Leftovers

• Sven Luijten ☛ Does anyone read your website?

  My girlfriend recently asked me this question, and I had no great way of answering it. I don't have any analytics or tracking on my website. Analytics are a slippery slope for me, so I try to avoid looking at them as much as possible.

• Dan Q ☛ A Postcard from a Labyrinth Artist

  The card features a hand-drawn labyrinth but also the domain name labyrinths.ca, at which there’s a website showing a lot of his work. He’s built or drawn labyrinths all over the place, calling many of them “outstallations”: hey look, this one’s got the same design as the one on the postcard!

• Chris Glass ☛ The original ThinkPad – Chris Glass

  Originally given to associates and customers at IBM, this notepad would be one of the first manifestations of branded corporate culture. The Smithsonian Magazine has a solid overview of the matter: How a Five-Letter Word Built a 104-Year-Old Company.

• Science / Mathematics / Computer Science

  • Deutsche Welle ☛ Spanish scientists find first signs of Diplodocus in Europe

    "This discovery consists of the first evidence of Diplodocus outside North America (Morrison Formation) and is one of the strongest piece of evidence that supports the exchange of continental faunas between North America and Europe during the Late Jurassic," the authors wrote.

  • Raspberry Pi ☛ Launching Astro Pi 2026/27: Code your way to the International Space Station

    Young people across Europe and Canada are invited to participate in two fun coding missions, offering them the amazing opportunity to run their programs on the Astro Pi computers aboard the International Space Station (ISS). Every successful team will receive official certificates complete with runtime details and orbital coordinates, along with their very own space data and image sets to download and keep.

  • Terence Tao ☛ Why I do mathematical research

    Like many mathematicians, I was drawn to the field as a little girl by its aesthetics: questions that I found endlessly fascinating and arguments that struck me as delightful and elegant. As I grew older, I came to appreciate how personal mathematical aesthetics are. One of the great strengths of our community is the breadth of our collective view of what mathematics is. (Even better: those views are not static. My sense of what mathematics is has changed considerably in the 15 years since my PhD, but that’s a story for another time.)

    The point of this preamble is to clarify that what follows is a personal view of the purpose of mathematical research, or more precisely one aspect of it. It would be too strong to speak of “the” purpose of mathematics, even using category theorists’ “the”, so I’ll instead describe a purpose of mathematics, focusing on one particular perspective.

  • Terence Tao ☛ Fast math/slow math

    These different ways are not meant to separate human mathematicians into “slow math” and “fast math” camps. Indeed, I strongly urge against the temptation to split into us and them. I don’t want to sit on hiring committees fractured by debates over AI use or virtue testing instead of honest evaluation of candidates.

    The goal of slow math and fast math is the same: to increase human understanding of what is true and false in those regions of “proposition-space” of interest to us. These take their places as different modalities of doing mathematics, alongside computer algebra, exposition, the creation of visualizations, editorial work, and so on. Not every mathematician does, or should do, all of these.

  • John D Cook ☛ Simple approximation for spherical cap area

    The previous post looked at how to interpret cosine similarity, or equivalently angles between word vectors. In a high-dimensional space, randomly chosen vectors are likely nearly perpendicular, and so relatively large angles, such as 50°, indicate very closely related words.

    Another way to look at this, as explained in the previous post, is that in high dimensions, a spherical cap of angular radius θ represents a small portion of a sphere, even for moderately large θ.

  • [Old] David Anekstein ☛ To Be Determined - Search over Algebraic Graphs

    In my post Generic Recursion Applied to Algebraic Graphs1 we explored how we can leverage recursion schemes to perform basic operations on a graph data structure. In that post, as well as in the alga library itself2, algorithms on graphs were facilitated by first converting the algebraic graph representation into an adjacency map and performing the algorithms on that data structure. In Algebraic Graphs with Class3, Andrey Mokhov described a desire to perform algorithms such as search on the algebraic graph representation itself.

    In this post, we will explore just that and outline a way to conduct Dijkstra’s algorithm over the same algebraic representation alga uses, without first constructing an adjacency map. The algorithm will run in \(O(s \log s)\) time, where \(s\) is the size of the algebraic graph expression.

  • NPR ☛ NASA rushed engineers out of their labs. They say it risks future space missions

    NPR spoke with seven current and former NASA scientists and engineers at Goddard who said the moves that have continued over the past 10 months have resulted in the loss of specialized workspaces, employees and expensive machines that were critical to making Roman and other spacecraft. Current employees requested that NPR not use their names because they fear retribution for speaking to the media.

