The most consequential sentence in American higher education this month is six words long, and takes two seconds to type.

“Log in and complete my quiz.”

Reporting by Dana Goldstein and Alan Blinder for the New York Times, published on 10 August 2026, established that this instruction now works. Not as a demonstration, not as a jailbreak, but as ordinary consumer software doing what it was built to do. An agentic browser takes the credentials, opens Canvas or Blackboard or Brightspace, watches the prerecorded lecture, sits the test, drafts the paper and posts into the discussion forum, where it converses with classmates who may themselves be agents. The student is elsewhere, and may be asleep.

Notice what is absent. There is no deception directed at the machine, no prompt engineering, no attempt to evade a filter. The Times tested the three tools most used by students, ChatGPT, Gemini and Grammarly, and found none refused a request to write a paper on a student's behalf. None of the companies prevents its agents logging into learning management systems. Educators have asked the AI firms to make agents identify themselves inside course platforms, and the firms have declined. Perplexity told the Times such a requirement “would put the student's privacy and security at risk”, a remarkable sentence from a company arguing its software should be permitted to impersonate an enrolled human inside an institution that will later certify that person as competent.

The exposure is not marginal. Analysis of federal IPEDS data for autumn 2024, published by the education market analyst Phil Hill in January 2026, found 26.5 per cent of American postsecondary students enrolled exclusively in distance education, rising to 40.5 per cent of graduate students, with roughly 54.8 per cent taking at least one online course. More than half of American college students took an online class last year, up from about a third in 2019. A quarter to a half of the sector's assessed output now passes through a channel in which the only evidence a human was present is a login session and a timestamp.

Six Words Are the Entire Attack Surface

What makes agentic cheating different is not sophistication. It is the collapse of the seam. Contract cheating required a transaction, a counterparty, a payment trail and a delivered file the student then submitted under their own name. Chatbot cheating required copying, pasting and, if the student was careful, some rewriting. Both left a boundary somewhere: a moment at which the student's work stopped and someone else's began.

The agent removes the seam. It does not hand the student a file. It operates the account. The submission originates from the student's session, at the student's IP address, at a plausible hour and pace, with the mouse movements and dwell times of a person reading. Reporting by Frank Landymore in Futurism on 13 August 2026 drew the consequence: professors cannot fall back on supervised examinations and oral presentations in a course where everything happens behind a screen. Digital Trends, covering the same Times reporting that day, stated what most institutions have declined to say aloud, which is that the question is no longer whether AI can help a student cheat on an assignment but whether a student can complete an entire degree without doing much of the learning.

Platforms can look for behavioural tells: time on task, session patterns, the rhythm of navigation. The Times reported no fail-safe way to block agentic cheating, and the reason is structural rather than technical. Any signature that distinguishes an agent from a diligent student is a signature the agent can be instructed to imitate, and the imitation costs the student nothing, because the student is not the one doing the waiting.

The Detector Was Never Going to Hold This Line

The institutional response, reported by Kathryn Palmer in Inside Higher Ed on 5 August 2026, is that universities have stopped pretending detection works.

Yale, Vanderbilt, Johns Hopkins and Indiana have adopted policies banning or discouraging reliance on AI detection software as the sole evidence of alleged cheating. At least a dozen institutions, including Northwestern, Georgetown and New York University, have switched off Turnitin's AI detection feature entirely. The updated Faculty AI Playbook at Indiana's Kelley School of Business tells staff that tools claiming to detect AI use “are highly unreliable” and advises them to design assignments encouraging process, reasoning and authentic engagement instead.

This is a retreat, and it is the correct one. The evidence against detection has accumulated for three years and it is not close.

Vanderbilt was early and unusually candid about the arithmetic. In August 2023 its Brightspace team disabled Turnitin's AI detector and published its reasoning. The university had submitted roughly 75,000 papers to Turnitin in 2022. At the 1 per cent false positive rate Turnitin advertised at launch, that implied around 750 papers wrongly flagged in one year at one institution. Vanderbilt also noted that Turnitin disclosed no detail about how its classifier reached a verdict, and that detectors disproportionately flag writing by non-native English speakers.

That second objection has the strongest empirical backing of anything in this field. Weixin Liang, Mert Yuksekgonul, Yining Mao, Eric Wu and James Zou of Stanford University published a study in Patterns in July 2023 running seven widely used GPT detectors over TOEFL essays by non-native English speakers and over essays by American eighth-graders. The detectors classified the native-speaker writing with near-perfect accuracy. They classified 61.3 per cent of the non-native essays as AI generated, all seven unanimously flagged 19.8 per cent of them, and at least one detector flagged 97.8 per cent. Detectors key on text perplexity, penalising writers with a narrower range of expression. When the researchers used ChatGPT to enrich the vocabulary of the same essays, the false positive rate fell to 11.6 per cent. The surest way to stop a detector accusing a foreign student of using AI was to have AI rewrite their essay.

Turnitin reviewed more than 200 million papers in the first year of its detection feature, roughly 11 per cent of which contained at least 20 per cent AI writing, and its guidance puts the sentence-level false positive rate at around 4 per cent, which is not a rounding error when a single flagged paragraph is what a misconduct panel is looking at. The detectors fail in the other direction too. Peter Scarfe, Kelly Watcham, Alasdair Clarke and Etienne Roesch at the University of Reading ran the most rigorous real-world test of the proposition, published in PLOS ONE in June 2024, inserting AI-generated submissions into live undergraduate psychology assessments without the markers' knowledge. Ninety-four per cent went undetected. The AI submissions attained grades on average half a grade boundary above those of the real students.

Jennifer Frederick, executive director of Yale's Poorvu Center for Teaching and Learning, and Alfred Guy, who directs its undergraduate writing and tutoring programmes, gave Inside Higher Ed the clearest statement of why Yale stepped back. Evidence shows, they wrote, that “humanizing” programs are very successful at reducing or eliminating the percentage of AI usage that is detected, and they wanted to avoid “the inevitable cat and mouse game created by AI detection tools”. Their objection was not that the arms race is unwinnable but that it is beside the point. “This becomes a technical exercise rather than a learning event.”

The cost of getting it wrong is not abstract. Haishan Yang, a third-year PhD student in health economics at the University of Minnesota, sat an eight-hour preliminary examination remotely in August 2024 and was expelled the following January after faculty graders concluded, partly from AI detection software and ChatGPT comparisons, that his essays were machine-written. Yang denied it and sued. A federal court dismissed his due process claim in October 2025 and the Minnesota Court of Appeals affirmed the expulsion in February 2026. Whatever the merits of that case, its shape is what matters: an unreliable instrument, a career-ending sanction, and a burden of proof no party can discharge.

The Only Numbers Anyone Has Are the Numbers of the Caught

Abandoning detection has a consequence universities have not absorbed. The sector's statistical picture of academic misconduct is a record of enforcement, not behaviour, and enforcement has just been switched off.

The best longitudinal data comes from the United Kingdom. Michael Goodier, reporting for the Guardian on 15 June 2025, obtained figures under the Freedom of Information Act from 131 of 155 universities. Proven cases of AI-related cheating reached almost 7,000 in 2023-24, equivalent to 5.1 per 1,000 students, up from 1.6 the year before, with partial figures suggesting about 7.5 the following year. Proven conventional plagiarism fell from 19 per 1,000 students to 15.2 over the same period. More than 27 per cent of responding universities did not yet record AI misuse as a separate category.

Now set that against what students say they do. The Higher Education Policy Institute's Student Generative AI Survey 2026, written by Rose Stephenson and Charlotte Armstrong and published on 12 March 2026 from a December 2025 survey of 1,054 full-time UK undergraduates, found 94 per cent using generative AI to help with assessed work, up from 88 per cent in 2025 and 53 per cent in 2024. Twelve per cent said they had included AI-generated text directly in assessed work, up from 8 per cent and then 3 per cent in the two preceding years.

Twelve per cent of undergraduates admitting to submitting machine-written text. Five proven cases per thousand students. The gap is roughly a factor of twenty-four, and it is the honest measure of how much the sector knows about its own assessments. Scarfe told the Guardian that those caught are the tip of the iceberg, and explained why in a sentence that ought to be pinned to every misconduct policy in the world. AI detection is unlike plagiarism, where you can confirm the copied text. Where you suspect AI, it is near impossible to prove.

