Tootfinder

Sources: OpenAI is in advanced talks to hire OpenClaw founder Peter Steinberger and several team members; Steinberger would likely work on personal agents (Nick Wingfield/The Information)
https://www.theinformation.com/briefings/openai-advanced-talks…

OpenAI in Advanced Talks to Hire OpenClaw Founder, Team
OpenAI is in advanced talks to hire Peter Steinberger, the founder of the popular open-source personal agent project OpenClaw, and several team members. The team would likely work on personal agents at OpenAI. They are also discussing setting up a foundation for the existing OpenClaw project. Meta Platforms is also aggressively courting Steinberger. Steinberger said he has not made a final

This is NOT a formal job posting, just testing waters. I have a year of post-doc money. Esp. int'd in formal methods applied cogsci diagramming. If you do work tied to my research, reach out (see my page). Must have US work auth, sorry. Please feel free to share/boost!

Stories of AI "accelerating" work describe problems as though they were solutions. "This PM had an LLM generate their product strategy." "This designer shipped a prototype with a backend running in Google Sheets." "This engineer no longer reads the code they release."
Releasing things is not an unalloyed good. "Build to learn" only works if you can learn something - and overwhelming your org's ability to absorb information is NOT th…

Designers will never have influence without understanding how organizations learn
We confuse prototypes with demos, and validation with confirmation bias. As a result, we cannot lead — instead, we are led.

The existence of Israel is a manifestation of European antisemitism, both historically justifying the project of Zionism and European governments continuing to fund Israel instead of addressing the core role of antisemitism in maintaining (at least) the conservative elements of the neoliberal order.
If antisemitism props up Israel, then a core part of dismantling Israel (and thus saving the lives of Palestinians) is addressing European antisemitism. There are Israeli Jews who would leave if they felt safe to do so. There are Jews demanding Israel be armed, because they don't feel safe anywhere else.
Protest, boycott, take whatever action you feel is appropriate. There is a limit to your ability to convince governments to stop funding genocide, but you can learn about antisemitism and you can work to fight it, especially within "The Left." You can learn to distinguish between legitimate critiques of Israel, and antisemitic ones, and you can stand and call out antisemitic ones.
Honestly, this is some of the easy work that I think a lot of people don't consider even doing, don't even realize it is a thing that can be done.
I'm talking to Israeli folks who identify as being on the Left. It's hard because they've been through a lot of propaganda. Israel is a cult that terrorizes its members. This is important work that can have a huge impact, because it focuses on dismantling the networks of support that reinforce what's happening now.
Meanwhile, I still run into wild things like "Rothschild" conspiracy theories among people who identify with the idea of supporting Palestinians.
Not only can you support the liberation of Palestine and the Palestinian people while opposing antisemitism, but you must actually do both in order to do either. They are exactly the same fight, and anything short of both is thrashing against oneself.

So to follow up on this, I've caught it in action. Models, when quantized a bit, just do a bit more poorly with short contexts. Even going from f32 (as trained) to bf16 (as usually run) to q8 tends to do okay for "normal" context windows. And q4 you start feeling like "this model is a little stupid and gets stuck sometimes” (it is! It's just that it's still mostly careening about in the space of "plausible" most of the time. Not good guesswork, but still in the zone). With long contexts, the probability of parameters collapsing to zero are higher, so the more context the more likelihood you are to see brokenness.
And then at Q2 (2 bits per parameter) or Q1, the model falls apart completely. Parameters collapse to zero easily. You start seeing "all work and no play makes jack a dull boy” sorts of behavior, with intense and unscrutinized repetition, followed by a hard stop when it just stops working.
And quantization is a parameter that a model vendor can turn relatively easily. (they have to regenerate the model from the base with more quantization, but it's a data transformation on the order of running a terabyte through a straightforward and fast process, not like training).
If you have 1000 customers and enough equipment to handle the requests of 700, going from bf16 to q8 is a no-brainer. Suddenly you can handle the load and have a little spare capacity. They get worse results, probably pay the same per token (or they're on a subscription that hides the cost anyway so you are even freer to make trade-offs. There's a reason that subscription products are kinda poorly described.)
It's also possible for them to vary this across a day: use models during quieter periods? Maybe you get an instance running a bf16 quantization. If you use it during a high use period? You get a Q4 model.
Or intelligent routing is possible. No idea if anyone is doing this, but if they monitor what you send a bit, and you generally shoot for an expensive model for simple requests? They could totally substitute a highly quantized version of the model to answer the question.
There are •so many tricks• that can be pulled here. Some of them very reasonable to make, some of them treading into outright misleading or fraudulent, and it's weirdly hard to draw the line between them.