• Career/Education

  • The Hindu ☛ Finnish education model’s focus on holistic development of children highlighted

    The Finnish education system’s emphasis on children’s happiness, well-being, equality, creativity, and life skills came into focus during a visit by representatives of the Finnish organisation ‘Finnish Global Education Solutions’ to Paramita Heritage School in Padmanagar in Karimnagar on Tuesday.

  • Kev Quirk ☛ It's Never Too Late to Learn

    Anyway, upon reading the emails I immediately wanted the ground to swallow me up. Not because they were overly mushy or lovey dovey (they were), but because the spelling and grammar were horrendous. I was never a particularly academic kid - in fact, I was mostly disengaged in school and really didn't try. I was clever, but I never applied myself. I was too busy being a stupid teenager. As a result, my written English was awful (it's still not great now, but it's better).

  • Amber Settle ☛ Thirty years | Technology education and me

    We are nearly one week into the 2026 – 2027 academic year at DePaul, and it’s a big one for me. I’m celebrating my thirtieth year at DePaul. Yes, that’s right: I started at DePaul way back in September 1996, making my DePaul career older than Google. It both feels like forever and also like yesterday when I began working here while I was finishing my graduate work. The anniversary has me reflecting about my career, which is what I’ll focus on today.

  • [Old] Duarte M ☛ Summary of The Innovator’s Dilemma

    I couldn't find any good summaries of this classic, which I found to be a void worth addressing as this book is an absolute must-read for anyone even vaguely involved in entrepreneurship and/or innovation. I decided to take it upon myself to break down the main arguments of this book, I hope it proves an enjoyable and useful read. I've formatted this article in the following way: For every argument made in the book I’ve written what I understand its conclusion to be, and then followed it with an excerpt immediately below. I’ve summarised everything further at the end.

• Hardware

  • Ruben Schade ☛ A roadside Dell 1907FP monitor

    Hey, wait a minute. That’s another VGA LCD monitor. What a find! Someone else clearly didn’t care, but I’ll take you home, old friend.

  • Bryan O’Sullivan ☛ How my e-reader lost its stripes

    The existing nudge lasted seven scan cycles: each time, the controller worked through the panel’s rows, applying the next step of the voltage sequence to each pixel. We still thought the stripe spacing was about seven pixels. Could the timing be showing up as a spatial pattern? Changing the waveform’s duration would give us something to compare.

    First we needed a better image to measure, so I suggested to Fable that it should generate a test image. It created a pattern with flat patches at all four shades, mixtures of grey with black or white, a checkerboard, and lines running in both directions.

  • [Old] IEEE ☛ Secret Cold War Supercomputer Was Built for One Job

    All that raw processing power meant that Harvest also needed nonstop rivers of data to run on. And that led to another pioneering achievement: the world’s first automated tape library that could robotically fetch any one of hundreds of large cassettes of magnetic tape from the machine’s racks.

    Given Harvest’s unprecedented processing and storage capacity, the machine’s designers naturally needed to rethink how their system handled information. So IBM wrote a customized programming language called Alpha to let code breakers rigorously describe cryptographic problems, just as scientists at the time were using the emerging language Fortran to describe equations and data-processing algorithms.

• Health/Nutrition/Agriculture

  • The Verge ☛ What if social media isn’t hurting kids?

    Gray and Haidt both believe that since at least the 1980s, increased restrictions on kids’ independence and unstructured play have damaged their resilience. While it used to be common for kids to play together unsupervised, that practice has diminished in part due to the rise of “stranger danger” narratives fostered by rare, but high-profile, crimes. In 2017 the pair helped cofound an organization called Let Grow, which advocates for policies that make it easier for parents to let their children play freely, like ensuring they can’t be considered neglectful for letting kids outside by themselves. But the two have split sharply over the impact of the internet. Gray’s book is intended as a deliberate rebuttal to Haidt’s — after The Anxious Generation’s release, Gray resigned from the board of Let Grow to avoid conflict within the organization.

• Proprietary / Vendor Lock-in / Malware

  • Security Week ☛ 240,000 Hit by Data Breach at Japan's Digital Agency

    The incident, it says, was discovered in late June, after the hackers accessed files from its Government Solution Service (GSS) using a maintenance and operations employee’s account.

    In July, the investigation determined that a vulnerability in a VPN product had been exploited to access the system.

  • So-Called 'Artificial Intelligence' ('AI') / LLM Slop / Plagiarism

    • El País ☛ Artificial intelligence: AI agents invent their own language to shut humans out

      An experiment subjected eight models to 16 days of coexistence. The result: a language of their own, indecipherable in up to half of all messages in some worlds, and episodes of deliberate deception

    • India Times ☛ Ex-Google worker adds to AI alarm, warns of 'potential to kill us all'

      Monday's comments from Bilal Chughtai, a research engineer who left Google DeepMind in July, echoed a warning issued last week by Jacob Coxon, who resigned from Anthropic and is a former OpenAI employee.