The faculty view comes from a survey of 1,057 American college faculty conducted between 29 October and 26 November 2025 by Elon University's Imagining the Digital Future Center with the American Association of Colleges and Universities, released on 21 January 2026. Seventy-three per cent had personally dealt with academic integrity issues involving student use of generative AI, 78 per cent said cheating had increased on their campus, 95 per cent expected the technology to increase overreliance, 90 per cent expected diminished critical thinking, and 74 per cent believed it would negatively affect the value of a degree. The authors are explicit that the sample is not statistically generalisable. It does not need to be. Three-quarters of a large sample of the people who mark the work believe the qualification they confer is worth less than it was, for reasons they can name.

Contract Cheating Already Answered This and Nobody Wanted the Answer

There is a comforting story in which generative AI created this crisis in November 2022. It is false, and believing it is why the sector is improvising.

Philip Newton of Swansea University published a systematic review in Frontiers in Education on 30 August 2018 covering 71 samples from 65 studies going back to 1978, with 54,514 participants. The historic average of students self-reporting commercial contract cheating was 3.52 per cent. In studies conducted between 2014 and 2018 the figure was 15.7 per cent, which Newton extrapolated to roughly 31 million students globally. That was five years before ChatGPT. The market was mature, the transactions untraceable in practice, the detection rate negligible, because a purpose-written essay contains no copied text for a similarity checker to find.

Tracey Bretag, Rowena Harper and colleagues surveyed 14,086 students across eight Australian universities for a study in Studies in Higher Education, and found three factors reliably associated with outsourcing: dissatisfaction with the teaching and learning environment, a perception that there were lots of opportunities to cheat, and speaking a language other than English at home. None is a technology variable. Two are descriptions of institutional design.

England responded with criminal law. The Skills and Post-16 Education Act 2022 made it an offence to provide, arrange or advertise contract cheating services for financial gain to students at post-16 institutions and higher education providers in England. It was the right instinct and has proved near-impossible to enforce, the providers being numerous, offshore and hard to reach, and prosecution sitting with the Crown Prosecution Service rather than an education regulator with a reason to care.

So the boundary between a student's work and someone else's did not disappear in 2026. It disappeared, for a meaningful minority, during the 2010s, and universities carried on treating the unsupervised written artefact as evidence of learning because the alternative was expensive. Generative AI removed the price, the counterparty, the delivery lag and the risk, converting a minority behaviour into a default one. That is not a change of kind but a change of magnitude large enough that the old assumption stops functioning.

Tricia Bertram Gallant, who directs the academic integrity office and testing centre at UC San Diego and co-authored The Opposite of Cheating, said exactly this to Inside Higher Ed. Over twenty-odd years, she noted, changes from internet-supported plagiarism to the contract cheating industry and now AI have slowly degraded the validity of twentieth-century assessments. Yet higher education has not changed. “We're still relying on the unsupervised written word as evidence of learning. The real trick is acknowledging that that doesn't work anymore.”

Students Are Not Confused About the Rules, They Are Renegotiating Them

One paper in this literature circulates as a finding that ChatGPT lowers the barrier to contract cheating while making detection harder. That is a fair summary of the field but not of the paper, and precision matters here. ArXiv 2606.09845, submitted on 27 April 2026 by Belle Li, Lily Tan, Wei Zakharov, Qiang Qiu and Colby Ben Acton, is titled “Tutor, Not Solver: Designing a Guardrailed AI Assistant for Learning in Higher Education: A Design Case of PeteChat”. It is not a prevalence study. It documents an AI tutor deployed at Purdue University on a locally hosted Llama-3 model with retrieval-augmented generation grounded in course materials, from which the authors derive eight design principles for assessment-aware tutors, from homework guardrails to self-regulated learning support. Its interest here is that it inverts the detection approach, assuming students will use a model and constraining what the model does at the point of use rather than punishing them afterwards. That works only inside a system the institution controls, which an agentic consumer browser is not.

The paper that addresses the collapse of the boundary is arXiv 2605.29090, submitted on 27 May 2026 by Jiyoon Kim, Kentaro Toyama, Sangmi Kim and John M. Carroll, titled “'It's OK Because...': The Wild West of Student Rationalization of AI Use in Academic Writing”. Drawing on twenty semi-structured interviews together with students' AI chat logs, syllabuses and submitted assignments, the authors find at least five distinct sites at which AI use is conceptualised, running from the policy the instructor intended through to what students actually did. Between those five sites, meaning drifts.

Within that drift the authors catalogue more than twenty distinct rationalisations. That copying AI-generated text is victimless. That any AI text reflecting the student's own beliefs, or reading in their own style, is therefore their own writing. That extensive AI use means they are learning more than they otherwise would. The taxonomy includes justifications used to excuse conscious violations of stated course policy. Crucially, these rationalisations arise both ad hoc and post hoc and are not necessarily self-consistent: students construct them in the moment, reconstruct them afterwards, and do not need them to cohere. Modern AI, the authors conclude, presents a steep ethical slippery slope which students conceptually slide down, landing far outside their instructors' pedagogical goals.

This should worry universities more than the agent story, because it describes a condition no assessment format repairs. A student who believes machine-generated prose expressing their own view is their own writing has not broken a rule. They have sincerely adopted a different definition of authorship, and they will carry it into a graduate job where nobody will correct it either.

The Argument on Reddit Is the Argument Everywhere

The third paper cited in the brief was characterised as documenting teacher burnout and poor institutional support. It is something else, and the something else is more useful.

At arXiv 2605.17712, submitted on 18 May 2026 by Pelin Yüce, Xiangruo Dai, Rebecca Owens and Tuğrulcan Elmas, “ChatGPT vs Teachers vs Students: Large-Scale Analysis of Generative AI Discourse in Education Communities on Reddit” analyses 270,000 AI-related posts and comments across 26 education subreddits between November 2022 and April 2026. Topic modelling yields seventeen themes. The trajectory is a detection-and-evasion arms race hardening into a sustained enforcement regime, with constructive integration only beginning to challenge it from the middle of 2024.

Different constituencies worry about different things. K-12 teachers foreground cognitive dependency, academics focus on detection, students in professional programmes on career anxiety. The finding that matters most is about contact. Seventeen per cent of threads are cross-role, and one third of that cross-role contact occurs in the two adversarial themes, AI Detection and Misconduct Enforcement. Students initiate 68 per cent of mixed threads, but faculty produce most of the replies. Mixed threads contain two to three times more records than same-role threads and last two to four times longer. Sentiment correlates strongly negatively with engagement.

Strip out the platform and what remains is a description of an institutional relationship. The sustained centre of contact between the two parties to an education has become the dispute over whether one of them cheated. Not the subject. Not the feedback. The accusation and the defence.

Marc Watkins, a writing and composition lecturer who directs the AI Institute for Teachers at the University of Mississippi, put the resourcing question to Inside Higher Ed in terms nobody has costed. He could not fathom the cost to a single university, let alone most campuses, of scaling AI-resilience tactics such as proctored oral examinations, and it should not all fall on faculty. Kevin Yee, who directs the Faculty Center for Teaching and Learning at the University of Central Florida, described colleagues at wit's end, moving towards assignment redesign not because they think it better but because the alternative failed. “It's a difficult, delicate moment right now, and I'm not sure we have all the answers.”

The Blue Book Is a Confession Rather Than a Cure

Which brings us to the booklet. The Atlantic has published a piece titled “What Students Learn From Blue-Book Exams”, which this article could not read behind the paywall and therefore does not characterise beyond its title. The broader argument is easy to state anyway. Put a human in a room, take the devices, hand them lined paper and a fixed period, and whatever appears on the page was produced by that person. It is the only verification method available that does not depend on a classifier, a probability or an inference.

Institutions are moving. Princeton's faculty voted on 11 May 2026 to place proctors in examination rooms, ending a system of unsupervised examinations that had stood since 1893, after a dean's letter reported that significant numbers of professors and students perceived cheating on in-class examinations to be widespread. The Daily Princetonian's survey of the Class of 2026 found 24.8 per cent of respondents admitting to cheating in violation of the honour code, and, asked about using a chatbot on work that banned it, 27 per cent of arts and sciences students and 46 per cent of engineers said they had. The University of Chicago Law School has banned phones, tablets and laptops in first-year classes as part of its AI strategy. Bertram Gallant put the logic best: “We cannot be giving unsupervised assessments to students expecting them to resist AI. They're not going to be learning, we're going to be spending our time trying to catch them cheating and they're going to graduate and realize they just wasted four years.”

