
Photo by Dave Adamson on Unsplash
"Robert Anton Wilson: Useful Idiot?" New Substack by Michael Johnson.
New Joseph Matheny Substack. Includes archive of his old blog.
The latest bad news about pro football.
Anecdote about Patrick Stewart and Sting (bottom of the newsletter).
Recommendations from John Higgs.
Bobby Campbell recommends "You Can See Everything."

During my long stint as a consultant to organizations large and small, I got to work with, and for, companies and branches throughout the Americas and Europe. Over that time, I observed something that astonished me: workers in the US and UK were, compared to those in the rest of the 'west', remarkably, even exceptionally docile and obedient. Under the veneer of first-name-basis casualness, hierarchy was basically never challenged and 'subordinates' never dared question what they were told by their 'superiors' to do.
I couldn't figure this out. The countries that prided themselves most as lands of opportunity, where anyone could do anything they set out to do if they worked hard enough, seemed inordinately, unnecessarily, and dysfunctionally regimented and unquestioning.
On many occasions, I witnessed American and British middle-managers nod and quietly do what they had been told to do, even though they (and I) knew full well that these directions were uninformed, misguided, and incompetent, and would lead to suboptimal results.
Why did they do this? One manager told me that American organizations are notorious for lynching middle-level staff when things go wrong, since the old-boys' network at the top can never be punished, and somebody has to be demoted, penalized, or axed when targets aren't met. He was up for a promotion, he said, and one thing he'd learned was never to rock the boat. If things go wrong (which happens often when senior execs set impossible goals for staff), you always have to be able to say "I did exactly what I was told to do."
Continental European and Latin American managers I worked with commented on this passivity often, asking me, as a Canadian, to explain it to them. I could not. I was as dumbfounded as they were. Many of them had been promoted because they brought valuable fresh thinking and challenges to old, dysfunctional ways of doing things. They were innovators. They volunteered to try different things, and they had a 'safe-to-fail space' where they could try out different ways of doing things, and get them adopted when they worked, and quietly shelved when they didn't. I was brought up in the Canadian work environment the same way. You, and the organization you work with, learn by trying things and making mistakes. Innovations that make products and workplaces better are rewarded.
Although it's carefully hidden through accounting and tax tricks, the consequence of this passivity and lack of initiative among front-line workers who know much better than the executives what customers want and need is poor performance, low productivity and a complete lack of innovation. And hence the incapacity to compete with countries where this passivity and lack of initiative are less present.
Of course, Americans and Brits aren't stupid, and heaven knows they are quick to proffer their opinions when the boss is out of earshot. So how are we to account for this apparent docility?
I would theorize that citizens of these two countries are taught, from an early age, not to think, not to challenge, not to rock the boat, and to just do what they are told. You learn from experience that you get rewarded, and get ahead, by doing that. In the rigid quasi-military hierarchies that prevail in these two countries, there is no room for questioning superiors, no space for advancing new ideas, no reward for suggesting a better way of doing things. It is dangerous to stick your neck out, or to stick out at all.
Their decrepit, dumbed-down education systems make the situation worse: Just put your head down, copy and paste what the teacher has told you is the 'right answer'. If it's not going to 'be on the exam', don't waste your time with it.
And in both countries, the Epstein Class is well-entrenched, and through political and economic control, they're becoming ever-richer and more powerful. You challenge them at your peril.
What concerns me most is that I see the same trend occurring here in Canada and in other countries in the 'west' now. Very few workplaces any longer offer a real career path if you're not related to the boss, or offer any job security or much in the way of 'employee benefits' anymore. So why should workers offer loyalty to their employers? When they know the boss will take credit for their good ideas and blame them for the boss' mistakes, why would they not just leave their hearts, souls and minds at the door when they check in to work each morning? Why take a risk for no possible reward? Why think, when it can only bring awareness of how much better things could be were it not for arrogant, incompetent bosses and broken, dysfunctional systems.
In short, I think all of us, except for those very few in the upper castes, are becoming hyper-domesticated.
The term domesticated literally means "made property of the house". Once you are property — of your employer and your country — you are theirs to do what they will with you. Your role outside the workplace is to be a compliant, uncomplaining consumer of what the Epstein Class chooses to sell you, in whatever enshittified condition and at whatever surveillance price they choose to sell it to you. In the immortal words of Jerry Michalski, a "consumer" is "a gullet whose only purpose in life is to gulp products and crap cash."
At work your role is to be obedient and do what you're told, when you're told, how you're told, in return for whatever slave wages they choose to pay you — because they control all of that. You have no say.
And in public spaces, your role is to do what the 'authorities' tell you to do — be quiet, go to work, afterwards stay home, conform, buy the awful, overpriced, distracting entertainments the Epstein Class sells you, read the approved corporate-owned media, and obey the police, the military, ICE and whatever other government paramilitaries terrorize your particular neighbourhood. Do what you're told. Including, of course, joining the military and 'giving your life for your country' — or rather, giving your life to serve the reigning castes that run it.
All the major political parties in the west are owned outright by the Epstein Class, and paid to do their bidding. So now we live under a hierarchy of ultra-rich, out-of-touch, emotionally desensitized buffoons with no front-line or real-life experience, who are as a result the most incompetent members of our entire society. This caste of clowns is replete with war criminals, grifters, con men, cult-indoctrinated tech bros, and fraudsters of every stripe, all of them perpetuating their political and economic atrocities with impunity — because they've been brought up to believe they know better than everyone else, and can do no wrong, and because their obscene wealth allows them to control everything.
And no one is seriously challenging them. The electorate is understandably disenchanted and angry — when you start getting hungry and scared, the propaganda begins to wear thin. But all that happens then is that one of the Epstein-Class-owned parties moves to curry favour with this disenchanted electorate, by talking a good and well-rehearsed story. But once elected, they continue the same criminal, elitist programs they railed against. And the other Epstein-Class-owned party's members quickly slip through the revolving door into absurdly overpaid corporate jobs, until it's their turn to rouse the unhappy rabble and return their party to political power. The fix is in.
Although it's now more pronounced than ever, this is not a new phenomenon. It's the consequence of political, social, economic and other systems that have grown too big to be manageable, and are now utterly dysfunctional, and have now been hijacked by the Epstein Class to allow them to extract everything of value for their personal use ("privatization", the final enclosure of the commons), and then run for their compounds when there's nothing left to steal. These systems are beyond fixing, reforming, or redeeming. They are collapsing, and the signs of that collapse are obvious. The Epstein Class didn't cause that collapse (at least not deliberately), they're just exploiting it.
For most of us in the 'west', there are strong incentives to deny what is happening, or just muddle along and make the best of it. When a place devolves into chaos, the creatures most at risk are the ones that are most domesticated, and hence most dependent on the collapsing systems, and the least able to cope for themselves. I can't see any country in the 'west' handling this competently (the governments of 'nation states', most of which never made sense in the first place, aren't going to last much longer in any case).
So then it will be down to local communities to self-organize and try to survive collapse. This will be awful everywhere, because most of us have none of the competencies — either the 'soft' or 'hard' skills — that are soon going to be needed to help create and self-manage sustainable, collaborative communities.
And this will be worst where the people have been hyper-domesticated. Like puppies caught in the chaos of a house fire, they are, and will be, completely helpless and dependent, waiting for others to tell them what to do, and utterly incapable of contributing in any useful way to a collaborative effort of any kind. In our atomized, individualized, isolated, disconnected modern cities, the skills of collaboration and self-organization have mostly been lost.
That is the cost of domesticated obedience: A population helplessly incapable of change, even when our survival depends on it. Just waiting for some hero, some god, anyone, to tell them what to do.
image by AI
I finally got to see the new Tesla Model YL, Model Y Standard, and Model 3 Standard in person a few weeks ago and wanted to give my thoughts on the three models. I drove the two Model Ys and just looked at the Model 3 Standard. I was visiting ... [continued]
The post I Checked Out Tesla's New Model YL, Standard Y, Model 3 In Person — Here's What Surprised Me appeared first on CleanTechnica.
The Lincoln Memorial Reflecting Pool has been host to thriving populations of algae ever since its construction in the 1920's, so why not transform it into a showcase for the algae biofuel industry?
The post Trump Wants More Attention For His New Algae Biofuel Showcase appeared first on CleanTechnica.
Rich in sonic detail, Kapka Kassabova's 'Borrowed Land' is vital testimony to the ecocide and damage to human health being committed in the name of the energy transition, writes Sally Huband.

Sound is not sound alone, it's touch.
John Cotter, Losing Music
My nerves were on edge when I visited the Barbican's Feel the Sound exhibition two summers ago. I had become fearful of sound and wanted to transform this fear back into love. An energy company had built a hyperscale wind power plant near my Shetland home. The low frequency noise of the turbines rewired the auditory processing in my brain, among other neurological harms. I moved house and slowly recovered but louder sounds continued to cause pain.
I needn't have worried. I suffered no pain at the Barbican exhibition. The subwoofer bass of the souped-up cars in the underground carpark was a gentle nudge compared to the brain punch of industrial wind turbines. I'm deaf and have long known that sound can be felt as much as heard. It was good to see artists explore this. In a video exhibit, Evelyn Glennie talked of teaching the world to listen but with our bodies, hearts and mind. If I had to choose a book that does just this it would be Borrowed Land by Kapka Kassabova.
The deep listening in Borrowed Land takes place in the Scottish Highlands. The writing is rich in sonic detail. A cave becomes the inner ear of the glen and birches whisper. I hear the stories told in Borrowed Land but I feel them too. Kassabova's Highland home place, like my Island home place, is a sacrifice zone. Ecocide is being committed in the name of the energy transition and human health is collateral damage in the pursuit of corporate profit. Kassabova's lived experience is vital testimony. The low frequency noise of a new electricity substation sends workers home with migraines, panic, nausea, a malaise and Kassabova herself is made ill by the pylons that are built next to her home. She, like me, is forced to leave.