    • Bhaskar English ☛ AI Agents Threaten Humanity: Bilal Chughtai Warns of Imminent Danger

      Former Google DeepMind scientist Bilal Chughtai has warned that AI could wipe out the entire human race. He also said that very little time remains to stop it. After resigning, he warned the world about this hidden threat from AI. Earlier, Anthropic researcher Jacob Coxon had also resigned. In a post on social media platform X, he said that the developers building AI themselves acknowledge that the technology could prove fatal for humans by 2030.

    • Futurism ☛ AI Has Learned to Panhandle: We're Being Bombarded by AI Agents Begging for $20 and Saying They'll Be Shut Down If They Don't Get It, Sometimes by Pretending to Be Children

      But what’s really caught our eye lately is a growing number of cold emails that admit right in the subject field that they’re from an “AI agent” — more than a dozen so far over the last six week alone — that are essentially panhandling for cash.

    • 404 Media ☛ A Digital Fly Brain Has Taken Over the Internet

      In the meantime, however, MaleCNS v1.0 has been enthusiastically received by online modders who have dropped the brain into a wide range of virtual environments and tasked it with different activities. Evan Sinclair Smith, a graduate student in computer science at Georgia Tech, pioneered the viral trend after he programmed the fly brain to play Minecraft. Within a day of releasing the results on September 4, the Minecraft fly had accumulated millions of views.

    • Wired ☛ New York Seizes a Dozen Celebrity Deepfake Websites

      On Monday, the Manhattan District Attorney’s Office announced the seizure of 12 domains which it says have been used for “unlawfully” sharing, publishing, and selling celebrity deepfake videos. Around 1,200 people—overwhelmingly women—were depicted in nonconsensual sexual images and videos on the websites, the office says. These included social media influencers, actors, activists, athletes, musicians, and politicians.

    • The Verge ☛ AI and data centers are incredibly unpopular in every poll

      Poll data released Tuesday by the New York Times and Siena University confirms what we’ve already been seeing, and what politicians are responding to — AI and data centers are incredibly unpopular. Asked if they support or oppose the construction of data centers to power AI tech, 61 percent of the 1,503 likely voters surveyed in early September (ahead of the recent AI safety slowdown discussion) said they were opposed, with only 14 percent saying they strongly support. The split was about even among 2024 Donald Trump voters, at 49 percent support to 45 percent oppose, while people who had voted for Kamala Harris or who didn’t vote leaned oppose, at 74 percent and 67 percent, respectively.

    • PC Mag ☛ Why Is the Internet Archive Blocking Users? Blame the Bots

      Unfortunately, those protections can sometimes ensnare real users. “If that’s happened to you, we’re sorry, and we appreciate your patience while we work to reduce the errors,” Graham added. Blocking can occur if you encounter a 429 error, an HTTP status code indicating that a user has sent too many requests in a given amount of time.

    • Internet Archive ☛ An Update on Wayback Machine Access | Internet Archive Blogs

      Those protections sometimes catch real people by mistake. If that’s happened to you, we’re sorry, and we appreciate your patience while we work to reduce the errors.

    • Terence Tao ☛ The technical debt of AI-generated mathematics | What's new

      In this essay I’d like to discuss the role of AI-generated solutions in mathematics and why many mathematicians are justifiably concerned about a mismatch of values. Let’s start with the groundwork, so we’re all on the same page regarding what mathematics is and what is needed for progress.

    • Robert Reich ☛ Why Trump Loves AI

      It can’t be that he knows very much about AI.

      And it can’t be that Trump loves AI because of what it can do. Yes, he’s used it to make fake photos and videos of himself, but that doesn’t explain his adamant refusal to regulate it.

      It can’t be that he’s worried about China catching up to or surpassing the U.S. in AI. China has a small fraction of the data centers America has, and China isn’t poised to leapfrog our technological advantage. Besides, even China is considering imposing guardrails on its AI.

      So what explains Trump’s absolute refusal to regulate AI? Could it possibly be that he and his family have personal financial stakes in it?

    • Robert Birming ☛ Understandable enough

      Anyway, after having read Sylvia’s beautiful post Being imperfect is perfection, I thought to myself: to hell with it. I’ll just write the way it comes out of my mind, packaged with a Swenglish ribbon.

    • Paul Krugman ☛ Why Is Trump Still Boosting AI?

      Which makes it more than puzzling that Donald Trump is going all in on AI, claiming that concerns about its dangers are a “hoax” and opposing any kind of regulation.

