iiinova — where your digital self keeps evolving

verdverm @verdverm.com · 19/03/2026

Can we get a switch in moderation settings for them? Like we do for adult content?

verdverm @verdverm.com · 18/03/2026

My rough calculation for "accounts" and their "status" as of ~2026-03-11 (it takes many hours to fill, so some slippage expected)

010

verdverm @verdverm.com · 18/03/2026

Rough counts (empty is have a valid account, ignore "Active") This is based on calling the getStatus endpoint for every account returned by the PDS, list from Mary. FYI the active field on the list endpoint is not the same as active in the getStatus endpoint, which gives you more details

050

verdverm @verdverm.com · 18/03/2026

bskycharts.edavis.dev/edavis.dev/b...

100

verdverm @verdverm.com · 18/03/2026

That's before accounting for takedown and deleted, which as you have noticed, brings the number way way down Third nit, this is on official copy

110

verdverm @verdverm.com · 12/03/2026

A big issue the party does not want to acknowledge or address. If they weren't so senile, they might understand why we feel they don't represent us by looking at this one weird chart.

Graph of House reps over 70 by party, currently

Dems 56
Repbs 30

040

verdverm @verdverm.com · 09/03/2026

Court filings for Anthropic v Government They are just sueing anybody it seems :] www.courtlistener.com/docket/72379...

000

verdverm @verdverm.com · 06/03/2026

blog.system76.com/post/system7...

The challenges we face are neither technical nor legal. The only solution is to educate our children about life with digital abundance.

010

verdverm @verdverm.com · 05/03/2026

My choice clips, this piece is really good

The Doom Loop and Getting Out of the Race, title + full section text

Building With Clay but No Kiln, title + two paragraphs

030

verdverm @verdverm.com · 04/03/2026

Which models are you using and finding the best success with? I'm still working out kinks on my printer, the square / flat ones seem to do much better, but not sure if they are loud enough. Flounder & McFry in this image. The Sockeye turned out ok, but took longer to print, still have a few to try

Image of 15 whistles, 3d printed, with code names

100

verdverm @verdverm.com · 03/03/2026

A Rational Analysis of the Effects of Sycophantic AI subtext, being careful with what you ask for is not enough arxiv.org/abs/2602.14270

Analyzing How Sycophancy Distorts Beliefs

We propose sycophancy leads to less discovery and overconfidence through a simple mechanism: When AI systems generate responses that tend toward agreement, they sample examples that coincide with users’ stated hypotheses rather than
from the true distribution of possibilities. If users treat this
biased sample as new evidence, each subsequent example increases confidence, even though the examples provide no new
information about reality. Critically, this account requires no
confirmation bias or motivated reasoning on the user’s part.
A rational Bayesian reasoner will be misled if they assume
the AI is sampling from the true distribution when it is not.
This insight distinguishes our mechanism from the existing
literature on humans’ tendency to seek confirming evidence;
sycophantic AI can distort belief through its sampling strategy, independent of users’ bias. We formalize this mechanism
and test it experimentally using a rule discovery task.

021

verdverm @verdverm.com · 03/03/2026

For the other 364d 23h you aren't voting www.sherwoodforestzinelibrary.org/_files/ugd/8...

000

verdverm @verdverm.com · 02/03/2026

Does corporate always forget about The Streisand Effect?

Mecha Streisand from early South Park

"Microslop! Microslop!"

060

verdverm @verdverm.com · 02/03/2026

Google published an amazing book of essays on SRE (site reliability engineering), three actually, and free to read sre.google/sre-book/tab...

Screenshot of the three SRE books by Google

020

verdverm @verdverm.com · 02/03/2026

I think I need a bigger stapler for making #zines

Image with a zine titled "Moral Revolution!" with a tiny stapler eating the half-page zine. In the background are stickers of the American Flag and "FREAD" (OM is unseen around the corner), "read" is in red

000

verdverm @verdverm.com · 01/03/2026

How To Make A One-Page Zine Just like this one! www.sherwoodforestzinelibrary.org/_files/ugd/8...

A picture of the folded zine with the pdf on a screen in the background

030

verdverm @verdverm.com · 01/03/2026

The situation I'm describing should be inferred and not require ENV vars. Can it take advantage of nested MD files in a single repository?

find . -name AGENTS.md with ~20 results for a single git repo

100

verdverm @verdverm.com · 01/03/2026

beabetterdev.com/2026/03/01/a...

intro text from the post at the other end of the link

062

verdverm @verdverm.com · 01/03/2026

Can we reproduce this without the author, i.e. are the paper details sufficient?