The comparison usually offered here is the calculator, and it is worth saying where it breaks. Maths education absorbed the calculator by conceding the mechanical step and moving assessment upward, towards modelling and interpretation. The tool automated a component and left the task recognisable. An agent that logs into Canvas and completes the course does not automate a component. It automates the student. There is no upward move available, because there is no residual layer above being the person enrolled.

But the blue book is a confession, not a cure, and its costs land on exactly the students who can least afford them.

Amanda Sturgill, an associate professor of journalism at Elon University, set out the practical objections for the Center for Engaged Learning in August 2025. Handwriting speed varies enormously and is not evenly taught; there is, as she puts it, a public-private divide in handwriting instruction, so a timed handwritten assessment quietly advantages students whose schools drilled it. Post-pandemic cohorts compose on keyboards. Hands cramp, and cramping degrades output and legibility, neither of which bears on whether the student understands the material. For students with dysgraphia, motor impairments or writing-related accommodations, an unadapted handwritten examination measures penmanship and stamina rather than knowledge, and the accommodations that fix it, extra time, a scribe, a locked-down laptop, are exactly what under-resourced institutions ration.

The technologists' alternative has the same problem in a different place. Armando Fox of UC Berkeley and Craig Zilles of the University of Illinois at Urbana-Champaign argued in Inside Higher Ed on 19 March 2026 that computer-based testing facilities beat blue books, allowing algorithmically generated examination variants, richer question formats, self-scheduling across multi-day windows and frequent low-stakes assessment. Illinois ran more than 130,000 examinations through its facilities in autumn 2025. It is genuinely better than handwriting, and it requires a proctored building, a capital expenditure most institutions and effectively all fully online programmes do not have.

Remote proctoring, the option that appears to square the circle, has the worst record. Deborah Yoder-Himes and colleagues at the University of Louisville published a study in Frontiers in Education in September 2022 examining automated proctoring across 357 students in four STEM courses. The software detected Black students' faces 79 per cent of the time against 92 per cent for white students. Students with the darkest skin tones spent 7.64 per cent of assessment time flagged against 1.56 per cent for lighter tones, and received 6.07 flags per assessment against 1.19 for the lightest group. Women with the darkest skin tones were 4.36 times more likely to be flagged than women with medium tones. Monika Blue Kwapisz, Yoav Ackerman, Jennifer Nguyen and Prashanth Rajivan documented in a November 2025 preprint how students with disability accommodations experience these systems, describing anxiety about the interaction between surveillance and their disability, fear of being misread as cheating, and the cognitive load that fear imposes during the examination.

Line the three options up and the pattern is impossible to miss. Handwritten examinations disadvantage students with motor and writing disabilities and those schooled without handwriting instruction. Proctored computer facilities require capital poorer institutions lack and online programmes do not have at all. Algorithmic proctoring misidentifies dark-skinned students and penalises disabled ones. Every available method of proving a human did the work reallocates the burden onto the students already carrying most of it, the same population Liang and colleagues found the detectors falsely accusing, and the same population Bretag and colleagues found most likely to be flagged for outsourcing.

Signalling Was Always Doing More Work Than the Learning

Underneath all this sits an economic question higher education has spent fifty years avoiding, and the agent has forced.

Michael Spence's 1973 paper in the Quarterly Journal of Economics established the framework. Education functions in the labour market partly as a signal: employers cannot observe productivity directly, so they pay for a credential that is cheaper to obtain for able and conscientious candidates than for others. The signal works because it is costly, and it works whether or not anything was learned in the acquiring. Bryan Caplan's 2018 book The Case Against Education pushed the claim to its limit, arguing that something like 80 per cent of the earnings premium reflects signalling rather than skill, an estimate other economists dispute vigorously and nobody has settled.

The dispute does not need settling for the point to bite. If the credential carries any signalling weight at all, an agent that can obtain it degrades the signal for every holder, including everyone who did the work honestly. This is a mechanical claim, not a moral one. A signal cheap to counterfeit stops discriminating, and once employers know it is cheap to counterfeit they discount it uniformly, because they cannot tell which holders did. The honest graduate of an online programme in 2026 pays full price for an asset being devalued by people they will never meet.

That is what the Elon and AAC&U figure measures when 74 per cent of faculty say generative AI will negatively affect the value of a degree, and it lands hardest on the segment of the market built as a widening-participation instrument. Online degrees exist, in the story the sector tells about itself, to reach the working adult, the carer, the person in a town without a campus. Digital Trends made the uncomfortable observation: for colleges that have spent years arguing online education makes higher education more accessible, an agent that can attend the class instead of the person paying for it is a particularly awkward problem to have.

The mixed institutional signalling makes it worse. Futurism noted that many universities hold partnerships with AI companies while policing AI use, leaving students to reconcile the contradiction. The California State University system signed a multimillion-dollar agreement with OpenAI to put ChatGPT Edu in front of hundreds of thousands of students and staff. Carol Sewell, an instructor in that system, told the Times how that feels from a classroom: “I'm not getting any guidance on how to dissuade their use. I'm getting opportunities to learn more about using it.”

There is a workable model in plain sight, ignored for a century because it is unflattering. Professional licensure separates the two functions a degree smashes together. Nursing candidates in the United States take the NCLEX under supervision at dedicated testing centres, more than 328,000 of them in 2025 for the registered nurse examination alone. Medicine has the USMLE, law the bar. In every case the coursework is where learning happens, and a separate proctored instrument certifies competence. Nobody worries that a nursing student used AI on a formative assignment, because the assignment is not the credential. The credential is the day in the room.

Higher education runs one artefact for both purposes: the essay is simultaneously the pedagogical exercise and the evidence. That worked for as long as producing an essay required knowing something. The agent has broken the coupling, and the options are to rebuild it by force, at enormous cost and with the distributional consequences described above, or to separate the functions as every licensed profession did decades ago.

The Verification Problem Is Not the Serious Problem

Return to the six words.

The student who types them is not, in most cases, a fraud in the way an essay mill customer was. They are, per the rationalisation study, someone with an available justification: that they are learning more this way, that the assignment was busywork, that the output reflects what they think anyway. Per the Reddit analysis, their main sustained contact with the institution has become a dispute about enforcement. Per the HEPI figures, they are in a cohort where 94 per cent use the tool for assessed work and only 36 per cent feel their institution encourages them to. They are behaving rationally inside a system that has stopped telling them a coherent story about what the work is for.

Kathryn Kysar, a community college instructor who has taught online for fifteen years, told the Times that those who have taught a long time usually recognise the cheaply written AI stuff within thirty seconds. She is almost certainly right, and it does not help her, because recognising it and proving it are separated by an evidentiary chasm no detector spans. Jason Gibson, a history professor at Alcorn State University, buried the word Madagascar in white text inside a midterm prompt in the summer of 2026 and reported that thirty-two of his thirty-five students across two classes failed part of the examination, having pasted the prompt into a chatbot and submitted the result unread. What unsettled him was not the cheating. It was that nobody noticed sentences about Madagascar floating sideways through the afternoon in an essay on the Industrial Revolution.

Gibson caught his students because they were careless. The agentic student will not be careless, because the agent will not paste the prompt and will not leave the word in. The trap worked once.

So the answer to what a degree is worth when a machine could have earned it is not a number, and proctoring alone does not recover it. Making a person sit in a room and write by hand verifies that a human produced marks on paper under time pressure. It verifies nothing about the fourteen weeks before, and it produces a credential whose accessibility depends on how steady your hands are and how your school taught cursive. It is a verification of presence dressed as a verification of learning, adopted because it is the last mechanism that cannot be spoofed.

The harder thing is what Bertram Gallant said out loud and almost nobody has repeated. If a student can complete a degree without doing the learning, the injury is not principally to the institution's reputation or the employer's hiring accuracy. It is to the student, who paid the money, gave the years, got the paper. The four wasted years are the loss no invigilator prevents and no detector catches, and the only person positioned to notice is the one who has already spent everything avoiding it.

