really looks like big hunks of the memorable bits of the old testament were lifted directly from sumerian myths.

are you shopping? probably ...

sulrich@botwerks.org (steve ulrich) — Wed, 12 Aug 2026 07:04:30 -0500

Shopping is Impossible

this is the running joke in our house. if the phone is out … “you must be shopping”.

there’s never been more retail, but retail is dead. we have more forms of X to consumer than ever before and oh, so. many. options. and yet, it all sucks.

seafood plunder

sulrich@botwerks.org (steve ulrich) — Mon, 10 Aug 2026 07:38:13 -0500

Plunder and Danger on Argentina’s Sea Shelf

this is one of the most fascinating things i’ve read in a while.

there are no moats; build the infra

sulrich@botwerks.org (steve ulrich) — Mon, 10 Aug 2026 07:14:07 -0500

last week was apparently dominated by gossiping about the departures from google. SemiAnalysis and plenty of others spilled a lot of metaphorical ink in the topic. the SA take, Gemini is Cooked but GCP is Cooking reeks of gossip and twitter tea leaf reading with their usual ass load of charts and wordy analysis. sometimes, this is a lot of fun. sometimes, it’s just tedious AF. this was tedious AF and really just felt like feed filler.

there’s a reasonable take here that google doesn’t see the need to pour money down the frontier training hole and by diversifying their bets with investments into the various AI model shops they’re reaping the best of both worlds. google’s always had the best tech stack. they saw into the TPU future well before the industry did broadly. it’s entirely possible and i would suspect quite likely that they’ve simply opted to focus on owning the infra layer and providing solid work-a-day models a la flash that are reasonably useful and competitive but don’t require the massive training costs up front.

there’s a whole finance angle to this discussion which is uh, interesting. who knows how these SPVs hold out but that’s apparently a fear for another day. there’s also a significant concentration of risk. the customer base can reliably be counted on a single hand with fingers left over.

spin the flywheel of money harder

sulrich@botwerks.org (steve ulrich) — Sun, 09 Aug 2026 07:28:09 -0500

The great wealth transfer reality check

here’s an interesting writeup from that great bastion of progressive thought, visa. a couple of things to take away from this article.

there’s not as much trickling down as folks might think there is. it’s going to be siphoned off at various stages in the process
there’s wealth transfer and propping up taking place right now
this will keep the economic disparities locked in for the next few generations
consumer spending bumps are likely to be modest. of particular interest to visa who’s keen to get a couple points of fees on all of those transactions.

suck it texas. you wish you had the stones of winona

sulrich@botwerks.org (steve ulrich) — Sat, 08 Aug 2026 08:10:46 -0500

All of WPD’s Flock cameras stolen | Crime & Justice | winonapost.com

well played winona.

file under: griftocracy

sulrich@botwerks.org (steve ulrich) — Sat, 08 Aug 2026 08:04:29 -0500

Trump administration to pay German firm to halt US wind projects

workism is for chumps

sulrich@botwerks.org (steve ulrich) — Sat, 08 Aug 2026 07:49:49 -0500

The Future Of Knowledge Workers In An AI-Driven World

this is a great writeup. there’s a fair amount that i’ve discussed with friends and coworkers over the years. having worked in tech for my entire career, i don’t even bother trying to explain to people what i do, it’s wholly uninteresting.

this was an interesting little blurb to me.

Workism is a microcosm of the larger spectacle: a world where appearing busy, important and uniquely knowledgeable is as valuable as actually being any of those things, and where work needs only to appear impactful rather than actually be impactful.

there’s also a component of this that has historically been american status assertion. i’m soooo busy. which was often shorthand for some sort of subtle claim of importance or workist criticality in the pipeline.

underlying all of this is the spectre of AI. personally, i do not feel particularly threatend by a lump of linear algebra. but it is creating new dangers and i can understand the anxiety and dread that this induces. we’re likely cooked. enjoy what you can, do what you must.

finally, some non-AI crimin' news

sulrich@botwerks.org (steve ulrich) — Fri, 07 Aug 2026 09:21:31 -0500

AI model achieves breakthrough in forecasting cyclones — Google DeepMind

this is some of the best AI related news i’ve seen in a while. too bad it’s lost in the noise of google staff departures and misc. political dreck.

the slide should have been evident long ago

sulrich@botwerks.org (steve ulrich) — Mon, 03 Aug 2026 08:08:22 -0500

How Google helped destroy adoption of RSS feeds - Open RSS

well, this was a romp through some bad memories.

your management fees will be paid in tokens

sulrich@botwerks.org (steve ulrich) — Mon, 03 Aug 2026 07:27:31 -0500

AI financial advice is surprisingly good — especially if you ask the right questions | MIT Sloan

odyssey translator's film review

sulrich@botwerks.org (steve ulrich) — Sun, 02 Aug 2026 21:12:24 -0500

‘The Odyssey’ Translator Emily Wilson on Nolan’s New Movie

i still haven’t found the time to read this but i’m somehow sucked into everything around it.

this is a great bit of writing and has all sorts of fun pointers around it.

in a country where corporations are people, this is not surprising

sulrich@botwerks.org (steve ulrich) — Sun, 02 Aug 2026 08:25:22 -0500

AI firms must answer for rogue bots, says Hugging Face boss

LLMs these days … amiright?

premium dreck

sulrich@botwerks.org (steve ulrich) — Sat, 01 Aug 2026 17:40:53 -0500

dear apple -

it’s (checks watch and setting » general » about) iPadOS 26.6 …

what is this shit?

oh? you think?

sulrich@botwerks.org (steve ulrich) — Sat, 01 Aug 2026 17:11:57 -0500

GOP Staffers Say the Party Has a Groyper Problem - POLITICO

periodically, you must be prepared to burn it all down.

sulrich@botwerks.org (steve ulrich) — Fri, 31 Jul 2026 08:37:14 -0500

SlopCodeBench: Benchmarking How Coding Agents Degrade Over Long-Horizon Iterative Tasks

someone finally brings this topic to the fore. even small vibe coded projects can be horrifying under the hood.

nice test coverage, but whoa!

kimi 3

sulrich@botwerks.org (steve ulrich) — Fri, 31 Jul 2026 07:56:13 -0500

quick personal observations on kimi 3

there are no moats. kimi k3 running at a fraction of the price of claude is impressive. i wired it up to claude-code yesterday and gave it a reasonable task to refactor some code based on some consolidation that i had spelled out in a couple of paragraphs of prose. it ripped through the work in about 20 seconds. now whether this is a function of the model or the hosting (at fireworks.ai) or perhaps compute constraints at anthropic, i can’t realistically say. i would have expected claude to dither about on the changes for a minute or two.

there’s something i can’t quite put my finger on re: kimi 3 - it “feels” fresher and less encumbered then the interactions with claude do. i don’t know if that will change or if there’s just accretion and baggage that i’ve now developed with claude. i also suspect i need to give claude a dusting up with the advent of the model updates.

this dude might be channeling what i’m thinking/feeling here.

i’m less bullish on the american frontier model labs than many of my peers. i increasingly find myself more firmly in the camp of those who want high quality open weight models that can be run on a variety of providers. deepseek v4-pro was an impressive move in this space and kimi 3 is a significant jump from there. i’ll likely be poking at this for more of my personal projects.