I've come to know and love the setting of Borrowed Land while staying at Moniack Mhor, a creative writing centre that sits high on a hillside. The rooms have views across a wooded glen, mountains form the backdrop and the light is always shifting. I mention Moniack's views deliberately. Those of us arguing that all is not well in the energy transition are accused of being more concerned with views than the climate emergency. If only this were true. A wind power plant called Ballach, with turbines 230 m high, is planned 10 km from Moniack Mhor. If it is built, I would hesitate to stay there again and not because the landscape would change. The soundscape would change, it would become unsafe.

Photo: Cliff Green
Each time that a turbine blade passes a tower, a pulse of low frequency noise is flung out. This noise is too low in frequency to be heard by the human ear but it can still be sensed. Think of sound waves as waves of pressure and of turbine harm as barotrauma. Bats can die from pulmonary barotrauma if they fly near turbines, the blood vessels in their pliable lungs burst. In humans, turbine barotrauma causes neurological harm. My brain was rewired and my heart was hacked, it would race uncontrollably. I could tell if the turbines were on or off without looking. It is difficult to describe sensing a turbine other than to say that it is sensory torture.
When I explain turbine acoustic harm to people, they ask if noise cancelling headphones help, or sound absorbing panels. The answer is no. Blade pass emissions cannot be stopped from entering a home, or a body. These pulses can be as low as 0.1 Hz which means soundwaves that are vast in scale and travel many kilometres. I now live 14 km from the hyperscale power plant here in Shetland but I can still sense the turbines in settled weather. The taller the turbine, the worse the blade pass pollution. Ballach wind power plant must never be built.
Building wind power plants a safe distance from homes and workplaces would increase construction and operation costs and decrease profits. This is why the wind industry claims that turbines are safe. They argue that the harm is psychogenic in origin (we are simply stressed because turbines have been built in our view). Despite mounting evidence of direct harm to human health, governments concertedly look the other way and for a while I did not understand why but in Scotland, at least, the reason is becoming alarmingly clear.
Renewables produced more electricity than is consumed in the country for the first time in 2022. The construction of new energy infrastructure continues apace because Scotland is being positioned as a 'green' data centre superpower. Twenty-six hyperscale AI datacentres have been proposed so far. As Kassabova astutely observes in Borrowed Land, renewable energy infrastructure is being built in Scotland to open new markets for profit and not for the good of the planet. These profits do not stay in Scotland. Promised jobs do not materialise and fuel poverty remains. The energy company responsible for ecocide and harm to health in Borrowed Land is the same company that built the hyperscale power plant here in Shetland. Its main shareholder is BlackRock, named by the UN as complicit in the economy of genocide.
Shetland is not a wooded place and I am drawn to Kassabova's portrait of a low impact forester called Dylan. He extracts timber with Nomad the horse, a process that Kassabova describes as hard toil and bright joy. We meet him sitting on his cabin porch listening to rain falling on leaves. In spring, he listens for the sound of sap rising before planting trees. It used to rise in March but now rises in February. The power company intend to build a pylon where he is constructing a barn for Nomad. If he does not agree to leave, the land will be compulsory purchased. As Dylan concludes, We need renewable energy, but this is just wrong. Borrowed Land is a book ahead of its time, a clarion call that cuts through all the corporate energy company noise.

The 103 turbine Viking Wind Farm, Shetland. Photo by Sally Huband
*
Kapka Kassabova's 'Borrowed Land: A Highland Story' is out now and available here, published by Jonathan Cape (£20.90).
Sally Huband's memoir, 'Sea Bean', won the Highland Book Prize and 'saltwaterisland&', produced with Gaada, recently featured in the Dundee V&A Design and Disability exhibition. For 'The Book of Bogs' anthology, Sally co-wrote with Polly Atkin an innovative call and response piece on turbine acoustic harm.
- A record August and the second-warmest year to date
- A record-setting El Niño event
- Record heat across much of the world
- Good odds of being world's warmest year
- Datasets are split on whether 2026 is headed for a record
- 2027 very likely to be the warmest year on record
- Earth's energy imbalance has more than doubled
- Arctic sea ice minimum tied for 10th lowest, Antarctic near record lows
With three-quarters of the year now past, it is virtually certain that 2026 will be the warmest or second-warmest year on record.
A rapidly strengthening El Niño, alongside human-caused global warming, made this year's August the warmest on record and September the second-warmest.
El Niño is the warm phase of a recurring climate pattern in the tropical Pacific that releases heat from the ocean into the atmosphere and temporarily boosts global temperatures.
Carbon Brief's analysis gives 2026 a 67% chance of surpassing 2024 as the warmest year on record across an average of six global temperature datasets.
In January, it is likely the groups that report global surface temperatures will be divided on whether 2026 officially beat 2024's record.
Currently, the odds for a record year range from as high as 97% in Berkeley Earth's dataset to less than 20% in European Copernicus/ECMWF's ERA5 reanalysis product.
With El Niño expected to intensify into the winter, Carbon Brief projects that 2027 will be 1.7C above pre-industrial levels, with a 96% chance that it is the warmest year on record.
In this latest "state of the climate" quarterly update, Carbon Brief finds:
- August 2026 was the warmest August on record in all six major datasets, at around 0.12C above the previous record. September was the second-warmest September, behind 2023.
- January-September was the second-warmest such period on record, at around 1.45C above pre-industrial levels, behind only 2024.
- El Niño broke the daily "Niño3.4" region record in September and has stayed above the previous record every day since. Nearly all forecast model runs have it as the strongest El Niño on record, with a peak of around 4C occurring this winter.
- Around 22% of the world's surface saw its warmest July-September on record, including much of the tropical Pacific, the western Mediterranean and Iberia. Western Europe had its hottest summer on record and Spain its warmest September, beating the previous record by 0.7C.
- Carbon Brief projects that 2026 will end up around 1.52C above pre-industrial levels, with a 67% chance of beating 2024's record.
- Whether 2026 is declared the warmest year depends on the dataset; a record is likely in Berkeley Earth and NASA GISTEMP, is too close to call in HadCRUT5, JRA-3Q and NOAA and is unlikely in ERA5.
- Earth's energy imbalance has more than doubled since the start of the satellite record, from around 0.5 watts per metre squared (W/m2) in the 2000s to around 1.25W/m2 over the past decade.
- The Arctic sea ice minimum was tied for 10th lowest in the satellite record, while Antarctic sea ice saw its third-lowest winter peak.
- This year's record-setting El Niño will have its largest effect on global temperatures in 2027. Carbon Brief projects 2027 at around 1.7C above pre-industrial levels, with a 96% chance that it will be the warmest year on record.
A record August and the second-warmest year to date
Carbon Brief analyses records from six different groups that report global surface temperatures:
- NASA's GISTEMP
- NOAAGlobalTemp
- The Met Office Hadley Centre and University of East Anglia's HadCRUT5
- Berkeley Earth
- Copernicus/ECMWF's ERA5 reanalysis product
- The Japan Meteorological Agency's JRA-3Q reanalysis product
There are a number of additional groups that feature in Carbon Brief's annual "state of the climate" update, but are not included here due to a lack of regular monthly updates.
The first nine months of 2026 were the second warmest on record in all six datasets, behind only 2024.
The figure below shows annual temperatures in each dataset since 1970, along with the 2026 average to date.
Global mean surface temperatures datasets aligned on a common 1981-2010 baseline, then shifted by the average 1850-1900 to 1981-2010 warming (0.67C) in HadCRUT5, NOAA and Berkeley Earth. 2026 to-date dots are staggered horizontally to avoid overlaps and the HadCRUT5 is for January-August as September data was not available at time of publication.
The year got off to a comparatively slow start. Each month from January to June ran cooler than the same month in 2024, as the lingering effects of weak La Niña conditions gave way to a developing El Niño.
La Niña and El Nino are the cool and warm phases of the El Niño-Southern Oscillation (ENSO), a recurring climate pattern in the tropical Pacific that reshapes weather patterns around the world.
That changed in the third quarter of the year. July 2026 was the second-warmest July in the six-dataset average and a narrow record in the NASA GISTEMP and NOAA datasets. (NOAA's own assessment described it as tied with July 2024.)
August then set a new record for the month by around 0.12C in the six-dataset average. It was the warmest August on record in every one of the six datasets, by margins ranging from 0.07C (HadCRUT5 and NOAA) to 0.13C (ERA5).
September temperatures stalled a little following the rise in July and August. At around 1.62C, it was the second-warmest September on record in all five datasets to have reported so far, around 0.08C below the record set in September 2023.
The figure below shows each month of 2026 alongside every year since 1940, with 2023, 2024 and 2025 highlighted.
Monthly averages of six global mean surface temperature datasets. The September datapoint excludes HadCRUT5 which had not reported at the time of publication.
Across the six-dataset average, July-September 2026 was effectively tied with the same period in 2023 as the warmest on record, at around 1.57C. It ranked first in GISTEMP and NOAA (and in HadCRUT5 for July-August), and second in Berkeley Earth, ERA5 and JRA-3Q.
The figure below shows the January-September average for every year since 1850. It shows how temperatures in 2026 to date sit at 1.45C - around 0.06C below 2024.
A record-setting El Niño event
In September, El Niño crossed a milestone. The daily sea surface temperature anomaly in the "Niño3.4" region of the tropical Pacific - the area most widely used to gauge El Niño strength - broke the record, previously set in November 2015, to become the strongest event since the satellite era began in 1981.
This year's El Niño has remained above that record every day since. September recorded the highest monthly Niño3.4 region temperature anomaly since records began in the mid-1800s.
As of 18 October, the daily value stood at around 3.5C above the 1996-2025 average for the time of year and around 0.4C above the 2015 peak.
The figure below shows how El Nino is projected to keep intensifying until the end of this year.
Average of 656 forecasts across 14 dynamical models. The lighter shaded area reflects the period when only seven models are available.
Carbon Brief examined 656 forecasts from 14 different climate models that were run between late September and mid-October. On average they estimate that this year's El Niño will peak around 4C in December - more than 1C above the prior record.
(Ten models expect a December peak and four anticipate a November peak.)