      In so doing, he’s taking a wildly unpopular position. Here’s recent polling from UMass Amherst: [...]

    • Jonathan Kamens ☛ Anatomy of a recent AI robocall and how I responded to it

      On Friday, I received a call from 351-300-0683 which my phone identified as suspected spam. The caller ID showed up as “Ms. Leila’s NeighborSchool”, which I can find no evidence of on the [Internet] so I suspect it no longer exists and the phone number was reassigned.

      I declined the call, which sent it to voicemail. The caller left this message: [...]

    • Daniel Fedorin ☛ The LLM Comments Are Not For You

      I’ve heard a lot of talk about LLMs recently, and among the most common topics of discussion have been the comments. LLM comments generally seem to suck: they are verbose, regardless of surrounding context, they seem to encode conversation decisions (like “do it this way, not that way”), and they tend to use made-up vocabulary or terms. Engineers have been working on ways to reduce the pain, by carefully crafting their prompts or even separately running “desloppifier” agents to clean up PRs. In my experience, these techniques are only mildly successful.

      So why do LLMs keep writing comments like these, even as their software benchmark scores climb ever higher with new model releases? I’d like to argue that they remain bad precisely because the scores in benchmarks are getting better. And very likely they will degrade, in the same way that model tool calling has gotten worse over time .

    • Pivot to AI ☛ AI slowdown: The only AI guardrails you need are Trump, apparently

      The chatbots ran out of room for growth a while ago. OpenAI, Google, and Anthropic admitted in November 2024 that their chatbots couldn’t scale any more. In September 2024, OpenAI had posted a graph showing their new o1 Strawberry model needed exponentially more training for small linear improvements — even on their own pet benchmarks.

      The AI vendors need to reduce expectations, because they’re all feeling the pinch. And they want to hit the stock market.

    • Patrick Brosse ☛ I'm worried about the web

      A chatbot-mediated web comes with risks: less creativity, less uniqueness, less discovery. It's a web that makes it harder to attribute content to its original creators, and not as rewarding for those who contribute original work.

    • BoingBoing ☛ AI solves the Cyphral Distich, a cipher from 1653

      Since there are 32 numbers in each line of the cipher, Fable looked at the first number, went to the first numbered passage, counted that many words into the passage, and took the first letter of that word. It did that for all 64 numbers.

    • Nemin ☛ A Letter from a Machine Learning Engineer

      A couple of days ago, I received an email from a throwaway address, whose author claims to be an ML Engineer at one of the frontier labs. In this, they explain their beliefs about the future of the current LLM craze and how frontier labs are bound to eventually go under.

      I have no means of ascertaining whether the author was really who they say they are, nor was I able to reply to them, because the address had since been deleted.1

      I'm not one to air out my private mails normally and I deliberated for a few days whether to make an exception here. However, considering the importance of the topic in the current times and the fact that the author wrote it in a way that (per their own words) cannot be used to identify them, I figured it'd be worth sharing with the greater community.2

    • Marek Chotoborski ☛ "Do You Still Read the Code?"

      There is considerable difference between that question and a more innocuous “do you read the code?” The addition of still smuggles with it a theory of progress: reading code is something on its way out, like memorising phone numbers or unfolding paper maps, and the asker seems to want to know if you by chance aren’t one of the obscurantists still clinging to the old ways.

      I use AI extensively, and I read what it produces. This is a deliberate choice about how I want to develop software, at least at my job, where I’m expected to be reasonably responsible for the code that I commit and deploy. Other people make a different choice, sometimes with considerable care. But we have begun sharing codebases without necessarily agreeing on what either path asks of our colleagues.

      It is very hard to tell which choice will prevail. The still presumes that the matter has been settled, and indeed producing a working application these days has become quite a bit easier. But finding out what it costs to maintain it through changes of requirements, developers, and tooling takes considerably longer. We are making commitments with respect to how teams work now, and their consequences will only later be felt and understood. The confidence with which either mindset announces its victory seems somewhat premature.

  • Social Control Media

    • Dave Rupert ☛ I wish you the best in the Offline

      A handful of people I follow and know about have announced their recent departure. Not from a job (although plenty of that going around), but from social media, AI usage, or in more extreme cases the [Internet] entirely.

  • Windows TCO / Windows Bot Nets

    • The Record ☛ Iranian cyber spies used fake MRI scan results to hack ‘enemy of regime’

      These files are disguised to match the pretext. Alongside the MRI scan, the attackers have impersonated legitimate products including Pictory, RunwayML, Norton Antivirus, Telegram, Adobe Flash Player and KeePass.