"LLMs work (somewhat) for coding computer programs. As everyone knows this is the highest form of human endeavor—unsurpassed by any other lesser activity such as project management, design, art or writing. Therefore LLMs will excel in every other field."
I really believe this is the crux understanding why so many programmers (including good programmers) fall for it in a way that can only be described as a cult, were any criticism is not only not allowed but reflexively is seen as either laughable or belligerent.
Anyway, LLMs are good* at writing code because writing code is easy and highly repetitive and doesn't actually take a lot of skill; unless it's novel ways to write code which LLMs cannot do.
Taking this as a sign LLMs can do other "lesser" activities is saying a lot about the hubris of programmers and not a lot of the capabilities of LLMs.
*for some definitions of "good"

For those who have been hearing of a Fly Brain being uploaded, the work you've been hearing of is impressive, yet as always the pop science media has warped what happened a bit:
„The Fly Brain Breakthrough Is Real. The “First Brain Upload” Narrative Is Not.”
https://…

The Fly Brain Breakthrough Is Real. The “First Brain Upload” Narrative Is Not.
Scientists mapped a fruit fly’s 139,255 neurons in extraordinary detail. Then a startup turned that milestone into a story about brain…

Does the Bandwagon.fm player not work on iOS?
#bandwagonfm #music #indie

How is there no space for people in the humanities who actually work with *and* interrogate AI? not knee-jerk condemnation and not fanboy acceptance, but using the tools and thinking through their consequences. so strange.

Only work-product of 45’s DoJ task force about weaponization of DoJ by Biden is a report with thesis Biden DoJ enforced the FACE Act (meant to stop attacks or blocking health clinics) unevenly, focusing mainly on anti-choice people/orgs.
On page 17 it points out how the Biden DoJ enforced evenly, and intended to keep the law in place. They unmake their thesis (and the Task Force’s purpose). The current DoJ has convinced me Biden’s DoJ was not weaponized.

Advice if you build websites: Use the W3C HTML *and* CSS validator. Not because I want to tell you that "you're bad if your HTML isn't valid", but because it'll help you find trivial bugs that can make your life hard.
Especially also use the CSS validator. I found 3 issues that made things not work as they should in recent days. I had a bug where it took me a while to figure out why it didn't work til I found some stackoverflow post, and the css validator would…

RE: https://indieweb.social/@web3isgreat/116403486703540080
“Highly curated” is hilarious given we know Apple’s app store is not thanks to work from CMA under DMA.
I hope Apple makes those people whole after its failure to honor its promise to cu…

Paramount responds to Hollywood creatives' letter against its WBD acquisition, reiterating its commitment for 30 theatrical feature films annually, and more (Ted Johnson/Deadline)
https://deadline.com/2026/04/paramount-warner-bros-merger-creatives-…

Paramount Responds To Letter From Hollywood Figures, Says WBD Merger Will Ensure Creatives Have “More Avenues For Their Work, Not Fewer”
The company argued that the Warner Bros. Discovery merger will create "more avenues" for work, not fewer.

I do love it when I come across people online who I've not seen in AGES - reporting a bug and it getting fixed by someone you used to work with decades ago, in an unrelated job.

Most developers now use coding assistants. I do too—#Copilot at work, #ClaudeCode at home. As a developer, I prefer not to repeat myself. This post explains why and how to avoid repetition with skills.
#DRY

CS researchers: NSDI 2027 explicitly declines to include work in some topics that are technically in scope (search for "Moreover, NSDI does not include work in…"). Are there other elite conferences that are also doing this? Pointers welcome, thanks!
https://www.usenix.org/conference/nsdi27/c

NSDI '27 Call for Papers
The 24th USENIX Symposium on Networked Systems Design and Implementation (NSDI '26) will take place May 11–13, 2027, at the Omni Providence Hotel in Providence, RI, USA. NSDI focuses on the design principles, implementation, and practical evaluation of networked and distributed systems.

I’m a stationery fetishist, just not a calendar guy. Maybe that’s because of my tendency to consistently struggle with categories. 😶‍🌫
https://thegoodwork.blog/posts/i-love-to-write-on-paper-but-i-cant-deal-with-calendars

I love to write on paper but I can’t deal with calendars - The Good Work
After quite a few failed attempts, I’ve decided that analog calendars just aren’t a thing for me. Not for the short, medium, or long term. It’s not the calendars’ fault,...

The 22 is really not built as a cross-town commuter line, but so much of the RT office work is now in Mission Bay, creating a really slow and packed ride.