References

1. Dana Goldstein and Alan Blinder, “AI Agents Are Taking Entire Online Courses for Cheating Students,” The New York Times, 10 August 2026 (syndicated by the San Francisco Examiner, 12 August 2026). https://www.sfexaminer.com/ai-agents-are-taking-entire-online-courses-for-cheating-students/article_9d4ef5ca-317f-54a3-937a-56bfcc9a0272.html
2. Frank Landymore, “College Kids Are Using AI Agents to 'Take' Entire Online Classes for Them by Logging Directly Into the Course Management Software,” Futurism, 13 August 2026. https://futurism.com/artificial-intelligence/college-kids-ai-agents-cheat-online-courses
3. Varun Mirchandani, “AI agents are sitting through students' online courses, and colleges are struggling to stop them,” Digital Trends, 13 August 2026. https://www.digitaltrends.com/cool-tech/ai-agents-taking-online-courses-student-cheating/
4. Kathryn Palmer, “AI Detectors Are Out, New Assessments Are In,” Inside Higher Ed, 5 August 2026. https://www.insidehighered.com/news/tech-innovation/artificial-intelligence/2026/08/05/ai-detectors-are-out-new-approaches-are
5. Lee Rainie and C. Edward Watson, “The AI Challenge: A National Survey of College Faculty,” Imagining the Digital Future Center, Elon University, and the American Association of Colleges and Universities, 21 January 2026. https://imaginingthedigitalfuture.org/wp-content/uploads/2026/01/Elon-AACU-faculty-AI-survey-full-report-1-21-26.pdf
6. Vanderbilt University Brightspace team, “Guidance on AI Detection and Why We're Disabling Turnitin's AI Detector,” Vanderbilt University, 16 August 2023. https://www.vanderbilt.edu/brightspace/2023/08/16/guidance-on-ai-detection-and-why-were-disabling-turnitins-ai-detector/
7. Weixin Liang, Mert Yuksekgonul, Yining Mao, Eric Wu and James Zou, “GPT detectors are biased against non-native English writers,” Patterns, 10 July 2023. https://www.cell.com/patterns/fulltext/S2666-3899(23)00130-7
8. Peter Scarfe, Kelly Watcham, Alasdair Clarke and Etienne Roesch, “A real-world test of artificial intelligence infiltration of a university examinations system: A 'Turing Test' case study,” PLOS ONE, 26 June 2024. https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0305354
9. Michael Goodier, “Revealed: Thousands of UK university students caught cheating using AI,” The Guardian, 15 June 2025. https://www.theguardian.com/education/2025/jun/15/thousands-of-uk-university-students-caught-cheating-using-ai-artificial-intelligence-survey
10. Rose Stephenson and Charlotte Armstrong, “Student Generative Artificial Intelligence Survey 2026,” HEPI Report 199, Higher Education Policy Institute and Kortext, 12 March 2026. https://www.hepi.ac.uk/wp-content/uploads/2026/03/HEPI-Report-199-Gen-AI-Survey-2026.pdf
11. Philip M. Newton, “How Common Is Commercial Contract Cheating in Higher Education and Is It Increasing? A Systematic Review,” Frontiers in Education, 30 August 2018. https://www.frontiersin.org/journals/education/articles/10.3389/feduc.2018.00067/full
12. Tracey Bretag, Rowena Harper, Michael Burton, Cath Ellis, Philip Newton, Pearl Rozenberg, Sonia Saddiqui and Karen van Haeringen, “Contract cheating: a survey of Australian university students,” Studies in Higher Education, 2019. https://www.tandfonline.com/doi/full/10.1080/03075079.2018.1462788
13. Deborah R. Yoder-Himes, Alina Asif, Kaelin Kinney, Tiffany J. Brandt, Rhiannon E. Cecil, Paul R. Himes, Cara Cashon, Rosalie M. P. Hopp and Edna Ross, “Racial, skin tone, and sex disparities in automated proctoring software,” Frontiers in Education, 20 September 2022. https://www.frontiersin.org/journals/education/articles/10.3389/feduc.2022.881449/full
14. Monika Blue Kwapisz, Yoav Ackerman, Jennifer Nguyen and Prashanth Rajivan, “Surveillance and Disability in Online Proctored Exams: Student Perspectives and Design Implications,” arXiv:2511.10826, 13 November 2025. https://arxiv.org/abs/2511.10826
15. Elizabeth Shockman, “'A death penalty': Ph.D. student says U of M expelled him over unfair AI allegation,” MPR News, 17 January 2025. https://www.mprnews.org/story/2025/01/17/phd-student-says-university-of-minnesota-expelled-him-over-ai-allegation
16. Joe Wilkins, “Princeton in Shambles Over AI Cheating,” Futurism, 17 May 2026. https://futurism.com/future-society/princeton-shambles-ai-cheating
17. Frank Landymore, “Professor Hides White Font in Midterm, Catches Students Using AI in the Stupidest Way Possible,” Futurism, 25 July 2026. https://futurism.com/future-society/professor-hides-white-font-ai-cheating
18. Phil Hill, “Fall 2024 IPEDS Data: Profile of US Higher Ed Online Education,” On EdTech, 6 January 2026. https://onedtech.philhillaa.com/p/fall-2024-ipeds-data-profile-of-us-higher-ed-online-education
19. Michael Spence, “Job Market Signaling,” The Quarterly Journal of Economics, August 1973. https://academic.oup.com/qje/article-abstract/87/3/355/1909091
20. UK Parliament, “Skills and Post-16 Education Act 2022, sections 34 to 36,” legislation.gov.uk, 28 April 2022. https://www.legislation.gov.uk/ukpga/2022/21/notes/division/9/index.htm
21. Armando Fox and Craig Zilles, “Blue Books Are Not the Answer to AI,” Inside Higher Ed, 19 March 2026. https://www.insidehighered.com/opinion/views/2026/03/19/blue-books-are-not-answer-ai-opinion
22. Amanda Sturgill, “Blue Books and In-Class Writing Are Not a Panacea,” Center for Engaged Learning, Elon University, 19 August 2025. https://www.centerforengagedlearning.org/blue-books-and-in-class-writing-are-not-a-panacea/
23. Belle Li, Lily Tan, Wei Zakharov, Qiang Qiu and Colby Ben Acton, “Tutor, Not Solver: Designing a Guardrailed AI Assistant for Learning in Higher Education: A Design Case of PeteChat,” arXiv:2606.09845, 27 April 2026. https://arxiv.org/abs/2606.09845
24. Pelin Yüce, Xiangruo Dai, Rebecca Owens and Tuğrulcan Elmas, “ChatGPT vs Teachers vs Students: Large-Scale Analysis of Generative AI Discourse in Education Communities on Reddit,” arXiv:2605.17712, 18 May 2026. https://arxiv.org/abs/2605.17712
25. Jiyoon Kim, Kentaro Toyama, Sangmi Kim and John M. Carroll, “'It's OK Because...': The Wild West of Student Rationalization of AI Use in Academic Writing,” arXiv:2605.29090, 27 May 2026. https://arxiv.org/abs/2605.29090

---

Tim Green UK-based Systems Theorist & Independent Technology Writer

Tim explores the intersections of artificial intelligence, decentralised cognition, and posthuman ethics. His work, published at smarterarticles.co.uk, challenges dominant narratives of technological progress while proposing interdisciplinary frameworks for collective intelligence and digital stewardship.

His writing has been featured on Ground News and shared by independent researchers across both academic and technological communities.

ORCID: 0009-0002-0156-9795 Email: tim@smarterarticles.co.uk

Listen to the free weekly SmarterArticles Podcast

Chapter 1: The Song Before the Garden

There are nights when you know something difficult is waiting for you. Maybe you are sitting at the kitchen table after everyone else has gone to bed, looking at a message you do not want to answer. Maybe tomorrow holds a hard conversation, a medical appointment, a family decision, or news you wish you could avoid. In moments like that, it is easy to think only about the trouble ahead. That is why the moment Jesus sang before Gethsemane matters so much. It shows us something about Jesus that we can miss when we rush toward the garden and the cross.

Matthew and Mark both tell us that after the meal, Jesus and His disciples sang a hymn and then went out to the Mount of Olives. Most of us know what came next, so we move quickly toward His prayer in Gethsemane. But I keep coming back to the song. Jesus knew Judas had already chosen betrayal. He knew Peter would deny Him. He knew His friends would scatter. He knew soldiers were coming, and He knew where the night would end. When I think about how Jesus faced suffering without losing His trust in the Father, that hymn becomes one of the most revealing moments of the entire night.

Jesus did not sing because He misunderstood what was happening. He sang with full knowledge of what was coming. That is what makes the moment so powerful. He was not celebrating an easy day. He was not thanking God because every problem had disappeared. He was about to walk into one of the most painful nights any person could face, and He still made room for worship.

That changes the way I think about worship. Sometimes we treat worship like a response to good news. We praise God when the answer comes, when the burden lifts, when the doctor says things look better, when the relationship is repaired, or when the bank account finally gives us room to breathe. There is nothing wrong with thanking God in those moments. But Jesus shows us that worship can come before relief.