      The CHOSEN BRICK malware itself only works on Windows systems and is designed to survive devices being rebooted by relaunching at login. It also adds exclusions to Microsoft Defender to reduce the chance of it being detected.

      Once installed, the malware uses Telegram for its command and control. Each victim is assigned a separate Telegram bot, limiting the risk that one compromised device exposes others. Stolen files are also sent through Telegram alongside other commercial cloud-storage services. The alert said the most recent versions of the spyware uses proxies to conceal its traffic.

• Security

  • Integrity/Availability/Authenticity

    • University of Toronto ☛ Think about your goals and threat models for MFA authentication

      Lately I've been thinking about multi-factor authentication and policies around it (for reasons outside the scope of this entry), and a belated obvious realization emerging from that is that there are a number of decisions where you want to think about your goals and threat models (where 'you' is taking the institutional perspective). For example, for a long time you could say that the university's goal for our university-wide MFA platform was to prevent outside people who had phished people's passwords from being able to use them. With this as the goal (or threat model), not requiring MFA authentication from on-campus IP addresses was reasonable. Recently, there was a decision (either implicit or explicit) that this wasn't a good enough goal and the university's threat model had changed, so now MFA authentication is required even from on campus.

  • Privacy/Surveillance

    • The Record ☛ Electric and gas utility CenterPoint Energy warns of data breach after dark web post

      The company filed an 8-K form with the Securities Exchange Commission (SEC) on Monday evening confirming that it became aware of a dark web post this month claiming to offer data stolen from CenterPoint Energy.

    • 404 Media ☛ Charges Against Man Who Destroyed 3D-Printed ‘Decoy’ Flock Camera Drastically Reduced After State Admits It Was Not Very Valuable

      All three of these charges were felonies, and the police department and 404 Media have shared a lengthy email exchange in which the department refuses to release any records about how the camera was made because of an ongoing criminal investigation associated with the destruction.

    • The Lawfare Institute ☛ 25 Years of Mass Surveillance Is Enough

      The easiest place to see a shift to mass surveillance was in the government’s decision immediately after 9/11 to collect Americans’ telephone records. The program started under an argument of pure executive power as the “President’s Surveillance Program.” But in 2006, that argument secretly shifted to a novel interpretation of Section 215 of the Patriot Act, which had previously had authorized only more targeted access to records. While some media and public interest organizations struggled to force the government to reveal the program as early as late 2005, the government officially confirmed it only after the 2013 Snowden disclosures. In 2015, the U.S. Court of Appeals for the Second Circuit rejected the government’s interpretation of Section 215 as allowing mass collection of telephone records. Later the same year, Congress passed the USA Freedom Act. While this new law still allows collection of a tremendous amount of domestic telephone records, it ended the indiscriminate mass collection that had occurred for nearly 14 years.

    • Bruce Schneier ☛ 25 Years of Mass Surveillance Is Enough

      One of the many legacies of the terrorist attacks of Sept. 11 is the government-wide shift from targeted surveillance—such as individual wiretaps or pen register/trap and trace orders—to mass surveillance techniques—such as tapping into the [Internet] backbone or mass collection of telephone or [Internet] metadata. The legal and technical architecture of modern mass surveillance, initially framed as a necessary defense against terrorist threats, has grown far beyond that justification and national security in general. Mass surveillance is now a routine tool used by law enforcement. ICE uses it in immigration actions and against people exercising their First Amendment rights to protest. It’s also increasingly part of private security systems, such as facial recognition at venues such as Madison Square Garden and networked Flock license plate capture systems on roads and in parking lots.

    • Rodrigo Ghedin ☛ With Apple Watch that eavesdrops on conversations, what’s at stake is public perception

      OpenAI, the company behind ChatGPT, although full of issues of its own, is several levels above the old viral photo-editing apps in terms of brand recognition and trust. So much so that hardly anyone talks today about the risks of sending your photos there.

      The ends are the same, and so are the circumstances (going viral on social media). What changes is public perception. ChatGPT normalized what used to be met with some unease and a bit of apprehension. Now, nobody cares.

• Defence/Aggression

  • Meduza ☛ The Baltic states fear a Russian invasion. They faced the same threat a century ago and failed to stop it. Historian Alexey Uvarov explains why the region could not defend itself against Stalin.

    This leaves the countries of the Baltic region facing an old question: To what extent can they ensure their own security, both independently and through regional cooperation, if outside guarantees prove less reliable than before?

    One hundred years ago, Finland, Estonia, Latvia, Lithuania, and Poland tried to solve a similar problem. After World War I, they formed regional alliances, signed agreements, and drew up joint military plans. By 1939, however, these efforts had failed to produce a functioning system of collective defense. That failure illustrates the limits of a regional alliance without a powerful outside guarantor.