This looks like an interesting take on the whole “AI" (IMHO its not AI and its not leading directly to AGI).
I've only watched this preview but it looks like it covers things from many points of view. The creator of the film
The AI Documentary: Or How I Becamean APOCALOPTIMIST.
#AIHype

This is why I don't fully trust 48hills, even though they're on "my side" in being pro-labor and anti-billionaire. Tim Redmond thinks it's a good thing that a 100% affordable housing development was forced to lop a floor off to preserve views 😫
"The staff reached out to the neighbors,[...]took feedback, and as a result made some changes that lowered the height of the new buildings. That, in a reasonable world, is how this should work" No it's not!!…

Supes approve affordable senior housing project—but discussion raises larger issue - 48 hills
Nonprofit housers are good neighbors. Speculators are not. But Wiener's law treats both the same way

One of the best features of GrapheneOS is that the Twitter app does not work at all: fails attestation, won't log in, and softblocks your account for suspicious activity. More folks should switch!

I've been wondering lately if my job is #bullshit.
I've given it a lot of thought, and I think it's not directly bullshit. I'm doing stuff that's meaningful, at least in a narrow scope, both in my dayjob and my #FreeSoftware / #Gentoo work.
That said, with the arrival of all the bullshit CEOs, CTOs, all their bootlickers, wannabe bootlickers, and all the CEO/CTO/bootlicker cosplayers, the whole software industry is becoming filled with bullshit to the brim.
Even if my work is meaningful, it contributes more and more to software that's either scam in itself, used to scam people or pure unadulterated bullshit. Even if the tools used to be useful, they either gain bullshit parts or bullshit dependencies.
I hate this, and it's making me hate what I'm doing.

This week's career limiting accessibility regression from @… is to not show comments in pull request diff any more. There's a separate comment panel which you can look at (but not at the same time as the diff) where there are indications of content replies but these cannot be interacted with using standard accessibility tools.
I review very long pull requests for work and as a volunteer maintainer and I am not sure why we're okay with this constant march of degraded experience. #a11y

The Senate returns to work today, while the House will hold a brief procedural session before getting back to regular business on Tuesday.
Lawmakers have still not passed a funding bill to reopen the Department of Homeland Security (DHS) subagencies affected by the record-breaking partial government shutdown, now in its ninth week.
During the two-week recess, House Republican speaker Mike Johnson took no action to advance a Senate-passed measure that would reopen agencies like th…

This is exactly the groundwork we need to be laying right now, and I’m delighted to hear it coming from a US Senator (!):
If Trump tries to prevent a fair election from happening, then we need “a true national strike in the sense that, if they do this, if they try to overthrow our democracy, if you are allied with democracy, do not go to work . If you’re a pilot, do not show up. If you drive a train, do not show up. If you’re a teacher, do not show up. We grind the country to a halt.”
https://www.theguardian.com/us-news/2026/feb/06/donald-trump-voting-midterms-democrat-national-strike
1/

Fresh out of the shower (ew) staffer stands in front of mirror at 5:30am, almost chanting, “I’m good enough, maga enough, have had enough work done, he likes me, he really likes me. He let me destroy evidence and didn’t deport me when I was a witness. I got this. Just brush my teeth with Ivermectin, spray hydrochloroquine behind the ears. Tough on crime, not our crimes. Who’s the best staffer, I am.”
Then, full chant, “we’re winning the war we’re losing, we’ve lost the war we’ve won.”

AI Fails at 96% of Jobs (New Study)
Artificial intelligence has been hailed as one of the most transformative technologies of the century. That may be so, but just not yet. In this episode, we take a look at a study that pits humans directly against AI for paid work. The results were surprising.
📺 https://www.youtube.com/watch…

Graney: Klint Kubiak wins people over with hard work, not words https://www.reviewjournal.com/sports/sports-columns/ed-graney/graney-klint-kubiak-wins-people-over-with-hard-work-not-words-3618343/

The other day, my friend shared this video on consulting collectives, which is when "a small group of independent consultants or freelancers share clients, resources, and revenue on specific projects while maintaining their independence on others" :
https://www.instagram.com/p/DUDtOqGkTmY/…

JANEL ABRAHAMI | Career Advice for the Future of Work on Instagram: "Are you already doing this? How does it work for you? ⚡️ Consulting Collective (noun): when a small group of independent consultants or freelancers share clients, resources, and revenue on specific projects while maintaining their independence on others. Like the Avengers, but for client work. The solo consultant life is lonely and limiting, because you can only take on so much work, you have no backup when you’re sick or on vacation, and big clients often want teams, not individuals. But going back to a traditional agency or corporate job means giving up your autonomy + flexibility. So, consulting collectives are the clever middle ground no one’s talking about. Here’s how they actually work. @freelancingfemales @dreamersdoers @linkedinnews @entrepreneur @entreprenistas"
55K likes, 1,355 comments - janelabrahami on January 28, 2026: "Are you already doing this? How does it work for you? ⚡️ Consulting Collective (noun): when a small group of independent consultants or freelancers share clients, resources, and revenue on specific projects while maintaining their independence on others. Like the Avengers, but for client work. The solo consultant life is lonely and limiting, because you can only take on so much work, you have no backup when you’re sick or o…

“So I guess what I’m trying to say is, the new workday should be three to four hours.“
Yup! That's what every worker knows and should have been fighting for with solidarity for decades. Every neurodivergent person knows that we can't do concentrated work for more than 3 hours, and that extended hyperfocus blocks drain our energy for the next day. It's not sustainable.
Steve Yegge writes about how AI Capitalism creates an energy vampire

The AI Vampire
This was an unusually hard post to write, because it flies in the face of everything else going on.