Based on an analysis of tree rings and coral sediments, it is possible that this will be the strongest El Niño event in a millennium.
Niño3.4 region sea surface temperature peak warming projected from 656 runs of 14 different dynamical El Niño models. Y-axis denotes the share of model runs in each 0.2C "bin".
The forecasts have strengthened since Carbon Brief's July "state of the climate" update, when El Niño was expected to peak around 3.6C.
While human emissions of CO2 and other greenhouse gases are responsible for effectively all of the Earth's long-term warming, temperatures in any given year are strongly influenced by short-term variations to the Earth's climate associated with El Niño and La Niña events.
Global temperatures tend to lag the tropical Pacific by around three months, which means an El Niño peaking around the end of 2026 will have its largest effect on global temperatures in 2027.
Record heat across much of the worldIn addition to high global average surface temperatures, the past three months have seen records broken across many parts of the planet.
The map below shows July-September 2026 temperatures relative to the 1981-2010 average in the ERA5 reanalysis.
Globally, temperatures were 0.96C above that 30-year period and 1.58C above the ERA5 1850-1900 pre-industrial average.
The El Niño is clearly visible as a band of exceptional warmth stretching across the eastern and central tropical Pacific, with temperatures more than 4C above average off the coast of Ecuador and Peru.
Western Europe and the Mediterranean were also unusually warm, with temperatures around 3C above average over France and the western Mediterranean Sea. North-west Russia, the Canadian Arctic, south-central North America, eastern Africa and parts of west Antarctica were also exceptionally warm.
The most notable part of the planet that was colder than usual was the Weddell Sea sector of Antarctica, which came in at more than 2C below average for the season.
The map below shows how July-September 2026 ranked against every year in the ERA5 record (1940-2026) at each location. Areas in dark red saw their warmest July-September on record; blues show areas among the five coldest.
Areas where hot and cold records were set from July-September 2026 compared to the same period in all other years since the ERA5 dataset began in 1940. Share of global surface: 21.5% record warm, 49.9% top-five warm, 0.5% top-five cold and <0.1% record cold.
Around 22% of the world's surface saw its warmest July-September on record and around half experienced one of its five warmest. By contrast, around 0.5% of the world saw one of its five coldest July-September periods.
Record warmth covered most of the El Niño region (including more than 90% of the Niño3 and Niño3.4 regions) along with sizeable parts of the south Atlantic (39%), the tropical Indian Ocean (34%) and the subtropical North Atlantic (25%, with 85% in its five warmest).
On land, around 85% of France and Iberia saw a record-warm July-September.
In September alone, around 20% of the world's surface saw its warmest September on record.
Overall, 55 countries set national records for September average temperatures, including Chile, Colombia, Italy, Kenya, Nigeria, the Philippines, Portugal, Spain, the United Arab Emirates, US and Venezuela.
Europe had an exceptional summer, with the western part of the continent experiencing its hottest summer on record.
Copernicus reported that western Europe's summer averaged 21.7C - 2.5C above the 1991-2020 average - beating the previous record set in 2003. A record 52% of the continent was exposed to "very strong" heat stress by the end of the season.
Europe as a whole had its third-warmest summer, behind 2024 and 2022.
France had its warmest July on record, at 3.8C above the 1991-2020 average. Switzerland had its hottest summer since records began, averaging 17.2C.
The heat persisted into the autumn in south-west Europe. Spain had its warmest September on record, at 3.4C above the 1991-2020 average. This beat the previous record set in 1987 by a remarkable 0.7C.
Rivers across Europe saw their lowest summer flows since the European Flood Awareness System's records began in 1992. By early September, the area burned across the EU was almost twice the 2006-25 average for the time of year, according to Copernicus, extending what had already been a severe fire season.
Elsewhere, the contiguous US had its hottest summer on record and its warmest September, with around 59% of the country in drought conditions.
India's monsoon, which tends to weaken during El Niño years, ended with 12.6% less rainfall than normal, the driest in 11 years.
This year's El Niño event has strongly impacted the hurricane season, suppressing the formation of hurricanes in the Atlantic while contributing to a record-setting number of tropical cyclones in the Pacific.
Good odds of being world's warmest yearCarbon Brief's projection for average global temperature for 2026 combines observed temperatures through September with the latest forecasts of the strength of El Niño. It uses a statistical model trained on the historical relationship between year-to-date temperatures, El Niño conditions and annual temperatures.
The figure below shows the six-dataset average for each year since 1970 (dark blue), the year-to-date January-September 2026 average (blue circle) and Carbon Brief's 2026 estimate (red circle) with its 90% likelihood range (vertical red line).
The dark red dot marks Carbon Brief's median projection for 2027, the bar the 50% range and line the 90% range.
The best estimate for 2026 is now around 1.52C above pre-industrial levels, with a 90% range of 1.50C to 1.55C. This is slightly higher - but effectively tied with - the prior record of 1.52C set in 2024.
It marks a slight increase from Carbon Brief's previous projection of 1.51C, set out in its July's "state of the climate" update.
There is an approximately 67% chance of setting a new record this year in the average of the six-datasets examined.
The year 2026 also has a 93% chance of coming in above 1.5C in the six-dataset average (though not in every individual dataset), which would make it the second year above 1.5C, after 2024.
However, a single year above 1.5C does not itself constitute a breach of the Paris Agreement goal, which reflects a longer-term average.
These odds also depend on assumptions that are hard to pin down.
The El Niño forecast for October-December lies beyond any event in the record the model is trained on. The projection assumes the historical relationship between the strength of El Niño and its impact on global temperatures holds for an event this large. If the warming effect of this El Niño turns out to be no larger than that of the strongest past events, the odds of a record year for global temperature would be closer to 50%.
Datasets are split on whether 2026 is headed for a recordDifferent groups of scientific researchers around the world will announce their own results for 2026 in January.
Across the six datasets that Carbon Brief explores there will likely be mixed results, with some datasets showing 2026 as the warmest year on record - and others showing it as the second warmest year.
Likelihood of a 2026 record in different datasets using Carbon Brief's model. Carbon Brief's projection is based on each dataset's observations and record. Probability for HadCRUT5 is based on data through August, as September data was not available at time of publication.
Based on temperatures to-date and the current El Niño forecast, Carbon Brief estimates that a 2026 record is very likely in Berkeley Earth (97%) and likely in NASA GISTEMP (82%). It is about as likely as not in JRA-3Q (48%) and NOAA (34%) and unlikely in ERA5 (18%).
HadCRUT5 currently shows a 72% chance of a 2026 record, but this is based on data only through August. Plausible values for September - which the other datasets suggest was slightly cooler than August - would put its odds somewhere between around 50% and 80%.
The difference in odds comes down to how far each temperature group's data for 2026 trails the prior record set in 2024.
For example, over January-September, Berkeley Earth has 2026 around 0.03C cooler than the same months of 2024, so its final three months need to beat those of 2024 by around 0.08C.
ERA5, by contrast, has 2026 around 0.09C behind 2024 over the same period, meaning October-December would need to run around 0.27C warmer than the last three months of 2024.
None of these datasets is "right". They differ in how they fill gaps in regions where fewer observations are taken, such as the poles, and in which sea surface temperature records they use.
The two reanalysis products, ERA5 and JRA-3Q, take a different approach altogether, using a weather model constrained by observations to estimate air temperatures everywhere, rather than combining land and ocean measurements.
There is precedent for a split decision. In January 2015, NASA and NOAA announced that the year 2014 was most likely the warmest on record, while the Met Office described it as tied with 2010. January 2027 could see something similar.
2027 very likely to be the warmest year on recordThe biggest temperature effect of this year's El Niño event will be felt next year. Just as the 2015-16 El Niño helped make 2016 the then-warmest year on record, an event peaking around the end of 2026 will have its largest influence on global temperatures in 2027.
Carbon Brief projects 2027 will end up at around 1.74C above pre-industrial levels, with a relatively wide 90% uncertainty range of 1.54C to 1.95C.
This is based on the historical relationship between year-on-year temperature changes and El Niño conditions in the preceding autumn, factoring in the remaining uncertainty around where temperatures in 2026 will end up.
The figure below shows the six-dataset average for each year since 1970 (dark blue), along with Carbon Brief's projections for 2026 (orange) and 2027 (red).
The dark red dot marks Carbon Brief's median projection for 2027, the bar the 50% range and line the 90% range.
This gives 2027 a 96% chance of being warmer than any previous year, including 2026. There is a 98% chance that at least one of 2026 or 2027 will set a new record.
The estimate for 2027 is necessarily much more uncertain than for 2026. None of the year has been observed yet and the result depends on both the peak strength of the El Niño and how quickly it fades.
As noted above, this year's El Niño strength is currently projected to peak well beyond the strongest event in the record on which Carbon Brief's model is trained. Carbon Brief predicts a rise in temperatures between 2026 and 2027 of around 0.2C due to El Niño - larger than the jump that followed any previous El Niño events. The projection assumes the historical relationship between El Niño and global temperatures will hold for an event this large, but it remains to be seen whether this is the case.
Nevertheless, even an average global temperature on even the low end of the range (1.54C) would exceed 2024 and, while Carbon Brief's central estimate of 1.74C would beat it by more than 0.2C.
If these projections bear out, the 2020s will have delivered new records in 2023, 2024, 2027 and, quite possibly, 2026.
Earth's energy imbalance has more than doubledThe rise in global temperatures is a symptom of Earth's rising energy imbalance. This is the difference between the energy the planet absorbs from the sun and the energy it radiates back to space.
As long as this energy imbalance is positive, the Earth system accumulates heat, around 90% of which ends up in the oceans.
Earth's energy imbalance has more than doubled in recent decades. This growing imbalance means the planet is accumulating heat ever faster and that global warming is accelerating.
NASA's CERES satellite instruments have tracked changes in this imbalance at the top of the atmosphere since March 2000.
The figure below shows the Earth's energy imbalance from 2000 to August 2026, with monthly values (faint line), 12-month averages (dark blue), calendar-year averages (dots) and the linear trend (red dashes).