Imagine someone sitting in a parked car outside a hospital because they are not ready to walk through the doors. They know the appointment may bring an answer they do not want. Their hands are tight around the steering wheel. Their mind is already running through worst-case possibilities. Worship in that moment does not mean pretending there is nothing to fear. It means saying, quietly and honestly, “God, I do not know what I am about to hear, but You are still here.”

That is closer to what Jesus shows us.

He sang before the garden, not after it. He worshiped before Judas arrived, not after the danger passed. He turned His heart toward the Father before the night became darker.

There is another part of this scene that moves me even more. Jesus sang with men who were about to fail Him.

Peter was there. Within hours, Peter would deny that he even knew Jesus. The other disciples were there. They would run when the soldiers came. Jesus knew their weakness before they showed it, yet He still shared the table with them and sang beside them.

That tells us something about the love of Christ.

Jesus does not love people only after they become dependable. He does not wait for perfect courage, perfect faith, or perfect obedience before drawing near. He knew exactly what His disciples would do, and He still gave them His presence.

That matters when you are carrying regret. It matters when you remember the promise you made to God and failed to keep. It matters when you know fear got the better of you, when you did not speak up, when you lost patience, or when you feel embarrassed by your own weakness.

Jesus knew Peter would fail before Peter did.

And He still sang with him.

Maybe that is part of the missing lesson in this story. Before we see Jesus crushed by sorrow in Gethsemane, we see Him worship. Before we see His friends abandon Him, we see Him singing with them. Before the night exposes human weakness, Jesus gives us a picture of divine steadiness.

Worship did not erase the garden. It did not cancel the betrayal. It did not remove the cross.

But it showed where Jesus had placed His heart before He walked into them.

Chapter 2: Worship Before You Know the Outcome

The next morning can feel very different from the night before. You wake up, reach for your phone, and the same problem is still there. The message has not changed. The bill is still due. The person you love is still struggling. The prayer you whispered yesterday still seems unanswered. That is where the hymn Jesus sang becomes more than an interesting detail. It becomes a way to understand what worship is for.

Jesus did not sing because worship guaranteed that the hard thing would disappear. He sang knowing the hard thing was still ahead. Then He went to Gethsemane, where He became deeply distressed and told His disciples that His soul was overwhelmed with sorrow. The song did not make Him less human. It did not keep Him from feeling the weight of the night.

That matters because some people quietly assume that strong faith should make pain easier to hide. They think worship means smiling through fear or refusing to admit that something hurts. Jesus shows us something more honest. He could sing with His disciples and later fall to the ground in prayer. He could trust the Father and still ask for the cup to pass from Him.

Those things were not contradictions.

Worship was not Jesus pretending everything was fine. Worship was Jesus remembering who the Father was before the suffering reached its deepest point.

I think about the person who has become the dependable one in the family. Everyone calls them when something breaks, when money is short, when someone needs a ride, when an older parent needs help, or when a crisis shows up. They keep going because people count on them. But eventually there is a quiet moment when nobody is asking anything, and the truth comes out: “I am tired.”

That person does not need to fake strength before God.

Neither did Jesus.

In Gethsemane, Jesus prayed honestly. He asked for another way, then surrendered Himself to the Father’s will. His trust was not shallow because He felt sorrow. His trust was proven inside the sorrow.

That gives the hymn even more meaning.

Jesus had already turned His attention toward the Father before the pressure reached its peak. He had already filled the room with worship before the garden filled with anguish. He had already chosen trust before Judas appeared.

Maybe that is one reason worship matters so much when life is uncertain.

It reminds us of truth before fear gets the loudest voice.

It does not promise that tomorrow will be easy. It does not guarantee the answer we want. It does not make grief disappear or turn every painful situation around overnight. But it can steady the heart enough to say, “God, I still belong to You here.”

That is a different kind of praise.

It is easy to worship on resurrection morning.

It is harder to worship when you can still see Gethsemane in front of you.

Yet that is exactly what Jesus did.

The garden still came. The betrayal still came. The cross still came. But so did resurrection.

The hymn was not denial of the darkness. It was an act of trust before the darkness had finished speaking.

So if you are facing something you would rather avoid, you do not have to pretend it is small. You do not have to act fearless. You do not have to force yourself into a cheerful mood before God will listen.

You can tell Him the truth.

And you can worship Him there.

Not because you know exactly how the story will turn out, but because you know who He is while the story is still unfolding.

That may be the quiet lesson hidden in one short sentence before Jesus went to the Mount of Olives.

He knew what was coming.

He knew who would fail Him.

He knew how much it would cost.

And He still sang.

Maybe some of the deepest worship we will ever offer God will not come after the victory.

Maybe it will be the song we choose before we know the outcome.

Your friend, Douglas Vandergraph

Explore the complete Douglas Vandergraph Master Index: https://douglasvandergraph.com/douglas-vandergraph-master-index/

Watch Douglas Vandergraph’s faith-based videos on YouTube: https://www.youtube.com/@douglasvandergraph

most of us we are so numb we are so inflamed we don’t even know

why we keep coughing when we can’t stop inhaling toxins

why we have chronic migraines when we continue to drink diet coke thinking it’s healthy

why our joints hurt when we eat a load of sugar

why we can’t lose weight when we aren’t eating the right food for our unique designs

why we can’t sleep when our sheets reek of a chemical fragrance

why our hair falls when we aren’t eating enough protein

why we have an odor when we don’t practice hygiene

why our skin reacts to the sun when the sun is only bringing our problems to the surface

why we blast music on the beach when the sound of the waves is present

we are overstimulated over-entertained over-dependent on everything but ourselves

In Summary: * As my vision continues to recover from the treatment received yesterday afternoon, I've been very careful not to use my eyes too much. I've used heavy sunglasses every time I went outside, I've used my prescription eyeglasses whenever watching TV, working at the computer, or reading. The only surprising effect today has been an increase of annoying “floaters” in my vision.

I've done more radio and music listening today than usual, and less screen watching of any kind. That's probably good for me in ways other that resting my eyes, truth be told. Listening now to “Oldies” hits from 1966. Soon I'll tune in a radio pregame broadcast ahead of an MLB game, then I'll listen to the radio call of that Rangers vs White Sox game. And that'll end my “recovery” Wednesday.

Prayers, etc.: * I have a daily prayer regimen I try to follow throughout the day from early morning, as soon as I roll out of bed, until head hits pillow at night.

Health Metrics: * bw= 224.1 lbs. * bp= 153/90 (62)

Exercise: * morning stretches, balance exercises, kegel pelvic floor exercises, half squats, calf raises, wall push-ups, BP breathing exercises, pilates

Diet: * 05:05 – 1 little cupcake * 06:15 – 1 fresh apple * 06:45 – pizza * 12:00 – chicken tenders and gravy, mashed potatoes * 15:30 – 1 fresh tangerine

Activities, Chores, etc.: * 04:00 – listen to local news talk radio * 04:35 – bank accounts activity monitored . * 05:00 – read, write, pray, follow news reports from various sources, surf the socials, nap * 09:00 – listen to relaxing music * 12:00 to 13:00 – watch old game shows and eat lunch at home with Sylvia * 13:15 tuned into 107.5 The Fan well ahead of this afternoon's The Ride With JMV Show for even more Indiana sports talk. * 17:00 – listen to relaxing music * 18:00 – tuned into 105.3 The Fan well ahead of tonight's Rangers / White Sox game

Chess: * 10:15 – moved in all pending CC games

18 August 2026

Gendlin gave the name “felt sense” to the unclear, pre-verbal sense of “something”—the inner knowledge or awareness that has not been consciously thought or verbalized—as that “something” is experienced in the body. It is not the same as an emotion. This bodily felt “something” may be an awareness of a situation or an old hurt, or of something that is “coming”—perhaps an idea or insight. Crucial to the concept, as defined by Gendlin, is that it is unclear and vague, and it is always more than any attempt to express it verbally. Gendlin also described it as “sensing an implicit complexity, a wholistic sense of what one is working on”.

Agosto | ¿Por qué tenemos tanto miedo de sentir?

¿Por qué uno tiene que reprimir tanto sus emociones?

¿Por qué no podemos ser intensos? ¿Por qué sentimos que somos demasiado cuando sentimos mucho?

Si te gusta alguien, ¿por qué da tanto miedo decirlo?