I connected my Samsung TV to the internet to try for wireless display and casting from Android working and I regret it instantly.
Worse, there's gamepads ads and Ramsey giving me the middle finger on my home screen. It split my 4 inputs into two different tabs: one's a new games tab to push me into using their cloud gaming apps. No option to move consoles back to the main tab.
Wireless display to play videos from PC is nice, but casting doesn't work. Not worth it and …

Want a Masterclass in How Not to Work With Artists? Ask Google
https://

Want a Masterclass in How Not to Work With Artists? Ask Google | Artnet News
Renown artist Judy Chicago shares her misadventures with an ambitious public art commission for the tech giant.

Every now and then, Vivaldi.. on my work computer - and only on my work computer - flashes and rebuilds its whole UI. Like the whole window goes blank for a second then its back.
Which _totally_ doesn't make me think they have something watching / scraping what Im doing. :/
edit: and not Vivaldi I mean, Im willing to bet its something work has installed on here.

#Researchers or other professions who read articles for your work - which do you find better for focus, information processing and memorisation: reading on an electronic medium (computer, tablet, phone...) or on actual paper?
I thought I liked the electronic format but I'm wondering if it makes it harder to focus or remember things..
(This is only for work-based and not leisure…

For the first time, 50% of employed US adults say they use AI at work a few times per year or more; leaders are more likely to see AI's impact as positive (Andy Kemp/Gallup)
https://www.gallup.com/workplace/704225/rising-adoption-spurs-workforce-chang…

Rising AI Adoption Spurs Workforce Changes
Half of U.S. workers now use artificial intelligence. AI adoption links to organizational disruption and individual productivity gains but not transformational changes to work.

Work in progress website mockup. Not sure if I'll ever get to finishing this or turning it into code, but I always enjoy working on web design stuff :)

🧺 Device that can extract 1,000 liters of clean water a day from desert air revealed by 2025 Nobel Prize winner
https://www.

Device that can extract 1,000 liters of clean water a day from desert air revealed by 2025 Nobel Prize winner — claimed to work in desert air with 20% humidity or lower, delivering off-grid ‘personalized water’
The system can function in air with 20% humidity or less. But these 1,000 liter a day machines are not small, at around shipping container size.

Machine Learning techniques are upending multiple scientific fields. Operational 5-day forecasting of air quality in 1 minute in this paper from Chinese researchers.
This is awesome work with very clear public health implications.
EDIT for clarity: I am.not suggesting LLMs have anything to do with this work, but many people hear AI and imagine LLMs. And many of them.are perhaps rightly sceptical of AI as a result.
But AI or ML techniques can be useful for lots of things, not just chatbots. And we should probably invest more in those.

https://www.nature.com/articles/s41586-026-10234-y

Please, for the love of all that is holy in this world, Microsoft, please fix Defender so that it is not constantly running, or constantly running when I turn on my machine, or whatever. I could handle a 15-minute suck of all my system resources once per day, ideally outside a primetime window, but whatever you're doing is ridiculous. I can't work.

Still too damned many not showing up. Left work at 15:30 should be at home at just after 16:00. Arrived at home at almost 17:00!
https://www.ctvnews.ca/ottawa/article/oc-transpo-cancels-fewer-trips-as-more-buses-hit-the-roads/

Hey @… any idea why nvidia-kernel-dkms (on trixie at least) does not depend on linux-headers-$arch?
The end result of installing it without kernel headers is that the driver doesn't work, because it can't compile without the headers. I can't think of any situation in which this is a desirable state

'Not enough' work being done in Kyiv as city faces heating emergency, Zelensky says: https://benborges.xyz/2026/02/01/not-enough-work-being-done.html

I’m at a spa trying to get my shit together in a 90ºC (~194ºF) sauna after my 2025 was beautiful madness but madness no less.
I’ve finally started to write a new newsletter & if you wanna know more about the madness or what’s coming up or a cute dog photo, sign up at https://buttondown.com/hynek

Hynek Did Something
Greetings friend! This is a list for those who’d like to be notified whenever I, Hynek Schlawack, do or publish something noteworthy, somewhere. Examples are new blog posts, YouTube videos, and talks, significant updates to existing work, or if I need to be bailed out of jail. It does not include announcements about new software releases. But it does usually include a cute dog photo. As a thank you for your trust, the emails aren’t just links to content. They contain a kind of “director's…

Helion has just announced they got D-T fusion working in their reactor; this is pretty big; they're not claiming break even yet - but Helion is one of the odder (non-tokamak) designs which is much further along than most others, but weird enough that no one is too sure if they'll get it to work - the next important thing is to see if they get their direct energy recovery to work;

How we conducted and measured D-T fusion
Helion’s first D‑T fusion campaign marks a major step forward in our mission to build practical fusion power plants.