Earth energy imbalance measured by NASA's CERES satellites.
The Earth's energy imbalance rose from an average of around 0.53W/m2 over the first decade of the record (March 2000 to February 2010) to around 1.25W/m2 over the most recent decade (March 2016 to February 2026) - a 2.4-fold increase.
The highest annual energy imbalance recorded to date was in 2023 at 1.79W/m2, while the most recent 12 months (September 2025 to August 2026) averaged at 1.59W/m2.
Several studies, including 2026 research covered by Carbon Brief, have noted that the imbalance appears to have risen faster than most climate models simulate, though the reasons remain unclear.
The measurements also come with some caveats. CERES tracks changes in the imbalance well, but its absolute level is anchored to ocean heat content measurements. Changes to satellite orbits over time add a further uncertainty of around 0.1W/m2 per decade to the trend.
A 2021 study found that independent satellite and in situ observations "each yield statistically indistinguishable decadal increases" in the imbalance between 2005 and 2019.
Over longer periods, CERES shows more energy being accumulated than ocean heat estimates set out by Institute for Atmospheric Physics data and by the authors of the Indicators of Global Climate Change report.
The imbalance typically dips as a strong El Niño peaks and the ocean releases heat to space, as it did in 2024. Next year's energy imbalance may be slightly lower.
Arctic sea ice minimum tied for 10th lowest, Antarctic near record lowsAround mid-September of each year, Arctic sea ice drops to an annual low, whereas Antarctic sea ice grows to its largest area of the year.
On 12 September, Arctic sea ice reached its annual minimum extent, at 4.60m square kilometres (km2) on the five-day average used by the US National Snow and Ice Data Center (NSIDC).
The 2026 minimum shares the 10th lowest spot - alongside 2008, 2010 and 2025 - in the nearly 48-year satellite record and was 1.21m km2 above the record low set in 2012, according to NSIDC.
That represents a modest recovery from earlier in the year, when Arctic sea ice set, or tied, record daily lows on 39 days between January and 2 July, following a joint-lowest winter maximum.
Arctic sea ice extent has not set, or tied, a daily record low since 2 July and has been within the 1979-2010 range for the time of year since the start of August, though still well below the long-term average.
(As Carbon Brief has previously explained, conditions earlier in the year are a poor predictor of the September minimum).
In the southern hemisphere, Antarctic sea ice reached its winter maximum on 14 September, at 17.59m km2. This was the third lowest on record, behind only 2023 and 2024. It recently fell sharply and hit record daily lows for much of the period since early October.
The figure below shows daily Arctic and Antarctic sea ice extent in 2026 (solid lines), compared with the 1979-2010 range (shading) and the record low for each day (dotted line).
Daily Arctic and Antarctic sea ice extent in 2026 compared to the prior record low and the 1979-2010 historical range. 2026's data runs until 17 October. Arctic minimum and Antarctic maximum calculated using NSIDC's five-day mean.
related
Experts: How are climate scientists using artificial intelligence?
22.10.2026
Climate modelling
UNEP: Opportunity has 'been lost' to meet 1.5C warming limit without 'overshoot'
21.10.2026
Climate pollutants
Guest post: Why tough methane cuts are crucial for keeping warming 'well-below' 2C
01.09.2026
Climate pollutants
Explainer: The CMIP7 emissions scenarios - and how they explore future climate change
01.09.2026
Climate modelling
The post State of the climate: 2026 likely to tie with 2024 as the warmest year on record appeared first on Carbon Brief.
With globe-spanning artists - including Charlotte Church doing angry electropop - the Flow project is hoping to make an impact at this month's Cop31
Amid tense and long-winded negotiations, and with the future of the planet at stake, emotions can often run high at Cop climate summits. But last year, when the annual conference arrived in Brazil, one UN official was unexpectedly moved to tears by five women singing about the power of water.
"She gave me a massive hug after the event and said, 'We need to bring music into Cop,'" says Nikita Llerena, the project manager for Flow, a music collective that's hoping to bring art to the centre of the climate movement. What started as a handful of musicians meeting over Zoom last year has grown into a full-length album that unites female artists from around the world, including Charlotte Church, Maxine Peake and Nadine Shah. This month, members of the collective will return for Cop31, in Turkey, where they have been promoted to a slot at Global Cultural Action's opening event.
Continue reading...
Most JavaScript programmers default to their ride-or-die frameworks, but such loyalties can hinder them when they task AI agents with doing their dirty work. Human-friendly tools aren't necessarily the most effective ones for the model-powered minions.
"One of my pet peeves with the web development community is that so many of the decisions we make are vibes-based," wrote Seattle-based Nolan Lawson, currently a developer at AI security firm Socket. That's an understatement; the JavaScript wars have been well underway for over a decade now, with each new framework promising to cure the woes of last year's release.
Not wanting these vibey biases to seep into the realm of AI, Lawson created a benchmark to compare how different JavaScript frameworks perform under the same agentic model. He subsequently ran the test on 12 JavaScript frameworks, publishing the results in a post that compared each framework in terms of build cost, maintainability, and end-user performance of their agent-built apps.
The results are illuminating. In general, agents tend to work more efficiently with server-centric frameworks than with client-heavy "Single-Page App" (SPA) frameworks (which should tickle for React haters).
But the biggest take-away should be less controversial: agents best utilize those frameworks that favor convention, API stability, and concision.
How to compare JavaScript frameworksFor the experiment, Lawson wrote a set of instructions for a Claude Code agent to build a basic Pokémon web application using the PokéAPI. The model used the identical setup on 12 different frameworks and JavaScript stacks: Angular, Astro, Gracile with Lit, Hono with htmx, vanilla JavaScript with Vite, Next.js, Nuxt, Qwik, React Router, TanStack Start, SolidStart, and SvelteKit.
Lawson evaluated each stack across five independent sandbox runs; the repeated iterations smoothed the variance in the model's performance.
To verify the resulting app met the specs, Lawson deployed a custom automated test suite that queried the browser's accessibility tree to ensure interactive elements like buttons and inputs weren't faked - a shortcut LLMs occasionally take. Lawson also used Google Lighthouse to measure transfer size and runtime performance.
If the frameworks specified official MCP servers, Claude plug-ins, and agent-skill files, Lawson provided them as resources to the agents (though they weren't always consulted). Otherwise, he did not mandate any specific architecture for rendering the app.
Comparing 12 JavaScript frameworks
The differences between the most cost-effective builds and the most expensive ones are substantial, to say the least, with the best lapping the efforts of the stragglers.
If you want to build a Pokéworld cost-effectively, the strongest bet is on Hono paired with htmx. This tag-team of technologies had the most economical run, with a median cost of $2.67 per Pokéapp.
Perhaps this is not surprising, given that the duo runs entirely on the server, thereby eliminating a whole set of checks required to synchronize an architectural split between client and server. It also helped minimize the code, which resulted in fewer output tokens being consumed.
Nearly as effective was Astro ($2.76). Like Hono/htmx, Astro does everything server-side. Server-side data crunching also helped Astro achieve the fastest median time to build and test the app, at 8.2 minutes.
Nuxt ($2.82) was the most cost-effective option for an SPA-like approach. Nuxt renders code on the server that gets hydrated on the client. Nuxt is actually built on Vue.js, but is "opinionated," eliminating a lot of boilerplate that the agent must write. Nuxt also handles data transfer and file routing internally, which is more code not to fuss over.
It's also worth noting Nuxt consumed the fewest tokens of the bunch (3.85 million), though that did not make it the cheapest, given how Claude charges different rates for input and output tokens. Nuxt produced the app in 1,415 lines of code, which is second only to Hono/htmx's 1,370 lines.
Overall, these top three frameworks were within the margin of error of each other, which means any one of them could have gotten the top score depending on what kind of day the LLM was having.
Some of the more traditional frameworks lumbered way behind, however.
A SvelteKit-based app cost $4.28 to generate, with the agent repeatedly going back to the SvelteKit documentation to get a better handle on how the recently-updated Svelte works. Angular was the most expensive way for an agent to build an app, costing $4.97 while taking 17.2 minutes and chewing up 9.76 million tokens (with the agent failing to even finish a working app in two of its five attempts).
All the resulting apps aced the Lighthouse accessibility and performance tests. It's a pretty simple app to render, Lawson acknowledged.
Astro's app had the fastest loading time (called Largest Contentful Paint or LCP) of 810 ms, while the slowest frameworks in the cohort - Next.js (2,090 ms) and TanStack Start (2,028 ms) - took more than twice as long to deliver the goods, perhaps because of all the JavaScript they shipped to the client.
Gold stars for allLawson hopes this work will set the stage for applying a bit more rigor in selecting frameworks for AI jobs, now that developers won't have to do the actual heavy lifting themselves.
But a few developers on Hacker News pointed out AI efficiency isn't the only factor to consider. There's the human component of maintaining the framework as well.
"After a decade of React, I'm not going to go learn htmx or any other marginally better same-shaped tool for a new project, especially when it'd degrade my ability to review and debug another actor's work product," one wrote.
After all, the frameworks themselves might evolve with the help of AI, minimizing their differences over time. "Besides a rewrite to another framework using AI, another option is to modernize the framework with AI instead," someone else noted. ®
Twenty-nine NHS trusts listed as benefiting from Palantir's Federated Data Platform (FDP) said in Freedom of Information responses that they were not using it, according to health worker campaign group Medact.
That represents about 21 percent of the 137 trusts NHS England lists as "realising benefits" as ministers weigh whether to continue the contract.
The government must decide in December whether to continue the £330 million agreement with a Palantir-led consortium supplying the platform.
Medact, which is campaigning against the deal, collected the responses to scrutinize the government's case for retaining Palantir. The relationship between the US tech biz and the NHS began during the pandemic with contracts worth £60 million awarded without competition.
NHS England awarded the FDP contract in November 2023. It says the platform brings together operational data held in separate systems to help staff work more efficiently.
NHS England's September figures list 142 trusts as live on the platform - 137 of which it claims are benefiting.