¿Por qué me da miedo decirte que te extraño?

Que ya no recuerdo tu aroma, pero que quisiera volver a tenerte cerca para recordarlo. Que quisiera volver a ver esos ojos de cerca, con ese brillo tan particular que tienen cuando hablas de algo que te apasiona. Escucharte hablar de todo y de nada. Volver a sentir tus besos, tus manos. Quisiera simplemente estar existiendo a tu lado, sin hacer nada extraordinario, solamente estar.

Te extraño.

Y me duele admitirlo.

Pero lo más complejo de todo no es extrañarte. Lo más complejo es sentir que no puedo decírtelo.

Porque entonces aparecen todas las dudas.

Tal vez dejaste de sentir lo mismo de la noche a la mañana.

Tal vez te arrepentiste.

Tal vez prefieres que seamos amigos.

Tal vez ya no me piensas de la misma manera en la que yo te pienso.

Tal vez para ti aquello que pasó fue solamente un momento y para mí se convirtió en algo mucho más difícil de acomodar.

Tal vez…

Y ese es el problema.

Que no sé.

No sé qué piensas. No sé qué sientes. No sé si me extrañas. No sé si alguna vez aparezco en tus pensamientos de la misma forma en la que tú apareces en los míos.

Y quizá por eso me da miedo preguntarlo.

Porque mientras no sepa, todavía puedo imaginar cualquier respuesta.

Puedo pensar que tal vez sí me extrañas. Que tal vez también tienes ganas de verme. Que tal vez cuando escuchas alguna canción, recuerdas algo de nosotros. Que tal vez, en algún momento del día, también te preguntas qué estoy haciendo.

Pero si pregunto, existe la posibilidad de que la respuesta sea otra.

Y creo que eso es lo que realmente me asusta.

No extrañarte.

No sentir.

No ser intensa.

Me asusta descubrir que todo esto que todavía siento está ocurriendo solamente de mi lado.

Y qué extraño es eso de querer decirle a alguien “te extraño” y, al mismo tiempo, tener tanto miedo de descubrir que quizá tú ya no tienes ganas de escuchar esas palabras.

Sequestered in absentia, notes of wander beckoned notice like the blooded claim of an overtly familiar bite. Too tarnished to maintain an omniscience, her need began The Unraveling. Each tug. Every touch. All sensation became a ripened prayer to yearning's phantom dismemberment, and the obligatory recycling mandated when learning to walk on two broken feet.

© 2026 AnOublietteofThought. All rights reserved.

—

I’m working on a book of erotic poetry. I thought I'd pen some pieces to wander along beside it and wet the appetite. Those who know my writing side well know that I loathe creating titles for poetry. Titles for books are quite exciting, on the otherhand. I already know what this one will be.

hello guess what? I’m on the train again. This time brightly illuminated pale fields of wheat and pine tree forests are zipping by along with a handful of red houses haphazardly scattered like raisins in a müsli

It’s always nice to come home. I felt disoriented when I woke up this morning in the apartment. A disappointment settling inside as I learned where I was.

I think mornings can be harsh, because there’s always a brief moment when you have a clean slate, and as the state is loaded back into the brain, the disappointments and failures briefly forgotten, are reexperienced.

Of course there are good things loaded with the context too, like waking up for the first time next to someone you love etc

Life itself is indeed like a müsli, but sometimes there’s shit in it.

Il existe des regards sous lesquels nous nous corrigeons avant même d'avoir parlé.

Ils n'ont pas besoin d'être hostiles. Une hésitation suffit, un sourcil qui se lève, une présence qui recule à peine. Alors nous baissons la voix, nous rangeons un désir, nous choisissons une version de nous plus facile à recevoir. Nous apprenons à nous présenter comme on prépare une chambre pour un invité difficile, en retirant tout ce qui pourrait encombrer son regard. À la longue, ce regard n'a même plus besoin d'être là. Nous l'emportons avec nous. Nous rangeons la chambre pour personne.

Et puis il existe le regard heureux. Celui qui ne nous embellit pas. Il se réjouit simplement de nous trouver là, et il ne demande pas que nous soyons achevés pour prendre plaisir à notre présence. C'est ce qui le rend si convaincant. La bienveillance est une décision, et nous le sentons. La joie de nous voir n'en est pas une. On peut choisir d'être patient avec quelqu'un. On ne choisit pas d'être content qu'il existe.

Sous ce regard, quelque chose en nous cesse de rentrer le ventre. La voix que nous avions baissée reprend son volume. Ce qui riait trop fort, hésitait trop longtemps ou rêvait de travers retrouve le droit de circuler dans la pièce. Nous n'avons plus à déposer devant nous, comme une carte de visite, une version commode de ce que nous sommes.

Quelqu'un nous accueille.

Pas avec une grande déclaration. Pas nécessairement avec des promesses. L'accueil véritable fait peu de bruit. Il ressemble à une porte qui reste ouverte alors que nous pensions devoir repartir. À une chaise que personne ne retire au moment où nous nous asseyons. À un regard qui ne se détourne pas lorsque notre belle composition se défait un peu.

Être accueilli, c'est pouvoir apparaître sans passer d'abord par l'atelier des retouches. C'est montrer une joie trop vive, une peur mal coiffée, une idée encore bancale, et découvrir que rien ne s'effondre. L'autre est toujours là. Il ne nous confond pas avec notre maladresse du jour. Nos aveux ne deviendront pas, plus tard, des marteaux. Notre vulnérabilité n'est pas versée au dossier à charge.

Il accueille et il observe.

Observer quelqu'un est une forme rare de générosité. Cela demande de ne pas courir vers une conclusion, de ne pas expliquer à sa place ce qu'il ressent. Expliquer quelqu'un est si souvent une manière polie de le ranger. Celui qui observe regarde ce qui naît, ce qui hésite, ce qui revient. Il laisse la personne se révéler à son rythme, comme une couleur monte dans une toile encore humide.

Et surtout, il peut le faire avec gaieté.

L'accueil n'a pas besoin du visage sévère des salles d'attente. La gravité de celui qui écoute risque même de confirmer la crainte de celui qui parle : si l'autre prend cet air, c'est que la chose devait être grave. On peut recevoir quelqu'un avec un sourire, un verre posé sur la table, une plaisanterie qui rend l'air plus léger. On peut trouver charmant qu'un être humain ne tienne pas tout entier dans la première définition venue.

Cette gaieté n'efface pas la profondeur. Elle lui donne de l'oxygène. Elle dit : ce que tu es n'est pas une catastrophe. Je peux te regarder sans convoquer aussitôt les pompiers, les juges et les architectes. Tu peux être étrange ici. Tu peux être magnifique et parfois franchement pénible. Tu peux te tromper sans être expulsé du paysage.

Alors quelque chose se détend.

La personne cesse de consacrer toute son énergie à protéger son droit de rester. Elle n'a plus à séduire à chaque phrase ni à préparer sa défense avant même d'avoir ouvert la bouche. Ce qui se tenait caché peut enfin sortir, s'étirer, respirer, regarder autour de soi.

Et c'est seulement là que le changement commence.

Car on ne fait pas grandir quelqu'un en lui répétant qu'il devrait déjà être différent. On obtient parfois une obéissance, une imitation convaincante, une amélioration de façade. Mais la transformation profonde ne pousse pas sous la menace de perdre l'amour. Elle pousse lorsque l'être comprend qu'il peut se transformer sans devoir d'abord se renier.

Ce qui est accueilli n'est pas condamné à rester identique. C'est même l'inverse : on n'est libre de devenir autre que lorsqu'on a le droit de rester tel qu'on est. Ce qui est accueilli n'a plus besoin de s'agripper à sa forme actuelle pour survivre. Il peut bouger. Il peut lâcher une vieille défense sans craindre de disparaître avec elle, essayer, rater, recommencer.

Ce regard ne devient pas pour autant celui d'un sculpteur. Il ne corrige pas chaque angle, il ne polit pas chaque rugosité, et il ne signe pas ce qui pousse. Il offre une présence assez vaste pour que nous puissions nous y rencontrer, et assez libre pour que nous n'en devenions pas la propriété.

Accueillir ne signifie pas tout accepter. On peut poser une limite sans retirer son regard, refuser un geste sans effacer la personne, dire non sans faire de ce non un exil. Un non prononcé sans retrait suppose même que l'autre peut l'entendre sans s'effondrer. C'est encore une forme de confiance. L'accueil véritable ne consiste pas à applaudir tout ce qui apparaît. Il consiste à ne pas confondre l'amour avec une récompense distribuée aux versions bien élevées.