I’m actually shocked that folks I speak to here in Ireland are still planning holidays in the USA.
Here’s a testimonial you might not find on Tripadvisor:
“like a concentration camp, absolute hell” – Seamus Culleton
https://www.theguardian.com/us-news/2026/f

Fiber broadband tariffs work as cross-subsidies. Every higher plan helps fund the smaller, more affordable ones. So if you can, just click that upgrade. To my knowledge, not a single end customer provider is currently profitable with fiber. Definitely true for Germany, but probably for other countries as well.
#Internet

So, something from my car's buildlog made it into Google's AI summary (yes, it links my buildlog as the source and and only source for this)...
Problem for me isn't that Google uses it (that debate can go on for ages, not worth my time), the problem is that this "fix", isn't actually a fix.
It was a band-aid fix to make something work "alright-ish" for me to get certain usable data but isn't actually a proper fix.
It's incorrect in…

SO!
I can fix my code so that the arguments to FEXPRs are not evaluated, by deferring the call to EVLIS into APPLY; but this is not what the code given on pages 70-71 does, and so my code will then not directly follow the specification in the #Lisp 1.5 Programmer's Manual;
OR,
I can not do this, in which case FEXPRs will not work as specified on pages 18 and 19.
(Note that…

This is the part that commentators need to understand down does not work when you're dealing with narcissists:
"Maybe you could have just asked your husband to get that done, Melania? I hear he’s pretty high up in politics."
https://www.theguardian.com/commentisfr…

Mailbag: Players' involvement in contract talks? https://www.dallascowboys.com/news/mailbag-players-involvement-in-contract-talks

"Personal Knowledge Management is for Life, Not Just for Work"
#PKM

Not reproducible, however I have a VirtualBox snapshot of the bugged installation.
I can restart SDDM and login, again I get a black screen. Xorg.0.log has one error:
open /dev/dri/card0: No such file or directory
Maybe negligible. The same error occurs when I successfully startx and use twm instead of Plasma (X11).
'reboot -r' is not enough to work around the issue.
Plasma did work, once, after 'shutdown -r now', however this is not a consis…

"Reducing aircraft soot might not actually reduce the climate effects of contrails"
#Aviation #Emissions #Climate

Reducing aircraft soot might not actually reduce the climate effects of contrails
Reducing aircraft soot emissions may not reduce contrail clouds, according to in-flight observations of emissions from a passenger jet with modern "lean-burn" engines, reported in Nature. Contrails from aircraft contribute to the climate-warming impacts of aviation. The findings demonstrate that more work is needed to understand and reduce the climate impact of jet engine emissions.

For anyone who has the 2026 2nd gen Apple Studio Display (not XDR), does it work with macOS Sequoia (15.7.4)?
Please only reply if you can directly confirm that it works (or doesn’t).
(The specs say it requires macOS 26 but who knows if that’s true.)

An eight-month study at a US tech company finds AI tools didn't reduce work but intensified it, as employees worked faster and took on a broader range of tasks (Harvard Business Review)
https://hbr.org/2026/02/ai-doesnt-reduce-work-it-intensifies-it

AI Doesn’t Reduce Work—It Intensifies It
One of the promises of AI is that it can reduce workloads so employees can focus more on higher-value and more engaging tasks. But according to new research, AI tools don’t reduce work, they consistently intensify it: In the study, employees worked at a faster pace, took on a broader scope of tasks, and extended work into more hours of the day, often without being asked to do so. That may sound like a win, but it’s not quite so simple. These changes can be unsustainable, leading to workload…

Cats & dogs aren't the only animals that nice to be pet.
Not saying you should go out and do it but look at this zebra shark revel!
▶️ A Zebra Shark Wont' Let This Diver Work
https://youtube.com/watch?v=OtwGVf2CeNM&si=qwfnsNy5z6vpG7op

People Hate Datacenters, Survey Finds https://www.404media.co/people-hate-datacenters-survey-finds/

1. This post is an incisive critique.
2. I’ve said before and I’ll say again that the only alternative to widespread popular violence is widespread empowerment through nonviolent mechanisms — not robbing people of power, that’s not an alternative, that’s a cause! — and if the oligarchs persist in their aggressive capture of power, Luigi Mangione is going to be not a curious aberration but a harbinger, and I don’t want that, I don’t actually want to live in a world where assassinations are a regular part of how things work, that’s miserable, that’s no way to live, and that’s exactly where we’re heading right now, and it’s grim and driving me to write egregious run-on sentences, please make it stop, thank you. https://wandering.shop/@aesthr/116381847228695045

I've just finished a modestly onerous administrative task.
It lead me to think about effort and the 3 legs of my job: research, teaching and admin: While there are intrinsic reasons to do research and teaching to the best of your ability, for admin "good enough" is enough. Focus on efficiency not excellence, where efficiency includes making things work, not causing extra hassle in the medium term.