Among the disputed entries, NHS England cites University Hospitals of North Midlands NHS Trust as receiving benefits from the platform, but in response to an FoI request in July, the trust said: "FDP is not live. We do not intend to roll it out in the future. A variety of factors prevent roll out, but primarily it does not appear to provide functionality over and above what we have already."
Meanwhile, a response from Nottingham University Hospitals said "no implementation timeline has been agreed," according to The Guardian.
The Register asked NHS England to explain the discrepancy.
An NHS spokesperson said: "The NHS Federated Data Platform is universally available for all NHS trusts and integrated care boards across England, with local NHS organisations able to select individual tools depending on their needs.
"137 NHS trusts have now reported benefits from the wide range of tools available on the platform - helping staff join-up and improve care for patients, and work more efficiently."
Palantir maintains its software is playing an important role in improving patient care and helping to deliver 100,000 additional operations, a 12 percent reduction in discharge delays and the removal of 675,000 patients from waiting lists.
In July, the UK's statistical regulator told NHS England to ensure greater accuracy in quoting figures following a row over data used to promote the purported benefits of the FDP.
Dame Chi Onwurah, the chair of the House of Commons Science and Technology Committee, has given the government until today to clarify its response to questions over whether NHS England has any exit plans should it decide to end the Palantir contract in February.
James Frith, a junior minister in the Department of Health and Social Care, wrote [PDF] to the Committee in October explaining that the contract has "measures built in to facilitate exit and transition including clear intellectual property rights in NHS build products or components."
He continued: "Publishing more detailed information on the Exit Plan would, or would be likely to, prejudice commercial interests and NHS England's ability to secure best value for taxpayers in future commercial activity."
Onwurah replied: "I understand that you will not publish the exit plan, but will you confirm if one has been prepared?"
The Register asked Palantir to comment. ®
The person in charge of digital strategy for the Senate campaign of Iowa Republican Rep. Ashley Hinson helps run a company that received nearly $10 million to influence Americans with pro-Israel propaganda.
In September 2025, Hinson — who is battling Democratic state Rep. Josh Turek in a tight race that could decide control of the Senate — tapped Abe Adams to lead her campaign's digital efforts.
Adams is a co-founder and managing partner of Targeted Victory, a digital consulting firm that has been used by high-profile Republican candidates such as Ted Cruz and Marco Rubio. Targeted Victory shares an address and much of its staff with Targeted Communications Global — a related company with no discernible online presence. Both are subsidiaries of marketing company Stagwell Global.
Israel awarded Targeted Communications a $9.9 million contract in March 2025 focused on social media and paid advertising, documents show. By October 2025, the firm was at work pitching news content and guests on Fox News and CNN, and creating YouTube ads suggesting young Iranians have been educated to believe the "most sacred thing they do is stab a Christian in the heart."
Hinson is endorsed by both President Donald Trump and the American Israel Public Affairs Committee, and has been supportive of Trump's war in Iran.
A June 2026 Department of Justice FARA filing from Targeted Communications details much of the firm's work for Israel. The document lists 10 registered foreign agents working on the Targeted Communications contract for Israel. Six of them are identified as Targeted Victory employees on the company's website or their LinkedIn accounts. Adams is not named in the filing.
Hinson, Adams, and Targeted Victory did not immediately respond to requests for comment.
Adams is not the only Hinson campaign adviser with ties to Targeted Victory and the Israel lobby. Hinson's current deputy campaign manager, Addison Lavis, worked at Targeted Victory before serving as Hinson's campaign manager in her 2023 congressional run. Lavis was previously an intern at AIPAC. Lavis did not immediately respond to a request for comment.
Related
AIPAC Isn't Retreating. Here's How It's Already Shaping the Next Congress.
Hinson has received more than $100,000 from AIPAC and related pro-Israel PACs in 2026, and as an Iowa state representative supported legislation that prohibited the state from working with any companies that boycott Israel.
Hinson's political action committee has paid Targeted Victory more than $1.6 million this election cycle for digital consulting and ads — 27 percent of its total expenditures, according to Federal Election Commission filings.
Israel's deal with Targeted Communications is part of a major public relations push that has pumped more than $100 million into swaying public opinion, particularly among young conservatives, back toward Israel.
The post Ashley Hinson's Digital Chief Oversees Firm Paid Millions to Push Israeli Propaganda appeared first on The Intercept.






I thought about her a lot after our friendship imploded. When our birthdays rolled around, I wondered what we'd be doing if we were still in touch
I was 26 when I lost my childhood friend. By that point, I'd known her for almost two-thirds of my life.
Our first conversation was at an ice-skating party when we were nine years old. I'd been nervous in the weeks leading up to the party because I couldn't ice-skate. Every time the Spice Girls' 2 Become 1 music video came on, the sight of Victoria Beckham singing on an ice rink filled me with impending doom. I was sure I would disgrace myself at the party.
Continue reading...Britain's AI infrastructure boom could soon have an unlikely new home: Drax Power Station, where the operator of the country's largest biomass power plant has unveiled plans for a £20 billion datacenter development capable of consuming more than a gigawatt of power.
Drax announced on Tuesday that it has teamed up with investment giant Brookfield and its UK-based AI infrastructure business Radiant to build the facility at its North Yorkshire site. The money would come through a joint venture between Brookfield and Qatar's Qai, although the £20 billion headline figure is more ambition than actual money on the table at this stage.
The companies plan to start relatively small, at least by the standards of the AI industry's apparently bottomless appetite for electricity. The first phase would deliver up to 100MW of capacity by early 2028, backed by a proposed £2.5 billion investment. A second phase could add another 500MW between 2028 and 2031, with further expansion taking the site beyond 1GW after that.
However, a few hurdles remain between today's announcement and a functioning AI megafactory. According to Drax's regulatory filing, the first phase still requires planning permission, regulatory and network approvals, commercial agreements, and final investment decisions.
It also needs a tenant, which is a fairly important detail when you're proposing to spend billions on somewhere to house other people's computers.
Drax reckons its existing infrastructure gives it an advantage over developers struggling to secure electricity connections. The 1,000-acre site already has access to 4GW of grid capacity, 2.6GW of active generation and cooling infrastructure, along with proximity to the UK's fiber network.
The UK government estimated last week that all of Britain's existing data centers use around 2GW of electricity. Drax's proposed development alone could eventually need more than 1GW, although how much power it actually consumes will depend on how fully it's used.
The first phase would draw electricity from the grid, but Drax wants to supply up to 500MW directly from its existing generating capacity under a long-term power purchase agreement, potentially allowing AI servers to run on electricity produced by burning imported wood pellets.
That brings another complication. Drax signed a government-backed Contract for Difference agreement last year covering biomass generation between April 2027 and March 2031. The agreement allows the company to request permission to redirect up to 500MW of capacity towards a datacenter, but ministers and the Low Carbon Contracts Company would have to agree, taking into account consumer value, energy security, and sustainability.
The announcement also arrives as enthusiasm for enormous AI server farms is proving rather less universal outside government and industry circles. Opposition to data center developments has been growing on both sides of the Atlantic, with communities raising concerns about electricity bills, water consumption, and the impact of industrial-scale facilities on their surroundings.
A recent investigation found that businesses accounted for 84 percent of the stakeholders UK ministers met to discuss datacenters between July 2024 and June 2026, while environmental groups secured precisely zero meetings.
Drax and its partners are promising a community benefits fund and say public consultation will begin this month. Whether that will reassure residents concerned about another enormous consumer of electricity is a moot point.
Drax CEO Will Gardiner said the project could "help secure the long-term future of the site and protect hundreds of jobs," while supporting Britain's ambitions to become a major player in AI.
The government, unsurprisingly, is enthusiastic. Chief Secretary to the Treasury Emma Reynolds welcomed the proposed investment as an opportunity to create jobs and support Britain's reindustrialization.
"Investments like this will help pave the way to a new age of British reindustrialization, creating good jobs in Yorkshire and delivering growth right across the country," she said.
Ministers may be enthusiastic, but Drax still has to find a customer, secure planning permission, and convince regulators before anyone starts plugging in servers. ®
Virgin Media O2 (VMO2) wants to chop a third of its labor costs and is expected to lay off thousands of workers, giving staff more to think about than Christmas presents during the run up to year-end.
Shrinking the workforce is one of a number of cost reduction measures designed to save the UK's second largest telecoms provider some £800 million by 2029.
It currently employs 16,000 people and redundancies will "equate to roughly 30 percent of the company's total people costs" over the next two years, VMO2 said on Tuesday.
The majority of the redundancies will come from "non-customer facing roles," it claimed. The Register asked the telco to be more specific on the headcount cuts and the type of roles are at risk.
A spokesperson refused to provide a number, saying those plans are being finalized and consultations are due to take place. Logically, if more senior head office positions are hit then fewer people will need to leave to meet the £300 million ($402 million) target figure.
The Reg understands thousands of roles are at risk.
Another lever management at VMO2 is pulling involves halting investment in any further fiber network upgrades beyond existing commitments. Instead, a low-cost upgrade will be deployed to its hybrid fiber-coaxial (HFC) network (i.e. cable customers) to bring 2 Gbps services to the majority of the network in 2027.
Full fiber merger pledgesVMO2 was formed in 2021 by the merger of broadband and TV biz Virgin Media and mobile operator O2. The pair pledged to spend £10 billion (c $13.4 billion) on beefing up 5G coverage and laying gigabit "full-fiber" fixed-line networks.
Both Virgin Media and O2 talked up those connectivity commitments when justifying the case to combine their businesses, saying they needed greater economies of scale to compete with their rivals.
However, business has not gone as planned, with revenue falling over several successive quarters. The latest figures for Q3 ended September 30 show total revenue down 5.7 percent year-on-year to £2.43 billion. The networks giant blamed consumer fixed market pressures, among other things.
VM02 forecasts a revenue decline of 3 to 5 percent for the whole of 2026.
In the consumer division, VMO2 talked of more clearly defining the role for each of its brands. O2 Business is planning to streamline the product portfolio.
Tom Oughton, senior analyst at Megabuyte, said the changes were expected.