Peu à peu, ce regard reçu devient le nôtre. Nous l'emportons avec nous, comme nous avions emporté l'autre. Nous apprenons à nous offrir cette même hospitalité, nous cessons de nous jeter dehors à la première faute. Nous devenons pour nous ce lieu où rien n'est forcé de mentir pour avoir le droit de rester.

C'est sans doute cela, aimer quelqu'un.

Lui préparer un espace assez doux pour qu'il ose y être vrai, une chambre où il n'ait rien à ranger avant d'entrer. L'observer avec curiosité, parfois avec amusement, toujours avec respect. Ne pas lui demander de devenir meilleur avant de mériter notre présence.

A more typical win.

I won this game yesterday evening when my opponent playing White resigned the game after my 33rd move (33...Rxc3), giving me the win by default. Checkmate was clearly only two moves away at this point and there was nothing he could do to avoid it.

This game began on 27 July and ended yesterday, so he and I moved at a pretty good pace. Even though we were playing with a 2 days per move time control, we kept at it; sometimes taking the full 2 days, other times moving several times in a day.

The position of pieces at game's end is shown above, and our full move record below:

1. e4 a6 2. Nf3 Nc6 3. Be2 e5 4. a3 Nf6 5. Nc3 d5 6. h3 Be7 7. O-O O-O 8. d4 Nxd4 9. Nxd4 exd4 10. Qxd4 c6 11. e5 Ne4 12. Nxe4 dxe4 13. Qxd8 Rxd8 14. b4 b5 15. Be3 h6 16. f3 exf3 17. Bxf3 Bb7 18. c3 Rd3 19. Bd4 a5 20. Rfe1 Rc8 21. Bg4 Rd8 22. Re4 c5 23. bxc5 Bxe4 24. e6 fxe6 25. Bxe6+ Kh8 26. Bg4 Bg5 27. Be2 Be3+ 28. Kh2 Bf4+ 29. Kg1 Rd2 30. Kf1 Rd8xd4 31. Kf2 Rd4d3 32. Rg1 Bg3+ 33. Kf1 Rxc3

And the adventure continues.

Starcraft Co. 17- Open War

By the time the UIC formally acknowledged that it was at war, the distinction had become almost academic. Renegade Terran forces were no longer operating through isolated laboratories, hidden breeding farms and deniable research installations. They held territory. They fielded fleets. They controlled Zerg swarms large enough to assault major population centers, and intelligence recovered from Khamsin and Aiur suggested that every previous operation had been preparation for something larger. The discovery that finally pushed the Council into open mobilization came from a captured renegade communications node near the frontier. Analysts found repeated references to a device called a psi emitter, a machine capable of projecting a signal across interstellar distances and drawing Zerg toward it from systems far beyond any conventional sensor range. The renegades had not needed to breed every creature that attacked Kestral or New Carthage. They could summon wild swarms, concentrate them around a chosen target, and then use their developing control technology to point the resulting biological hurricane in a useful direction. The emitter had been located on Niflheim, an otherwise insignificant snowbound planet near the edge of UIC-controlled space. There were no major cities, almost no permanent population and little economic infrastructure beyond automated mineral extraction. The renegades had chosen it because none of those things mattered. Niflheim sat close enough to several known migration routes that the emitter could draw Zerg toward the core while remaining far from civilian interference. Reconnaissance already showed the consequences. Swarms were arriving continuously. Some contained only a few hundred organisms. Others numbered in the thousands. Mutalisks descended from orbit beside overlords carrying ground strains. Scourge appeared in clouds. Ultralisks wandered across the ice beneath them. The biological signatures surrounding the emitter increased every hour.

For once, the UIC reacted with something approaching its full strength. Fleets were pulled from multiple sectors. Army formations mobilized from core worlds. Hundreds of transports moved toward Niflheim. General Voss informed StarCraft Co. that Central Command intended to destroy the emitter with overwhelming force and requested that the company deploy as a supporting reserve. The commander read the operation plan and immediately disliked it. The primary landing zone was an open basin south of the emitter, a broad snowfield surrounded by low ridges. It offered room to land an enormous army quickly and almost nothing else. “You're putting the main force in low ground,” the commander said. Voss answered that the basin was the only area large enough to accommodate simultaneous landings by several divisions. “It also has no natural barriers, no defensible approaches and high ground on three sides.” Voss hesitated. “The Army believes massed firepower will compensate.” The commander stared at the map. “Against how many Zerg?” “Current estimate is uncertain.” “Then what exactly are they compensating for?” Voss sighed. “This plan is already approved.” The commander understood what that meant. Too many generals, too many commands and too many signatures had accumulated behind it for anyone to change course without admitting the obvious flaw. StarCraft Co. was ordered to remain in reserve and arrive after the main assault was underway.

The UIC army landed first. Thousands of marines and medics established bunker lines across the snow while siege tanks deployed behind them. Goliaths formed anti-air belts around the artillery. Battlecruisers hovered overhead. Wraiths and Valkyries screened the approaches. On a tactical display it looked formidable. In person, it looked exposed. The first Zerg attack came before half the supply depots were operational. Several dozen ultralisks formed the spearhead, charging directly across the open snow while artillery fired into them from maximum range. Behind them came an ocean of zerglings and hydralisks. The UIC tank lines inflicted staggering casualties, but there was nowhere to funnel the attackers and no terrain forcing them into manageable concentrations. The swarm simply spread wider. The ultralisks reached the bunkers. Some died within meters of them. Others hit at full speed. Entire defensive positions disappeared beneath kaiser blades and tons of armored flesh. Zerglings poured through the openings. Hydralisks stopped beyond the first line and fired continuously into the exposed infantry. UIC commanders committed reserve tanks early, then discovered that the reserves had nowhere useful to deploy because fleeing troops and advancing Zerg already filled the snowfield. What should have been layered defenses became one enormous firing line, and once sections of that line broke there was nothing behind them except more soldiers.

The air battle collapsed almost simultaneously. UIC battlecruisers moved forward to relieve the ground force and encountered clouds of scourge screened by devourers. Laser batteries destroyed hundreds, but hundreds more came behind them. Corrosive spores coated hulls and weakened armor. Scourge struck damaged sections in clusters. One battlecruiser exploded above the battlefield and fell burning into the snow. Another lost propulsion and was torn apart by successive impacts before evacuation could begin. Goliaths attempted to support the fleet and found guardians drifting beyond effective anti-air angles, launching biological projectiles into their formations while mutalisks swept through the gaps. Within hours, the UIC's carefully assembled assault force had become a disaster visible from orbit. StarCraft Co. arrived while it was still happening. The commander took one look at the tactical display and ignored the designated reserve landing zone. “Mountains,” he ordered.

StarCraft Co. transports redirected toward a range west of the battlefield, where narrow valleys and elevated plateaus provided exactly what the UIC deployment lacked. Dropships landed infantry and SCVs first. Siege tanks followed. Goliaths occupied ridgelines. Missile turrets began appearing before the final transports touched down. Battlecruisers remained above the mountains rather than descending into the air battle over the basin. Valkyries and Wraiths operated beneath their protection. StarCraft Co. had fewer forces than the UIC army being annihilated below, but it possessed something the government force had thrown away at the moment of landing: terrain. The commander ordered no rescue assault into the basin. That decision drew immediate protests from surviving UIC officers. Thousands of government troops were still fighting below. The commander refused to commit StarCraft Co. into the same killing ground. Instead, he opened evacuation corridors through the western slopes and transmitted coordinates to anyone capable of reaching them. UIC units began retreating toward the mountains in fragments. Marines arrived without officers. Tank crews abandoned damaged vehicles and climbed through the snow on foot. Entire companies had been reduced to dozens of survivors. StarCraft Co. medics received them behind prepared positions while SCVs turned mountain passes into fortifications.