People tell me that #AI code is fine, because you can run automatic tests. But tests can only tell you if code is doing the thing you want it to do.
To know what it SHOULD do, we used to have requirements. But now requirements are themselves vibe-coded slopotypes.
People were hoping that this would get them higher velocity, but testing the requirements in production only produces waste an…

Checking an LLM's work is a systemic, not an individual, problem.
Only a culture of design critique can prevent sloppy AI-generated problem definitions.

One thing I've noticed in recent years, not solely from government, is the assumption that processes are naturally parallelizable - ie they can be scaled with resources - whereas in reality this can take significant amounts of work to make true.
There's a lot of pre-cloud business processes - and licensing arrangements - that still work with the assumption of "one big computer".

The Government Told Courts It Could Easily Refund Unlawful Tariffs. Now It Says It Can’t.
When companies sued to block Trump’s IEEPA tariffs last year, one of the key arguments they made was obvious: if these tariffs turn out to be illegal, we’ll never get our money back. We…

I've been trying to use #LibreOffice as a replacement for #MSOffice but it's really not the same, lots of small things that do not work in the same way or that do not exist at all, whether in word/ writer or PowerPoint/ pages.
Is there another MSOffice-equivalent out there that is mor…

A well-executed video by B1M channel about the state of traffic infrastructure in Germany, how it should get better and why this still may not work out. With some impressive (for Germans probably well-known) examples. #traffic #infrastructure

Sonnet 005 - V
Those hours, that with gentle work did frame
The lovely gaze where every eye doth dwell,
Will play the tyrants to the very same
And that unfair which fairly doth excel;
For never-resting time leads summer on
To hideous winter, and confounds him there;
Sap checked with frost, and lusty leaves quite gone,
Beauty o'er-snowed and bareness every where:
Then were not summer's distillation left,
A liquid prisoner pen…

I'm with the auditors on this one! Companies deploying AI need more expert human auditing, not less, at least until fundamental flaws in 'AI' are addressed and it's reliably accurate
'KPMG gets money off its own audit by arguing AI makes accounting cheaper'

Reviewing applications for career funding is already hard, because you know the applicants’ situation all too well, and you’d like to reward their hard work.
But people really need to understand the goals of these programs! They’re intended to help you obtain a permanent position, not to simply continue what you’ve been doing, or to become even more specialized and to thus effectively reduce your chances of obtaining a permanent position.

Reviewing applications for career funding is already hard, because you know the applicants’ situation all too well, and you’d like to reward their hard work.
But people really need to understand the goals of these programs! They’re intended to help you obtain a permanent position, not to simply continue what you’ve been doing, or to become even more specialized and to thus effectively reduce your chances of obtaining a permanent position.

Related, do you know if anyone is currently working on such a project? I haven't followed @… closely for a while, but I'm not aware of anything beyond the blog and I'm not aware of other groups doing similar work.

In October, ice agents “smashed the car windows and detained all seven occupants” of a van that they thought suspicious because people in it were going to work together. Carpooling. Carpooling is foreign and unintelligible to the brilliant and diligent ice agent who testified as JB. The concept of carpooling is beyond JB’s comprehension. 🤦‍♂️
The intellect on JB could be studied (but might make anyone studying it dumber).

"The Rise of the Voluntariat" by Geoff Shullenberger (found via D. Graeber's "Bullshit Jobs").
"""
The voluntariat performs skilled work that might still command a wage without compensation, allegedly for the sake of the public good, regardless of the fact that it also contributes directly and unambiguously to the profitability of a corporation. Like the proletariat, then, the voluntariat permits the extraction of surplus value through its labor.
But unlike the proletariat’s labor, the voluntariat’s has become untethered from wages. The voluntariat’s labor is every bit as alienable as the proletariat’s — Coursera’s Translator Contract leaves no doubt about that — but it must be experienced by the voluntariat as a spontaneous, non-alienated gift.
And the voluntariat is not, like the proletariat, the instrument of its own dispossession. Rather, its contribution of uncompensated work accelerates deskilling and undermines the livelihood of those who do not have the luxury of working for free — in this case, professional translators who cannot afford to give away their labor.
"""
#AntiCapitalism

The Virginia State Bar rejects a disciplinary complaint from a press freedom group against a DOJ lawyer over the raid of WaPo reporter Hannah Natanson's home (Charlie Savage/New York Times)
https://www.nytimes.com/2026/02/12/us/politics/virgini…

Bar Punts on Ethics Complaint Over Application to Search Reporter’s Home
A press freedom group accused a prosecutor of violating an ethics rule by not telling a judge about a law limiting searches for journalistic work product.