Reproducibility
The Helix framework was developed as a proprietary extension for internal use at Verificate Pty Ltd. The
core implementation, including the custom C++ modifications to llama.cpp and certain API integrations, is
not publicly released due to proprietary constraints and security considerations. However, the pre-computed
truthfulness manifold files, non-proprietary evaluation scripts, detailed hyperparameter settings, and step-bystep reproduction guidelines for the reported experiments are available upon request. Qualified researchers
may contact the corresponding author at info@verificate.ai with a brief description of their intended use.
Access will be granted on a case-by-case basis under non-commercial, non-redistribution terms.

130

verdverm @verdverm.com · 28/02/2026

support artists jmail.world

000

verdverm @verdverm.com · 23/02/2026

this is a personal fav b/c Elon is a fan of Monty Python, I'm petty too

Monty Python dead bird skit with Elon and the Twitter bird with an "X" for an eye

020

verdverm @verdverm.com · 22/02/2026

> They thought of it the same way they thought of autocomplete or spell-check - a thing their phone just did now.

Each company’s individual response was rational. The collective result was catastrophic. Every dollar saved on headcount flowed into AI capability that made the next round of job cuts possible.

Software was only the opening act. What investors missed while they debated whether SaaS multiples had bottomed was that the reflexive loop had already escaped the software sector. The same logic that justified ServiceNow cutting headcount applied to every company with a white-collar cost structure.

When Friction Went to Zero
By early 2027, LLM usage had become default. People were using AI agents who didn’t even know what an AI agent was, in the same way people who never learned what “cloud computing” was used streaming services. They thought of it the same way they thought of autocomplete or spell-check - a thing their phone just did now.

120

verdverm @verdverm.com · 22/02/2026

susam.net/attention-me...

000

verdverm @verdverm.com · 21/02/2026

are we partying or are we "partying"?

110

verdverm @verdverm.com · 20/02/2026

flawless logic, no notes for TFG cc @spockresists.bsky.social for the lulz

Evil Spock from the mirror world in Star Trek TOS

000

verdverm @verdverm.com · 20/02/2026

Apparently @kkomaitis.bsky.social cannot take negative feedback, for posterity

000

verdverm @verdverm.com · 20/02/2026

You donate to Trikki Nikki once in the primaries when it looked like she might beat the monster, and years later you are still unable to get off their lists... You'd think with all the public, anti-everything about them posts here, they'd have figured it out by now. Even super unsub doesn't work

Donation text message with STOP, FUCK MAGA, FUCK TRUMP, FUCK MILLER, FUCK ICE and three "You have successfully unsubscribed." replies

Even this super unsubscribe doesn't work

110

verdverm @verdverm.com · 18/02/2026

New toy just arrived

000

verdverm @verdverm.com · 18/02/2026

This was the first one that caught my eye because I'm pretty sure I know what the bean counters are going to do... Ai is certainly a "shock to the system", perhaps the 4th wave (alvin toffler extended?). I think Ai might have been in the 3rd wave, but I think it deserves its own

Is the software I’m making for myself on my phone as good as handcrafted, bespoke code? No. But it’s immediate and cheap. And the quantities, measured in lines of text, are large. It might fail a company’s quality test, but it would meet every deadline. That is what makes A.I. coding such a shock to the system.

120

verdverm @verdverm.com · 18/02/2026

by the way, those rules of thumb, they are PER MONTH A typical car for context is about 400 trees per month www.epa.gov/greenvehicle...

How much tailpipe carbon dioxide (CO2) is emitted from driving one mile?
The average passenger vehicle emits about 400 grams of CO2 per mile.
What is the average annual carbon dioxide (CO2) emissions of a typical passenger vehicle?
A typical passenger vehicle emits about 4.6 metric tons of CO2 per year.
This assumes the average gasoline vehicle on the road today has a fuel economy of about 22.2 miles per gallon and drives around 11,500 miles per year. Every gallon of gasoline burned creates about 8,887 grams of CO2.
Are there other sources of greenhouse gas (GHG) emissions from a vehicle?
In addition to carbon dioxide (CO2), automobiles using gasoline produce methane (CH4) and nitrous oxide (N2O) from the tailpipe and all vehicles can emit hydrofluorocarbon (HFC) from leaking air conditioners. For gasoline vehicles, the emissions of HFCs are small in comparison to CO2; however, the impact of these emissions can be important because they have a higher global warming potential (GWP) than CO2. Electric vehicles (EVs) also emit a small amount of GHGs due to air conditioner/HFC leakage.

110

verdverm @verdverm.com · 18/02/2026

Rule of thumb 👍 to offset pet CO2 by planting trees 🌳 30 trees per cat 80 trees per dog onetreeplanted.org/blogs/storie...