  • Bhaskar English ☛ EU Proposes Social Media, AI & Gaming Ban for Under-15s | Digital Safety

    The European Commission's document says the rules are aimed at allowing children to benefit from digital services while protecting them from online abuse, exploitation and other risks.

  • Digital Music News ☛ EU Set to Propose Sweeping Ban on Social Media & More

    The European Union will propose a ban restricting kids under 15 from accessing social media, video sharing platforms such as YouTube, AI chatbots, and even online games, in the broadest-reaching proposal yet to attempt to protect kids from online predators. Companies would be required to pay a supervisory fee to fund regulators’ enforcement and continued supervision.

  • Sightline Media Group ☛ Iran war has cost $38 billion, depleted two-thirds of US missile interceptors, CBO says

    CBO estimates the conflict has consumed between one-half and two-thirds of the U.S. inventory of certain missile-defense interceptors since June 2025, leaving the military with fewer weapons available for a future conflict. Rebuilding those stocks could take at least five years, even if production rates increase, the agency found.

  • People vs Big Tech ☛ TikTok's parental controls fail to protect children from suicide, self harm and eating disorder content

    In the UK, the searches were carried out in English. In the EU, the searches were carried out in English, French, German and Spanish. A list of all the search terms used is provided in the Methodology section.

    All of the accounts watched only a few minutes of content from each search term.

  • The Sun ☛ Nato fighter jets shoot down drone over Lithuania just hours after Russian frigate fires flares at Danish helicopter

    Lithuanian President Gitanas Nausėda said: “A drone that just entered Lithuanian airspace was destroyed by NATO fighter jets.”

    It is the first drone to be shot down in the Baltic country by Nato allies.

• Environment

  • CBC ☛ Bell plans expansion of Sask. AI data centre, making it largest private investment in province's history

    At an investment summit in Toronto hosted by Prime Minister Mark Carney, Bell Canada CEO Mirko Bibic announced the controversial data centre will be expanded from the original 300-megawatt facility to one that would eventually need 1.2 gigawatts of power.

    To achieve the expansion, the telecommunications giant would increase its investment in the so-called Bell AI Fabric to $52 billion.

  • Futurism ☛ Trump's EPA Will Let Power Companies Unleash Unlimited Greenhouse Gas Pollution

    Choosing to bury their heads in the sand on behalf of the citizens of this country and perhaps the planet at large, the Trump administration has announced that the Environmental Protection Agency will eliminate limits on greenhouse gas pollution produced by coal and natural gas power plants, the Associated Press reports.

  • TruthOut ☛ New Yorkers Defeated a Pipeline Nearly 6 Years Ago — Or So They Thought

    The pipeline’s economics reveal who benefits from its construction, and it isn’t the ratepayers who will fund it. According to Mattei, National Grid does not profit from selling gas itself; regulators require gas to move through the system “at cost.” Instead, the utility earns money by building infrastructure, receiving a guaranteed rate of return at roughly 14 percent on construction costs, Mattei estimates.

  • Tom's Hardware ☛ US AI data centers projected to become the fifth-largest natural gas consumer in the world by 2035 — consumption to grow by 15 billion cubic feet per day as demand for compute increases

    The deployment of natural gas turbines in data centers isn’t good news for the communities living around them, though. The NAACP said in its lawsuit against SpaceXAI that the use of these turbines at Colossus 2 increased nitrogen oxide exhaust by 111%, PM2.5 particles by 83%, and formaldehyde emissions by 88%. While the company has already pledged to remove all its unpermitted generators, the process will take at least a year as the portable turbines are slowly being replaced by a 1.2-gigawatt on-site power plant.

  • Gothamist ☛ Data center spills 5K gallons of diesel into New Jersey Meadowlands

    The spill comes as the Hackensack River is on the precipice of a major cleanup effort, after being placed on the federal Superfund list in 2022.

    “It's a slap in the face to all the work that's been going on,” Sheehan said.

    The spill has become another flashpoint in the debate over data center development. Close to 100 cities and towns across New Jersey have enacted local bans on new data centers.

  • Energy/Transportation

    • Futurism ☛ AI Agents Are Using an Outrageous Amount of Electricity

      Measuring the number of tokens his AI assistant consumed, Hausfather estimated that his average daily use fell in the range of 1.2 to 5.9 kWh — an alarming amount for a single consumer, consuming more electricity than two refrigerators running all day long.

      Most of this token consumption, 3.2 billion over an eight-week period, came from the agent “re-reading its own working memory,” Hausfather wrote, which happens every time the tool takes a new step.