The GF and I share our locations with each other via Google maps. Because a) we're grossly cute like that b) its easier than asking "where the fuck ARE you?" all the time.
I didn't recall the GF going to work, but she could have slipped out and me not notice so...
Instead of walking 10' to look out the window to see if the car is still here, i just checked Google maps to see where her avatar was. Yep, still home.
Im such a

As usual the cats are not impressed with my work...
#cats #catsOfMastodon #FediCats

Finally got my messaging apps transferred over to the GrapheneOS device!
For Signal, I had to restore a backup file: device-to-device sync failed.
For WhatsApp, I had to use a 5GHz 802.11ac network for the device-to-device transfer; 2.4GHz 802.11ax and 802.11n did not work, resulting in the very helpful error message "something went wrong." Due to GrapheneOS's sandboxing, restoring data from Google Drive didn't work, either, even with ALL permissions granted to …

“How may the compulsive programmer be distinguished from a merely dedicated, hard-working professional programmer? First, by the fact that the ordinary professional programmer addresses himself to the problem to be solved, whereas the compulsive programmer sees the problem mainly as an opportunity to interact with the computer. The ordinary computer programmer will usually discuss both his substantive and his technical programming problem with others. He will generally do lengthy preparatory work, such as writing and flow diagramming, before beginning work with the computer itself. His sessions with the computer may be comparatively short. He may even let others do the actual console work. He develops his program slowly and systematically. When something doesn't work, he may spend considerable time away from the computer, framing careful hypotheses to account for the malfunction and designing crucial experiments to test them. Again, he may leave the actual running of the computer to others. He is able, while waiting for results from the computer, to attend to other aspects of his work, such as documenting what he has already done. When he has finally composed the program he set out to produce, he is able to complete a sensible description of it and to turn his attention to other things. The professional regards programming as a means toward an end, not as an end in itself. His satisfaction comes from having solved a substantive problem, not from having bent a computer to his will.”
—Joseph Weizenbaum, Computer Power and Human Reason, 1976

And another nice shader speedup. Not quite the massive boost of the vector frequency/phase filters, but it's next in line after them.
I'm gonna use this random-access LFSR trick on so many filters that work with scrambled data lol.
v0.1.1 (20M point memory depth)
* PRBS31 generation: 65 ms
* PRBS31 verification: 360 ms
Latest (20M point depth)
* PRBS31 generation: 3 ms (21.6x)
* PRBS31 verification: 1.7 ms (211x)

Watched a couple of episodes of #DeArktiskeReddere on DR (public broadcaster on Denmark) yesterday (series about #Greenland's search and rescue team - though they do a lot of emergency air ambulance type work too).
It's a really gripping and well filmed piece of work, sensitive on the difficult topics too and shows really well how different authorities work together.
I imagine it would be uncomfortable watching for USians who seem to imagine some kind of stone age society seeing healthcare emergencies dealt with so professionally with such excellent equipment (and for free at point of delivery).
I also find it hard to explain that the categories and separation between "Danes" and "Greenlanders" isn't always so clear cut and I think the programme got that over well.
It's filmed mostly in Danish but with Swedish, Norwegian and quite a bit of English, so may not travel easily but it's a really well done piece of TV and if you are at all interested in Greenland, it's worth seeking out.
De arktiske reddere https://www.dr.dk/drtv/serie/de-arktiske-reddere_450849

Mailbag: Players' involvement in contract talks? https://www.dallascowboys.com/news/mailbag-players-involvement-in-contract-talks

"Study finds biodiversity credits could boost rewilding, but fall far short"
#Environment #Biodiversity #Nature

Study finds biodiversity credits could boost rewilding, but fall far short
Payments that enable landowners to rewild ecologically degraded land—in the form of biodiversity credits bought by investors wishing to offset their impact on nature—could be an effective component of the emerging market for nature recovery, but will not work as a standalone approach.

Thanks to a work-in-progress Homebrew cask by Andi Péter (https://codeberg.org/GramEditor/gram/issues/15#issuecomment-11159294), I was just able to quickly install and play with the new Gram code editor (

An Irish man who has been held by US Immigration and Customs Enforcement for five months
despite having a valid work permit and no criminal record
says he fears for his life and has appealed for help from Ireland’s government.
Seamus Culleton said conditions at his detention centre in Texas were akin to “torture”
and that the atmosphere was volatile.
“I’m not in fear of the other inmates. I’m afraid of the staff.
They’re capable of anything.”
Speakin…

"The hardest part of acting is not being guaranteed work. Every job could be your last."
—Henry Cavill
#acting #coaching #inspiration

Apple reportedly signed a 3rd-party driver, by Tiny Corp, for AMD or Nvidia eGPUs for Apple Silicon Macs; it's meant for AI research, not accelerating graphics (AppleInsider)
https://appleinsider.com/articles/26/0

AMD or Nvidia eGPUs can work on Apple Silicon Macs, but not for graphic acceleration
Apple has signed a driver for AMD or Nvidia eGPUs connected to Apple Silicon but there are some big caveats, and it won't improve your graphics. Here's what they're for.