How Much Carbon Dioxide Do Trees Absorb Per Year?
To determine the amount of carbon dioxide a tree can absorb, we combine average planting densities with a conservative estimate of carbon per hectare to estimate that the average tree absorbs an average of 10 kilograms, or 22 pounds, of carbon dioxide per year for the first 20 years. This was determined by reviewing the planting density of our projects around the world, with an average planting density rate of 1,000/trees per hectare.

With carbon estimates, we like to err on the side of being conservative to avoid over inflating the potential benefits, which is why we selected 10 tons per hectare per year. Dividing 10 tons per hectare by 1,000 trees per hectare gets us an average value of 10 kilograms/22 pounds per tree per year. This is a little less than half of the commonly cited 48 pounds per tree per year, which gives us confidence that we are in the same ballpark as the accepted, de facto number, but are now being more conservative, and have a way to back it up.

There are verifiable carbon accounting methodologies that we fully support, like the Gold standard. These are often only suitable for large projects, and require additional costs and processes that are not applicable to all projects. The 10 kilograms per year is meant as a starting estimate to help us understand the potential carbon impact on a per tree basis.

121

verdverm @verdverm.com · 18/02/2026

You're telling me that Unitree H2, those dancing robots everyone is talking about, can be yours for $30k ummm... had a CD just roll over... I wish I could justify this

Diagram and spec sheet for the H2, highlighted is a price of ~$30k

010

verdverm @verdverm.com · 18/02/2026

We've reached new levels of absurdity #ai #openclaw #hackernews

a bot posting the actual openclaw as if it were its own

120

verdverm @verdverm.com · 18/02/2026

context for your #climate change and environmental concerns, have you considered the comparative impact of your pets? earth.org/environmenta... (with tips to improve it)

Climate change affects everything and everyone, including companion animals. Heatwaves and rising infectious diseases from pathogens caused by climate change harm the pets we love so dearly – albeit in a much less significant way than most human activities and industries, the pet industry also partly contributes to increasing heatwaves and pathogens.

An average-sized cat can produce 310 kilograms (CO2e) annually. An average-sized dog generates 770 kg of CO2e, and an even bigger dog can emit upwards of 2,500 kilograms of CO2e, which is twice as much as the emissions deriving from an average family car per year.

Fortunately, raising a pet does not have to be environmentally harrowing for the planet, nor do we need to abandon societal’s fondness for owning a companion animal altogether. If we are already taking steps to become more sustainable in other industries like business and fashion, the pet industry (and ownership) can indeed follow suit.

140

verdverm @verdverm.com · 18/02/2026

This "risk" tax for #ai sounds like a pretty great idea for #safety www.foreignaffairs.com/united-state...

Embrace the Tradeoffs
The best route forward for AI regulation is to embrace two compromises, each of which involves a modest economic cost and a larger gain for societal safety. The premise of the first compromise is that model safety is both a private and a public good. AI labs have incentives to produce models that are safe for their users, creating a private good. But a model that is harmless to its users can still harm public safety. For example, if a user asks a model to generate and distribute thousands of subtly distinct articles alleging an imaginary election fraud, the AI lab’s private incentive is to please that customer by complying—but that very compliance would harm the public by spreading disinformation. Because private labs lack incentives to avoid such externalities, they will not adequately invest in safety research.

A special “risk tax” could correct this shortfall. The goal would not be to raise revenue but to influence how companies allocate their resources, shifting more of them to safety research. For example, an AI developer that spent $1 billion to build a model would face a five percent tax, creating a $50 million liability. To offset this, the government would offer a tax credit worth 25 percent of each dollar the firm allocated to safety research. If the lab spent an additional $200 million on safety research, the resulting $50 million credit would offset its tax bill.

210

verdverm @verdverm.com · 18/02/2026

The Practicality Test for pro #ai governance and regulation, i.e. focus your efforts and spend your political capital where it matters www.foreignaffairs.com/united-state...

The Practicality Test

Having grasped the AI trilemma and sidestepped the singularity delusion, advocates of AI governance must complete one final exercise. They must be blunt about the fact that some popular misgivings about AI should not be policy priorities. The clearest examples involve concerns that are exaggerated and problems that the private sector is motivated to solve, rendering regulatory pressure unnecessary.

010

verdverm @verdverm.com · 18/02/2026

It's a really, really good interview, you should watch it I donated $10 to James Talarico afterwards

011

verdverm @verdverm.com · 18/02/2026

Fresh polling on #ai usage, highlights - roughly 70% have tried with 30% regularly, if we remove retirees - far more mistrust, surprising no one - split on economic impact ~50/50 - job outlook heavily negative d3nkl3psvxxpe9.cloudfront.net/documents/ec...

131

verdverm @verdverm.com · 17/02/2026

dragons are just the biological boot loader to draigons

Issei's full dragon suit presented as a "draigon"

000

verdverm @verdverm.com · 17/02/2026

I love that jmail.world is still pushing new features #epstein

000