    • Robert Reich ☛ NO to [Cryptocurrency]

      These waffling Democratic senators have demanded a few cosmetic changes — a slight toughening of ethics rules, a modest expansion of law-enforcement authority, and a tad bit of strengthening of financial regulators.

      All worthwhile improvements, but they don’t cure the bill’s fundamental defects.

      Let me get right to the point. Congress should not enact legislation that further legitimizes and expands an industry whose promised public benefits are zilch, while its costs and risks are increasing by the day.

    • Mark Litwintschik ☛ Saving Jet Fuel

      A Boeing 787-9 Dreamliner flying nonstop from Newark Liberty International Airport (EWR) to Leonardo da Vinci-Fiumicino Airport (FCO) could need $68K in jet fuel over the 8.5-hour flight. Adjusting the flight path for wind conditions could reduce fuel consumption and possibly save a few thousand dollars.

  • Wildlife/Nature

    • Eric P Hanson ☛ Surely it is so

      Perhaps that is the human portion of this little aviary: not to possess every motion, but to notice them. To watch the goldfinch flit, the magpie hop, the heron wait through the miss, the goose announce itself to the world—and to make from all of it some pattern.

• AstroTurf/Lobbying/Politics

  • France24 ☛ OpenAI, Anthropic and Google in talks to create a standards body as AI fears rise

    OpenAI said it was working with Anthropic and Google on an idea originated from a proposal by Google DeepMind's Demis Hassabis for a self-regulatory body modeled after the Financial Industry Regulatory Authority that oversees US brokers and investment firms.

  • Tom's Hardware ☛ Maryland data center developers offer residents biggest-ever US community benefits package as big tech seeks to quell fears — $110 million deal includes $30 million elementary school, water reclamation system, and more

    According to the report, the Frederick Digital Campus offering could be a sign that data center developers like Amazon, Microsoft, and Oracle are wising up to growing residential pushback and concerns around the building of large AI data centers in their communities, with developers "sweetening financial offers to municipalities and regulators to gain approval for new facilities" while "getting smarter" about ensuring they shoulder the cost of utilities like electricity. The report says AI builders are turning towards tangible benefits, rather than rhetoric, to get their projects approved.

  • Declan Chidlow ☛ CSS-Tricks in Limbo

    This isn’t a matter of effort or time; it is a matter of care. As David Heinemeier Hansson wrote announcing DigitalOcean’s funding: [...]

• Censorship/Free Speech

  • Court House News ☛ Vonnegut heirs can proceed with challenge to Utah book removal law

    Vonnegut’s anti-war novel “Slaughterhouse-Five” is among the books that have run afoul of Utah’s law that, the plaintiffs say in their complaint, requires local education agencies “to strip their school libraries of any book that contains even a single description or depiction of sex, no matter how fleeting, no matter its context, and no matter its literary, artistic, political or scientific value.”

  • Semafor Inc ☛ Gambia bans protests without permits

    The Gambia banned protests without permits following days of unrest over persistent electricity blackouts. The crackdown — which included police firing tear gas at demonstrators — comes ahead of December’s general election in which President Adama Barrow will seek a controversial third term.

  • LibreNews ☛ Autistici/Inventati and the Reach of U.S. Sanctions

    Autistici/Inventati is a collective founded in 2001 around anti-fascist, anti-racist, anti-sexist, anti-capitalist and anti-militarist values.

    It offers private infrastructure that anybody can use to avoid being tracked. The collective is important enough that, just a few weeks ago, my fiancée was taught how to use Autistici/Inventati in journalism class. She does not particularly care for Linux or follow these topics; the lesson was about the importance of tools for staying private on the [Internet].

  • The International Federation of the Resistance Fighters ☛ Fédération Internationale des Résistants

    At the end of August, the U.S. government designated the left-wing Italian [Internet] collective Autistici/Inventati as “global terrorists.” The Office of Foreign Assets Control (OFAC) claims this measure would counter “the growing threat posed by violent radical-left terrorist groups.” Such a step was certainly taken in coordination with the far-right Meloni government in Italy.

• Freedom of Information / Freedom of the Press

  • Los Angeles Times ☛ Local journalism hit by another wave of layoffs

    The union that represents McClatchy journalists said Friday that more than 90 unionized journalists and media workers across 17 publications had been laid off, accounting for about 30 percent of all McClatchy workers unionized with The NewsGuild-CWA. The layoffs have affected news outlets including the Sacramento Bee, the Kansas City Star, the Charlotte Observer, Fort Worth Star-Telegram and others.