Milwaukee & Madison DSA have endorsed Hong for Governor of Wisconsin.
“Working people have seen that the system doesn’t work for them. Time and again, the establishment has failed us so as not to upset their billionaire donors. As ICE threatens to terrorize our communities and kidnap our neighbors, Francesca Hong stands committed to fight back as the only candidate calling for their abolition.”

“You should be ashamed of where you work. Not just Grammarly or Superhuman or whatever comically dumb name you come up with next. Almost everyone running tech firms, most people in positions of responsibility, pretty much every C-suite type — congrats, you’re all making the world a worse place. People used to be excited about tech, now they dread what data you're going to steal next, they dread what violation of privacy or the environment will turn up next.”
https://www.moryan.com/an-open-letter-to-grammarly-and-other-plagiarists-thieves-and-slop-merchants/

A favorite dental assistant who long worked on my teeth once mentioned that she’d also had Pete Docter in her care at an orthodontia clinic when he was young, and she still vividly remembered his teeth — not in a good way or a bad way, just in a “she’s a professional and every set of teeth is unique” kind of way.
I thus feel a bizarre kinship with him, and cringe extra hard every time he puts his foot in his mouth like this, thus presumably undermining not only his reputation but also his orthodontic work.

"If you always put limits on everything you do, physical or anything else. It will spread into your work and into your life. There are no limits. There are only plateaus, and you must not stay there, you must go beyond them."
—Bruce Lee
#acting #coaching

Sunny Naqvi, a 28-year-old U.S. citizen, is now back at home
after her family says she spent about 43 hours in Department of Homeland Security custody.
Customs and Border Patrol officials are now disputing her relatives' claims.
Naqvi was born in Evanston and raised in the Chicago suburbs.
A few weeks ago, she was set to travel overseas on a work trip with five other people. That group included three U.S. citizens and three green card holders, all in the U.S. legal…

DHS detains US citizen from Evanston at O'Hare, releases her in Wisconsin after nearly 2 days
Family said agents told them Sunny Naqvi was not in custody despite her phone location showing her at the Broadview facility.

Imagine you had the blueprints for all car motors ever made. Some are joke blueprints or bad photocopies or missing sections. Some use metric units, some use imperial, and often it’s not specified. There’s any number of cylinders. Various cooling approaches. Different basic materials.
Now you make a new motor by taking the average of all these blueprints.
That’s how LLMs work.

There's a whole dystopian sci-fi plot prompt hidden here about a world that mirrors company stores, where everyone is paid partially in LLM tokens which they have to use for everything including work... Work, where they basically gamble the tokens away, always ending up in debt, needing to take out loans to buy tokens to then work for more tokens.
The story would be called "Sixteen Tons." Maybe I'll write it, but not unless it can be written as solarpunk instead of cyberpunk because the new rule is "don't write 'The Torment Nexus' because someone will think it's a good idea and implement it."

"The Holy See affirms its unwavering conviction that efforts to control, limit, reduce, and eventually eliminate nuclear weapons are not an unrealistic prospect, but a possibility and an urgent moral imperative." - Archbishop Gabriele Caccia, the Holy See’s Permanent Observer to UN, Oct. 21, 2025.

Holy See: Eliminating nuclear weapons is an urgent moral imperative - Vatican News
Archbishop Gabriele Caccia, the Holy See’s Permanent Observer to the United Nations, urges the international community to work towards preventing the ...

I’ve just had occasion to use Svelte / SvelteKit on a web project, and…
…it’s quite good. The core features are well chosen, and they work. The learning curve pays dividends. The resulting code is reasonably pleasant to read. As often as not, the surprises have sensible decisions behind them. It feels like it •is• the thing that React is •trying to be•.
There are gaps and quirks and barriers — it is a tool, after all — but if you’re writing a highly interactive SPA-style site, it gets the Paul Seal of Approval.

"The hardest part of acting is not being guaranteed work. Every job could be your last."
—Henry Cavill
#acting #coaching #inspiration

A story of courage and tenacity.
#strength

RMS Carpathia's Courageous Rescue Efforts
“You are not obligated to complete the work, but neither are you free to desist from it" - Fathers Ethics (2:21) I was struck by this story. It is not quite about the Titanic, but the ship and people that fought to rescue her: The Carpathia. "Carpathia received Titanic’