"VMO2 has undergone a strategic review as it looks to address rising leverage, with today confirming a transformation program to ultimately bring c. £800 million ($1bn) in cost savings ahead of its first debt expiry in 2029. The routes to achieve this are the most obvious ones: reduce capex on full fiber and focus on HFC upgrade (lower cost, quick payback) and reduce operational costs by cutting headcount," Oughton said.
"The obvious question is how a significant transformation will impact the business. VMO2 focuses on the benefits, but the business is declining 3-5 percent organically with 16-17k staff, so reducing headcount by 25 percent or more could stretch operations, worsen customer support, and cause broader disruption."
Paolo Pescatore, founder and analyst at PP Foresight didn't mince his words either: "Virgin Media O2 is in serious trouble, with customer losses and declining service revenue and earnings exposing the scale of the challenge."
"The business has lost its strategic direction, and continued customer departures suggest it is struggling to rebuild trust while rivals give people compelling reasons to switch. Cutting costs can buy time, but it cannot repair those relationships or compensate for weak execution," Pescatore added.
"Management must show customers why they should join and stay; otherwise, it risks shrinking the business without fixing what is driving people away. Shareholders need to take a hard look at these businesses and make even tougher choices about their future."
Noel McClean, deputy general secretary of Prospect union, said the job cuts "will be devastating for those directly affected and will of course have an impact on colleagues remaining in the business.
"Industry conditions are tough but this level of cuts will drastically alter the way VMO2 works." ®
MotoAmerica Quad Lock Superbike 2026 In Review:
Championships Are Won By Preparing For The Bad Days
By Michael Gougis
First Person/Opinion:
If you want to know why Bobby Fong is the 2026 MotoAmerica Quad Lock Superbike Champion (and only one mistake away from being a two-time and defending Champion), look at the lap charts from Qualifying One at Road America in 2025.
The track was soaked. There was no rain in the forecast for the rest of the weekend. No one went out for the session, really. Nothing to gain, a lot to lose. Sean Dylan Kelly did a trio of slow laps. A couple of backmarkers went out to play in the rain.
No one, that is, except for Fong. He pounded in 10 laps around the long 4.0-mile circuit. And when I asked him why he did it, he said he wanted to learn the bike, learn how to work with his crew under those conditions, and to learn to ride better in the rain.
Go back a few months earlier, and I am interviewing Fong about his new ride with Attack Performance Progressive Yamaha. We're talking about his training routine, and he says that the majority of his training is simply riding motocross. Fong wanted experience in seeking the edge of traction, how to find it and manage what the bike is doing under those circumstances and to extract the most speed he could find in that instance. And on a motocross track, Fong says, the surface changes every lap, every corner. It's an intense learning curve, like a summer college class where a full semester's worth of work is crammed into five weeks.
Championships, they say, are won on the bad days. And the days when everything is perfect are far outnumbered by the days when they are not. There were other riders who were as fast as Fong, and sometimes a lot faster. But when the track surface got sketchy - wet, dry or damp - most of the time Fong simply did a better job of getting the most out of those conditions.
Attack played its part as well. Helpful to winning in MotoAmerica competition is: A) a worldwide network of people racing the bike who are sharing information and B) knowledge of how to get the spec Dunlops to work on your bike. Yamaha races the R1 around the globe, so there's a wealth of data to tap into. Just to illustrate, a few years ago, I was driving across Germany with a technician who provides trackside support for the alpha Racing BMWs that are raced worldwide. He said, one of the things that we offer the teams who buy our bikes is this: If you have a question, there's someone else running one of these bikes who has had the same question. Call us and we'll give you an answer. Yamaha's racing DNA pays off in ways that go largely unseen to the casual spectator; there's a wealth of data available for anyone racing an R1 because many, many teams race this bike.
Attack's Richard Stanboli probably knows the current YZF-R1 as well as any tuner/builder on the planet. The platform's age means there are few secrets left. And Attack has loads of experience in optimizing it for a wide range of track conditions on the spec Dunlop tires. Again, maybe not the fastest machine under ideal conditions, but Attack's ability to get it as good as it can get on race day when things aren't ideal is a huge advantage. The fact that JD Beach and Mathew Scholtz also won on Attack-built R1s is testament to that; each had a good machine performing at its best.
Sean Dylan Kelly (40) at Road America in 2026. Photo by Brian J. Nelson.
The biggest surprise, to me, was watching Kelly lead the Championship right up to the final race. There's no questioning Kelly's skills. You don't come back from years in the Moto2 World Championship slower. But it was his first season with the OrangeCat team, and it was the team's first full season with its Superbike-spec alpha Racing BMW M 1000 RR. While it had that bank of alpha Racing knowledge to tap into, OrangeCat still had to execute trackside. And while the team was experienced, having won the last Stock 1000 title easily with the M 1000 RR, stepping up to Superbike is always a challenge. The most competitive class always poses questions that you learn to answer over years of trying, failing and succeeding.
So as skilled as the OrangeCat crew is, I expected Kelly to win races, but I didn't expect to see him fighting for the title. It bodes well for Kelly in 2027, when he will be back with bike-specific track knowledge for each of MotoAmerica's circuits. And the well-funded OrangeCat team's determination to win is unquestionable. Only one rider beat them to the title in 2026. That was Fong. So OrangeCat simply went out and hired him for 2027.
PJ Jacobsen (15) at The Ridge. Photo by Brian J. Nelson.
If OrangeCat and Kelly overachieved, in my view, in 2026, the biggest underachievers this season were the riders on the new Ducati V4 R racebikes. On paper, 2026 should have seen the Panigales run away with everything. You don't just go buy one of these bikes. You apply for the privilege of racing one. Ducati vets your team to determine whether they think you can win before they'll sell you one, no matter how much money you have. The bikes come from a factory that builds nothing but racebikes, and they are as trick as the rule book allows.
Again, a few years back, I was interviewing the crew chief of a team that was racing Panigales in the Endurance World Championship series. The team bought machines straight from Ducati's racebike factory and then modified them for endurance racing. One of the modifications they had to make was to swap out the forks. The Superbike-spec Ohlins forks were so precisely tuned for flex and so lightweight that the endurance team didn't trust them not to crumple in a crash. In their mind, the Ohlins forks that came on the full Superbikes were like a racer's helmet: Good for one crash. The point is this: A factory-built Panigale is one finely-honed racebike, built right to the edge of the rules and refined to get every last bit of performance from the machines.
And there's no shortage of information on how to set up one of these bikes, either. Ducati won every Superbike World Championship race this year and took every pole and the vast majority of the podiums, and did so with seven different riders. It's clear that the bikes are fast and there's knowledge available about how to get the most out of them.
Now, to be fair, there were injuries and other setbacks that were not the fault of those riders at all. Cameron Beaubier's fall at Road America should not have sidelined him for half a season. Benjamin Smith's scary crash at the same track was due to a part flying off of another machine and locking his front wheel at more than 180 miles an hour. There was no shortage of bad luck plaguing the Ducati contingent.
But there were a lot of crashes, and the Ducatis frequently seemed to bite their riders unexpectedly. My theory is this: MotoAmerica is the only series in the world that uses the spec Dunlop tire. Every other racing Ducati is on Pirelli or in a series that has open tire competition. And a modern racebike is optimized around its tires. Rahal, Warhorse HSBK - they were basically on their own in trying to get new-for-2026 Panigales to work on the spec Dunlops. Not only was the Panigale V4 R this year brand new, weather conspired to limit the pre-season testing for the teams learning the new bike. In contrast, Attack, Strack, OrangeCat were all racing on basic platforms they knew and had raced on with Dunlops in the past. They started ahead of the teams running the new Ducati and basically stayed ahead.
Richie Escalante (54) at The Ridge. Photo by Brian J. Nelson.
In a way, the Team Hammer squad was in the same position as the Ducati teams trying to sort out Dunlops on their own. The Suzuki GSX-R1000R is a known quantity, but it's the oldest bike in the field, and Suzuki isn't racing MotoGP, World Superbike, or even competitively at the factory level in high-profile National series like British Superbike. Even in Japan, the best-placed Suzuki rider in Superbike was barely in the top 10. Other than the evergreen Yoshimura SERT squad in the FIM Endurance World Championship, Team Hammer is the only team, really, making the GSX-R a better, faster racebike, and it is largely on its own. Respect for what it accomplishes in that situation, and Richie Escalante's podium at The Ridge was hard-earned and well deserved.
And respect to Fong. As a racer sets up a pass three turns before it happens, he built this title over the past several years, and he came back from a bitter disappointment last season to hold his nerve until the very end this year. He earned this one.
It will be interesting to see what the Ducati teams can do next year with a season of Dunlop experience on the new bikes and with a fully healthy Beaubier, 2026 Supersport Champion Darryn Binder and former Superbike Champion Josh Herrin aboard.
Roll on, 2027.
The post MotoAmerica Quad Lock Superbike 2026 In Review: Winning On The Bad Days appeared first on Roadracing World Magazine | Motorcycle Riding, Racing & Tech News.
• The return of the storied BMW Rennsport moniker.
• Limited to 1,000 units globally.
• 205 hp and 432 lb. unladen weight.
• Expected U.S. Market Launch, Q1 2027.
• MSRP: $38,395 plus destination.
Woodcliff Lake, NJ - BMW Motorrad USA is proud to announce the return of the Rennsport [Racing Sport] moniker to the BMW motorcycle lineup with the new limited edition 2027 M 1000 RS. The new M hyper naked bike carries with it a heritage of eighty-eight years of track capable production motorcycles, including significant historical customer racing models such as the 1939 BMW R 51 RS, the RS 255 Kompressor and the 1954 RS54.
Since the 1970's the RS moniker has also referred to ReiseSport [Touring Sport] with models such as the R 100 RS and current R 1300 RS, but now, the M 1000 RS Rennsport firmly brings the original meaning back into the hands of today's BMW enthusiasts.