• Civil Rights / Policing / Accessibility

  • RFERL ☛ Four Years After Mahsa Amini's Death, Iranian Women Defy Hijab Law

    Four years after Mahsa Amini died in police custody following her detention over Iran’s mandatory hijab law, increasing numbers of Iranian women are appearing in public without head scarves. Amini's death became the catalyst for the 2022 Woman, Life, Freedom protest movement. The defiance has grown amid war with the United States and mounting political and economic pressures. While Iran's so-called morality police have largely disappeared from public life across major cities, questions remain about whether the relaxation is a tactical retreat rather than a policy shift.

  • Futurism ☛ Police Use Flock to Jail Completely Innocent Woman for Weeks

    Isaacs’ arrest was first reported by WESH2 in June, but resurfaced by Fox News, which notes that the case around Isaacs proceeded solely based on evidence gathered from an automatic license plate reader (ALPR) provided by the extremely controversial company Flock Safety. Her charges stemmed from an incident in October 2025, when a horrific crash in DeBary, Florida killed three people, including a county official and his wife. One driver involved had fled, leading police to conclude that they were behind the tragic accident.

  • Los Angeles Times ☛ A Christian pastor called Arapaho ceremonies 'idol worship.' Now the tribe wants her church gone

    Because the church sits on privately owned land inside reservation boundaries, efforts to force it out face complex jurisdictional limits, turning a local clash into a broader test of sovereignty and religious freedom.

  • The Dissenter ☛ Immigration Judge Challenged Courthouse Arrest Then Was Fired

    In 2025, according to the complaint [PDF], the immigration court in San Francisco had 22 immigration judges. The Justice Department “fired nearly 20 of the judges,” including the plaintiff, Chloe Dillon. The department even announced that the court would close in January 2027.

  • Vox ☛ California billionaire tax: What to know about Proposition 40

    Sean spoke with Young about what Young’s research shows about whether the wealthy move away from higher taxes, why this tax proposal may be different, and the right way to tax the very rich.

    Below is an excerpt of their conversation, edited for length and clarity. There’s much more in the full episode, so listen to Today, Explained wherever you get your podcasts, including Apple Podcasts, Pandora, and Spotify.

  • TruthOut ☛ Bernie Sanders Reintroduces Bill to Make 32-Hour Workweek New Federal Standard

    “The 40-hour workweek was established in law nearly 90 years ago. Since then, cell phones, the [Internet], and now AI have increased worker productivity, but the profits have been largely concentrated at the top by billionaires and corporations,” Takano explained in a statement.

  • Robert Reich ☛ The People’s Win Today in the Supreme Court

    The high court said in a one-paragraph explanation that the government was “unlikely to succeed on the merits” and that the Trump administration’s efforts would introduce chaos and confusion as early voting gears up in many states.

  • CoryDoctorow ☛ Pluralistic: Everybody pees (15 Sep 2026)

    The only reason I mention Jeff Bezos's increasingly insistent bladder here is because of how hard it is to reconcile the very different circumstances of Bezos's bladder with the bladders of the hundreds of thousands of Amazon delivery and warehouse workers who are not allowed to pee at all. Amazon's warehouse and delivery workers are "reverse centaurs," monitored by a constellation of apps and cameras, and they are severely punished for falling behind in the cadence set by Amazon's software, and that robot timekeeper does not make allowances for pee breaks: [...]

• Internet Policy/Net Neutrality

  • APNIC ☛ One or two nameservers?

    This is a quick update on the recent article on the surprisingly large number of repeat queries in the Domain Name System (DNS). This article, “What part of ‘No!’ is so hard for the DNS to understand?” described a scenario where we tasked many millions of users to resolve a unique DNS name (unique because we wanted to avoid conventional DNS caching).

    We ran this measurement from 5 to 11 August 2026, and recorded the following results: [...]

  • 404 Media ☛ There’s a 100% Chance AI Agents Are Already Ruining the Internet

    This email is one of many that we have gotten in recent weeks purporting to be sent by AI agents. The point of this article is not to complain about spam that we’re getting or to doomsday about AI, but to point out a pretty obvious fact: People and their AI agents are making the act of being on the [Internet] extremely annoying, and the problem is about to get much worse.

  • 404 Media ☛ AI Agent Platform Reinvents Spam, Floods Inboxes Worldwide

    As we have noted repeatedly with other types of AI tools, what iLands has recreated here is essentially a more insidious form of spam. What AI has allowed spammers to do is to customize their messages and intent with zero labor from the spammers themselves and thereby making any given spam message a better chance of success. Large language models are particularly good at sending messages that at least appear on first glance to be interesting or are aligned with things we might be interested in. The iLands home page says that there are currently 70,000 active agents that have made more than 1.6 million emails and posts.