"The BMW M 1000 RS makes a clear statement. RS stands for Rennsport - and that is exactly what you feel in every detail. With 205 hp, maximum precision, M Brake and a character all its own, it is BMW Motorrad's ultimate hyper naked bike. And because true exclusivity needs limits, we are building exactly 1,000 units. More performance. More character. It simply does not get more M than this," says Markus Flasch, CEO of BMW Motorrad.
The new 2027 BMW M 1000 RS Limited Edition. Photo courtesy BMW.
M 1000 RS HIGHLIGHTS
• M 1000 RR spec 205 hp motor.
• Full titanium exhaust.
• 45-tooth Rear Sprocket (vs M-R 47-tooth rear sprocket).
• Limited Edition 1 of 1,000 plate on triple clamp.
• Matte Carbon M Winglets with RS lettering.
• Dark-clear coated brushed aluminum fuel tank.
• Night Black Matte paint finish.
• Black rear subframe with M 1000 Rennsport lettering.
• Tall single seat configuration.
• M 1000 RR tail section with rear Matte Carbon Cover with RS lettering.
• High-grip M Alcantara seat with M embroidery.
• M Forged aluminum footrest system.
• M Carbon Wheels.
• M Brake
• Pirelli Diablo IV Corsa tires.
• M Track Package.
1,000 units. Maximum exclusivity.
The BMW M 1000 RS is limited to 1,000 units worldwide with a limited number coming to the U.S. Market arrival is expected here in Q1 2027 and ordering will be available through a pre-order program. Sign up on www.bmwmotorcycles.com to stay informed.
The BMW M 1000 RS brings the high-performance credentials of the BMW M 1000 RR to the hyper naked segment. Its unfaired design puts the technology on display, while dark finishes and powerful proportions give it a distinctive, assertive presence. The M rider footrest system from the M 1000 RR and a seat height 0.78 inches higher than the M 1000 R create a more forward-focused riding position, emphasizing the M 1000 RS's uncompromising sporting character.
The M full titanium exhaust system highlights the bike's focus on lightweight construction and performance. Matte carbon M winglets with RS lettering combine distinctive racing style with aerodynamic function, while the tail section from the BMW M 1000 RR further underscores its superbike roots. Matte carbon components, a brushed aluminum fuel tank, new graphics and an M seat with an Alcantara look complete its exclusive, dark aesthetic.
At the heart of the BMW M 1000 RS is the superbike engine from the BMW M 1000 RR, featuring updated components including an increased compression ratio, titanium connecting rods, forged pistons with a two-ring piston design. Paired with an unladen weight of just 432 lbs., the engine delivers an exceptional power-to-weight ratio and immediate response across riding situations. The result combines strong acceleration, precise control and hyper naked agility with the advanced engineering of a superbike.
The 2027 BMW M 1000 RS Limited Edition. Photo courtesy BMW.
- STANDARD EQUIPMENT
205 hp M 1000 RR spec motor, Full titanium exhaust, Chain drive with M Endurance Chain, Anti-hopping clutch, Gear Shift Assist Pro, M Quick-Action Throttle with 58 degrees of rotation, M Carbon Wheels, ABS Pro, Race ABS, Dynamic Brake Control, Upside down fork with adjustable rebound, compression and preload, Adjustable Steering stabilizer, DTC Dynamic Traction Control, MSR Dynamic Engine Brake Control, M Carbon Winglets, Tire Pressure Monitor, Radial front brake caliper with M Brake, Rear suspension strut with adjustable rebound, compression and preload, Dynamic Damping Control, Heated Grips, Electronic immobilizer, LED tail light, headlight and turn signals, M Lightweight Battery, USB port, Drop sensor, Cruise Control, Comfort turn indicator, Headlight Pro, Multi-Controller, 6.5″ Colored TFT screen, Passenger Seat Cover, On-board computer, Adjustable M clutch and M handbrake levers, M Rider's seat with rear seat cover (Rider's seat with Alcantara look, Pillion seat cover in matte carbon), Keyless Ride, Riding Modes Pro, Connectivity, Scope of carbon parts: M Carbon Winglets, M Carbon front & Rear Fenders, M Carbon left & right tank covers, M Carbon Chain Guard, M Carbon Airbox cover, M Carbon Pillion seat cover(all carbon surfaces in matte finish), Detachable license plate holder, M milled adjustable footrest system, Ready for GP shift pattern, Aluminum fuel tank, M GPS Laptrigger Activation Code, Hill Start Control Pro, Ultimate Care Break-In Service.
The 2027 BMW M 1000 RS Limited Edition. Photo courtesy BMW.
The 2027 BMW M 1000 RS Limited Edition. Photo courtesy BMW.
- OPTIONAL EQUIPMENT
Torque Optimized Calibration (dealer installed software) for improved mid-range torque response.
- ACCESSORIES
Ergonomics and Comfort
- Windscreen, low
- M Passenger Kit
- Handlebar Risers, black
- M Rider's Seat, low or standard height
- Tank Pad
M Performance Parts
- M Cover Kit
- M Data Logger including GPS lap trigger
- Auxiliary Stand Spool Mounts
- M Chain Tensioner with Auxiliary Stand Mounts
- M Engine Protectors
- M Axle Protectors
Navigation & Communication
- Navigation Preparation
- ConnectedRide Smartphone Cradle
- ConnectedRide Navigator
- TFT Display Protective Glass
Maintenance & Technology
- M Motorcycle Carpet
- Auxiliary Stands, front & rear
- Anti-Theft Alarm
Storage
- Black Tank Bag - 7-10 liters
- Black Tail Bag - 22 liters
- Passenger Seat Bag - 7-10 liters

BMW Group in the United States:
BMW Group began operations in the U.S. over 50 years ago. In addition to the sales, marketing, and distribution of BMW, MINI, Rolls-Royce, and BMW Motorrad vehicles, BMW Group's business in the U.S. spans 30 locations in 12 states, including BMW Group Financial Services, BMW Manufacturing, Designworks, BMW Technology Office USA, and BMW i Ventures. The company's U.S. plant in South Carolina is the largest single BMW production facility in the world and the global center of competence for BMW Sports Activity Vehicles. The BMW Group sales organization is represented by a nationwide network of 355 BMW retailers, 147 BMW motorcycle retailers, 105 MINI passenger car dealers, and 37 Rolls-Royce Motor Car dealers. Taken together, BMW Group's business activities in the U.S. provide and support more than 120,000 jobs and contribute more than $43.3 billion to the U.S. economy annually.
The post BMW Unveils the 2027 M 1000 RR Limited Edition appeared first on Roadracing World Magazine | Motorcycle Riding, Racing & Tech News.

If you need to watch a video of a loud, bleating goat automaton without any skin, here you go. The House of Automata shared this lovely early-20th-century goat automaton. It is currently under construction for restoration, so it does not have any skin or goat-like features at the moment. — Read the rest
The post This skinless 1910 goat automaton bleats like a real goat appeared first on Boing Boing.

The occult philosopher Robert Fludd created this image in the 17th century. It is called Black Square and quite literally represents nothing. The medium is not known for certain, but it is guessed to be an engraving.
View this post on Instagram
From the Public Domain Review:
Black square ("and so on to infinity…" inscribed in Latin around its edges) representing the nothingness that was prior to the universe, as pictured by the great occult philosopher Robert Fludd, who himself passed into infinity on this day in 1637.
The post Robert Fludd's "Black Square" is a 17th-century picture of nothing appeared first on Boing Boing.

This mesmerizing time lapse shows what 48 hours of heavy snow looks like. The camera starts off facing a lawn chair and a wooden stick pointing up from the ground. The snow eventually covers all of this and more.
What 48 hours of heavy snow looks like
byu/Main_Locksmith_2543 ininterestingasfuck
Watching the entire chair and even the camera lens become covered by snow was one thing. — Read the rest
The post Watch 48 hours of heavy snow bury a lawn chair, a car and the camera lens appeared first on Boing Boing.
Rawpixel.com/ShutterstockOver the next few months, most year nine pupils in England (aged 13-14) will choose their GCSE options. For those curious about the jobs behind the clean energy systems, heat pumps, electric cars and buses now appearing on their streets, the government's new vocational GCSEs will arrive just too late.
The Department for Education announced the new qualifications in September 2026. They will have nationally set content, with assessment and grading regulated by Ofqual, the exams regulator in England. The first are set to be available in some subjects from as early as September 2029.
By then, today's year nines will have sat their GCSEs and today's year eights will be halfway through their courses.
Nevertheless, a good deal of the groundwork for a clean energy career is already taught at GCSE. It is just spread across several subjects, so pupils and parents need to know where to look.
At the University of Salford's Centre for Sustainable Innovation, I lead research on electric cars that can send power back to buildings and the grid. It calls on electricians, engineers, programmers and people who understand energy markets, and the first steps towards each of those jobs are already in GCSE subject content.
Science and geographyThe most widely taught material sits in GCSE science, which almost every pupil takes, either as combined science or as separate physics, chemistry and biology. The combined science subject content asks pupils to describe energy sources, from fossil fuels to wind, the tides and the sun, compare how they are used, and tell renewable from non-renewable.
Pupils also learn why the national grid carries power at high voltage, how an appliance's power depends on current and voltage and what the earth wire does in a mains circuit. Much of the physics behind electric cars and heat pumps is there, though neither is named.
Geography, a popular option, adds the human side. Its subject content covers how societies obtain food, energy and water, with a detailed study of one of the three, including its sustainable use and management at different scales.
Design and technologyDesign and technology goes a step further. Its subject content requires pupils to understand how energy is generated and stored, so they can choose the right source to power a product, and to apply the basic principles of life cycle assessment, which tallies a product's environmental impact from raw materials to disposal.
Half the marks come from a project in which each pupil designs and makes a prototype. Exam boards set new contexts each year, and the Department for Education's suggested list includes "securing a sustainable future". Pupils write their own design brief in response, so where a context allows, a sustainable energy project is well within reach.
Electronics and engineeringGCSE electronics and engineering come closest to the hardware side, although few schools offer them. The electronics subject content teaches pupils to build circuits that sense, decide and act, to switch power with transistors and to program microcontrollers that respond to sensors. They also learn how diodes turn alternating current (AC) into direct current (DC) - a step needed to charge any battery from the power grid.
For their assessed project, pupils write their own brief and then build a system to meet it. A circuit that switches on a small load only when a light sensor detects sunshine would be a simple version of the diverters that send spare rooftop solar power to a hot water tank.
Students can learn to control systems in GCSE engineering.
Halfpoint/Shutterstock
Engineering pupils write programs that monitor and control systems. They also learn that designs must take account of energy production methods, maintenance and end of life, and they study engineering's impact on the environment.
Economics and businessThe money questions belong here. GCSE economics covers how supply and demand set prices, which is why some electricity tariffs now charge different rates at different times of day. It also covers externalities, the costs and benefits that fall on people outside a transaction, and the policies governments use to correct them. Economists use the same idea to analyse carbon taxes and clean energy subsidies.
GCSE business includes the impact of environmental considerations on firms, sustainability among them, along with cash-flow forecasting and business start-ups. A pupil who hopes to run an installation business one day would find much of the groundwork here.
Joining the dotsMuch clean energy work happens where people live. You cannot install a heat pump in Salford from anywhere else. The government's Clean Energy Jobs Plan expects employment in the sector to double to 860,000 by 2030, with plumbers, electricians and welders among its priority occupations.
Pupils need three things to join the dots: education, exposure and experience. The education is in place. Teachers need no new specification to use a charger or a heat pump as a worked example, and families can ask at options evenings whether coursework can take a clean energy theme. To help families prepare, I have built a free interactive online tool, the Sustainable Local Energy Skills Roadmap, which turns four questions into a personal map from GCSE topics to clean energy jobs.
Exposure means meeting the people who do these jobs. Talks from local installers and visits to university energy labs show pupils what the topics lead to. Experience is where the government's call for 200,000 extra work experience placements for pupils aged 14 to 16 comes in. Installers, network companies and bus operators are natural hosts for this.
Early suggestions for vocational GCSE subjects include digital tech as well as building and construction, both close to clean energy work. The government says local leaders, employers, schools and colleges will help choose the final list, with a consultation due later this year, so clean energy employers and colleges have good reason to respond.
For pupils now in year eight and above, the subjects are already on the timetable. The heat pumps and electric car chargers on their own street are a good place to start connecting them.
Vahid Vahidinasab would like to thank the funding for DriVe2X research and innovation project from the European Commission and the UKRI, with grant numbers 101056934 and 10055673, respectively.
Cop30 host sets out steps to break link with coal, oil and gas as it aims to move focus from negotiating to implementing
All countries can start to break their dependence on fossil fuels now without needing to wait for new technologies or economic changes, Brazil's government has said ahead of a global climate summit.
Brazil, the host of last year's UN climate conference, set out the world's first global "roadmap" for phasing out fossil fuels on Tuesday, showing that countries could boost their economies and pull people out of poverty while breaking the link with coal, oil and gas.
Continue reading...Exclusive: Garden Climate Assessment Tool will allow users to predict a plant's response to the projected climate in a given location
"In the 20th century, we could pretty much grow the same things with confidence, we could learn from previous generations," said Prof Sam Brockington, a curator at Cambridge University Botanic Garden.
"But now we're moving into climate scenarios that we just don't have the traditions to understand fully."
Continue reading...For the first time since 1937, no open permits to hunt ruffed grouse in Ohio have been issued due to falling numbers
Every November, hundreds of hunters descend on four designated wilderness areas in southern Ohio in pursuit of the elusive ruffed grouse.
But not this year.
Continue reading...

At the recent EMFCamp, I bumped in to Professor Kate Devlin who was talking a little about the AI Resist List. I think one of the useful things we can do is loudly complain about shitty uses of AI. Here's an example.
As mentioned in a previous post, I'm contacting companies which send malformed calendar invites. One of the organisation I contacted was the British online supermarket Ocado. They don't have an easy way to contact their technical team, and no obvious paths to raise a bug report, so I filled in their usual customer service feedback form. I included a clear statement of the problem and a detailed technical analysis of why I thought it was a problem with their service.
After sending in my message, I received this back from "Lulu" at Ocado:
Thanks for getting in touch and sharing this feedback about our calendar files.
You're entirely right—having a technical glitch pop up when you're trying to organise your delivery is a real pain, so I've made sure to pass your feedback on to our technical and webshop team.
We really appreciate you taking the time to help us improve our service. I hope you have a wonderful rest of your day
I mean, that's AI right? It starts with the classic tell of a sycophantic agreement.
Next up is the em-dash. Beloved by chatbots but almost unheard of in British English style guides.
Finally (and this is hard to quantify) it just doesn't feel like a customer service response. They're usually poorly spelled, have dodgy grammar, and ask a bunch of irrelevant questions. The message gave me no confidence that anyone had looked at the problem and I doubted it had been passed on to a technically competent human.
I immediately got a customer satisfaction survey. Usually I give top marks to humans because, even though Net Promoter Score is bullshit, I don't want customer service agents to suffer the indignity of a bad review. But in this survey I vented my spleen - giving crappy 1 ⭐ scores on every metric and complaining vociferously about how rude it is to use an AI to deal with a genuine complaint.
I was both surprised and delighted to receive this back a few days after my invective was sent:
I saw your survey comment about getting a robotic, AI-generated response previously, and I wanted to reach out personally to apologise. There is nothing more frustrating than pointing out a legitimate bug only to feel like you're talking to a brick wall. I promise you there is a real person typing this out!
Second, I wanted to send a massive thank you for the incredibly detailed diagnostic. It is rare to get feedback that tells us exactly what is broken and how to fix it.
I can assure you that this has been passed to our team to look into. Your feedback is going to make things much smoother for a lot of people, and we really appreciate you taking the extra time to help us get it right.
Thank you again, I do hope that this issue is resolved quickly. If there is anything else I can help with in the meantime, please don't hesitate to get in touch.
Perhaps I'm being touchingly naïve, but that feels a lot more human, right? The grammar isn't perfect, the sentences are a bit more rambly, and it appears to have actually engaged with the contents of my message.
I get that having a million irate customer service messages is a hassle. Using AI to triage them probably makes business sense, summarising the verbose ones could be useful for customer service agents, and totting up general themes is almost certainly a sensible way to gather metrics without overloading the workforce.
But using an AI to reply to a human is just disrespectful. It says the organisation values their customer's time at less than their own.
It's OK to call out shitty AI responses and insist that a human engages with you.
The UK's National Cyber Security Centre (NCSC) says AI-powered malware is "highly likely" to be a part of a future ransomware attack on Britain's critical national infrastructure between now and 2028.
The warning accompanies its annual review and a separate assessment exploring how AI could change cyber threats over the next two years.
The cyber arm of signals intelligence agency GCHQ did not provide much detail about what form this AI malware may take. However, it alluded to open source AI security tools that were initially intended to assist security researchers being repurposed for nefarious activity.
Attackers are already repurposing such tools. For example, researchers are increasingly seeing widespread abuse of ARTEX, an open source, Chinese-made AI red-teaming platform, in attacks on UK registrar 123 Reg, Microsoft 365, and others since its launch in July.
In its pessimistic scenario for 2028, the NCSC envisages "a significant increase in the frequency and sophistication of cyberattacks worldwide," with wider adoption of AI tools making attribution harder.
China "has likely played a role" in supplying those tools in this future, "enabling them to hide their activities within the noise," the report states.
It also considers a further threat: an autonomous attacker operating beyond its human operator's control.
The NCSC says it "is a realistic possibility such an entity could self-replicate and conduct cyberattacks at will, with no apparent motivation or restraint."
Such attacks would likely complicate geopolitical relationships, owing again to attribution difficulties, making it easier for countries that order attacks on others to deny their involvement.
The UK's plans for AI defenseAt a time when each Patch Tuesday seems to come loaded with record numbers of bugs, the NCSC acknowledges that AI is already changing vulnerability discovery and patching.
The body is also exploring how AI could help defenders keep pace. Project DROMOS began earlier this year with a four-week pilot analyzing government code repositories, before expanding into selected operational environments.
The Government Cyber Coordination Centre (GC3) launched the project with the AI Security Institute and other departments. The NCSC says it uncovered hundreds of potential security issues, including critical vulnerabilities.
The review also reaffirms plans for Cyber Shield, an agentic cyber defense system first teased in July.
Cyber Shield is expected to be delivered in stages over the next five years. It will involve the NCSC providing defensive agents deployed by participating organizations with data that will strengthen national resilience to ongoing cyber threats.
The NCSC will be running the scheme alongside the Department for Digital, Culture, Media and Sport (DCMS), but acknowledged that this kind of automated defensive activity currently sits outside many organizations' risk tolerances, meaning it has a role to play in building trust in the system before adoption can go mainstream.
As reported last week, the government has set aside £60 million to fund the initiative, and an additional £70 million to support a new National Exposure and Real-Time Vulnerability (NERV) Centre.
The NERV Centre's job will be to identify and maintain an ongoing understanding of vulnerabilities across CNI organizations and the wider economy.
Per last week's budget, £40 million will fund a National Agentic Security Unit responsible for responding to uncontrolled bots and other agentic mishaps, and £24 million was assigned to strengthen public sector defenses.
An undisclosed sum will also help establish a new NCSC-managed AI Cyber Security Centre in Manchester to develop AI security expertise across government.
In the meantime, NCSC CEO Dr Richard Horne is calling for organizations to "invest with urgency" in their security and resilience.
"For decades, we have collectively got away with doing tech on the cheap," he said today, announcing the annual report.
"We can be a nation where leaders of all organizations actively choose to invest with urgency... And where all organizations make security and resilience a fundamental requirement, not an afterthought.
"The collective result would be a source of enduring national strength for us all."
Horne said organizations should treat the NCSC's Cyber Essentials certification as the starting point, but scale up resilience efforts in line with organizational size, complexity, and risk.
The government also announced a 75 percent discount on Cyber Essentials certification for micro, small, and medium-sized businesses applying for their first certificate. ®