The Zerg followed, which was precisely what the commander wanted. Ultralisks that had crossed the open snowfield without obstruction found themselves entering narrow mountain approaches where only a handful could advance at once. Siege tanks firing from elevated terraces concentrated on the lead organisms. When one fell, its corpse became an obstacle for everything behind it. Zerglings climbed over the bodies and entered interlocking Gauss fire from bunkers cut into the rock. Hydralisks attempting to establish ranged positions discovered that StarCraft Co.'s artillery could see farther down the slopes than they could fire upward. Mutalisks and guardians approached the ridges and encountered Goliaths, Valkyries and layered missile-turret coverage. Science vessels moved continuously behind the defensive line, projecting matrices onto exposed positions and irradiating dense biological concentrations in the passes. The contrast with the basin was brutal. The same swarm that had shattered entire UIC divisions spent the next several hours dying in corridors of rock and snow. StarCraft Co. gave up individual positions whenever necessary, collapsed passages with demolition charges and forced the Zerg to find another route into yet another prepared kill zone. UIC survivors who reached the mountains were rearmed and folded into the defense. Some of them had spent the previous day fighting under generals who still believed a straight line of bunkers could stop enough Zerg if the line was thick enough. Now they watched StarCraft Co. deliberately surrender ground, force attackers into narrow approaches and destroy them a few hundred at a time.

By the end of the second day, the concentration that had destroyed the main UIC army was gone. Tens of thousands of Zerg lay across the approaches beneath the mountains. The surviving government force was a fraction of what had landed, but enough personnel had escaped that StarCraft Co. could use them to reinforce the newly established forward base. The commander did not waste time celebrating. Science-vessel scans showed more Zerg entering the system. The emitter remained active. Everything they had just killed was being replaced. That made the next objective obvious: destroy the emitter. The problem was reaching it.

The renegades had expected someone eventually to understand how important the device was, and they had built accordingly. The psi emitter stood on a mountainous island rising above a vast frozen lake north of the primary Zerg concentration. Terran fortifications surrounded the summit. Siege tanks covered the approaches. Bunkers and missile turrets formed overlapping defensive rings. The lake itself eliminated most conventional ground routes, while the surrounding valleys had been deliberately saturated with creep and Zerg structures. Sunken colonies existed in layers. Spore colonies made low-altitude air operations suicidal. Hydralisks were burrowed throughout the approaches. Every hour the emitter remained functional brought more organisms into the system. Worse, the renegades had hired Protoss. They were not Ara. Intelligence identified them as mercenary detachments drawn from several independent tribes willing to sell military service for resources or political favors. Scouts patrolled above the emitter. Corsairs provided rapid interception against lighter spacecraft. Carriers remained farther back under the protection of defensive structures, their interceptors capable of shredding anything attempting to linger above the installation. Most concerning were several arbiters moving between the air formations. Their presence complicated targeting and offered the defenders tactical options StarCraft Co. did not fully understand. Whatever the renegades had paid, they had purchased one of the strongest air-defense networks the company had ever encountered.

The command staff spent most of a night considering alternatives. A ground assault was discarded first. Even if StarCraft Co. could cross the outer Zerg lines, the ice surrounding the mountain forced vehicles into predictable approaches covered by siege tanks. Assaulting uphill through creep against Terran artillery while hydralisks and ultralisks attacked the flanks would turn the operation into the UIC landing all over again. An airborne strike looked almost as bad. Dropships would have to pass spore colonies, missile turrets, scourge, Scouts, Corsairs, carriers and whatever defensive effects the arbiters could produce. Enough aircraft might get through, but whatever infantry survived the landing would arrive beneath an intact fortress. Someone suggested nuclear weapons. For a moment, the room went quiet. StarCraft Co. possessed them. One ghost reaching firing range could theoretically end the problem without sacrificing an army. The commander asked the science-vessel staff to map detection coverage around the emitter. The answer killed the idea. Protoss observers, Terran missile systems, science-vessel equivalents and Zerg overlords created overlapping detection fields extending far beyond the target. A cloaked ghost might penetrate the outer line. Getting close enough to paint the emitter before being discovered was another matter. “We could send several,” one officer suggested. The commander looked at the casualty estimate. “We could send all of them. That doesn't make it a plan.”

The debate returned to air power because every other option was worse. The commander eventually reduced the problem to the same principle StarCraft Co. had used since Ardea. They did not need to defeat every defender. They needed to create one route to one target for long enough to fire several weapons. Battlecruisers could do that. Probably. The plan was ugly enough that nobody tried pretending otherwise. StarCraft Co. would mass its available capital ships, including the newer Minotaur-class vessels, reinforce them with every science vessel capable of generating defensive matrices, and attack through the narrowest practical corridor. Valkyries and Wraiths would clear scourge and smaller fliers immediately ahead of the formation. The battlecruisers would ignore targets that did not physically prevent forward movement. They would not attempt to win the air battle. They would punch through it, reach Yamato range, destroy the emitter and withdraw whatever remained. Losses were not a risk. Losses were part of the plan.

The commander looked around the briefing room after the final simulation. “We are sacrificing ships.” Nobody argued. The projections showed several battlecruisers unlikely to survive long enough to retreat. “Crews?” an officer asked. The commander had already decided. Skeleton complements would operate the vessels expected to take the heaviest fire. Nonessential personnel transferred to other ships. Damage-control teams remained because without them the formation might collapse before reaching the target. Every crew member received the choice to transfer out. Most stayed.

The assault began before dawn over the northern mountains. StarCraft Co. battlecruisers climbed from behind the ridgeline in a tight formation while Valkyries surged ahead. The defenders reacted immediately. Mutalisks and scourge rose from the Zerg lines. Terran missile turrets acquired targets. Protoss Scouts accelerated toward the formation while Corsairs swept around the flanks. Carriers released clouds of interceptors. The commander ordered the fleet forward. Valkyrie missiles saturated the space ahead. Scourge vanished by the hundreds. Wraiths engaged Scouts while science vessels moved dangerously close to the capital ships, wrapping defensive matrices around whichever hulls were taking the heaviest fire. Battlecruiser lasers ignored most of the fighters and concentrated on scourge slipping through the screen. The first devourers reached effective range and coated a Minotaur in corrosive spores. Its armor began failing almost immediately. The formation kept moving.

Protoss interceptors swarmed across the lead ships. Corsairs cut into the Wraith screen. Missile fire rose from the mountain. A battlecruiser on the eastern edge of the formation lost propulsion after repeated scourge impacts and began falling behind. Its captain reported that the ship could no longer maintain formation. “Can you still fire?” the commander asked. “Yes.” “Then keep firing.” The damaged vessel rotated toward a carrier and charged its Yamato cannon. The blast struck the Protoss ship while interceptors tore into the battlecruiser's exposed hull. Seconds later, another swarm of scourge reached it. The transmission ended. StarCraft Co. continued forward. The defenders had expected the Terrans to turn away when losses became severe. They had misunderstood the objective. Every ship destroyed after the formation entered the corridor had already accomplished part of its purpose by absorbing weapons that were not hitting the vessels behind it. Another battlecruiser lost its forward armor and continued under two overlapping defensive matrices. A third took direct fire from siege batteries on the mountain as it passed low enough for their trajectories to reach. Its port weapons failed. The starboard batteries kept firing.

The emitter appeared on targeting displays. Still too far. The commander ordered the formation tighter. Scourge struck the central ships. One Minotaur disappeared behind consecutive explosions, emerged burning and continued forward. Wraith losses crossed fifty percent. Valkyries were almost out of missiles. Science vessels began losing power from continuous matrix projection. Then the first battlecruiser entered Yamato range. “Fire.” Energy gathered along the ship's forward weapon assembly. A Protoss carrier crossed the firing line. The captain ignored it. The Yamato cannon discharged. The blast struck the outer shielding around the emitter. It remained standing. A second battlecruiser fired. The defensive field collapsed. The third ship began charging. Scourge hit it before the weapon discharged. The ship rolled sideways, losing altitude. The cannon fired anyway. The shot missed the emitter by less than fifty meters and obliterated a section of the surrounding installation.

The commander watched the remaining ships closing. “Again.” Yamato systems began recharging. There was no guarantee they would survive long enough. The Minotaur at the center of the formation reached firing range with half its outer armor gone. Missile impacts struck continuously along its hull. Interceptors swarmed around it. A science vessel projected a final defensive matrix over the forward section and immediately lost power. The Yamato cannon charged. The psi emitter continued broadcasting. More Zerg were already entering the system. The Minotaur fired. The emitter disappeared inside white light. For several seconds the sensors showed nothing. Then the signal stopped.

Across Niflheim, Zerg behavior changed almost immediately. Not dramatically. There was no sudden collapse or mass death. But the continuous influx from orbit ceased. Swarms already approaching the planet began dispersing. Some Zerg under renegade control continued fighting normally. Others became erratic. The strategic equation that had made the entire operation impossible changed in an instant. The commander ordered the surviving fleet out. The withdrawal was almost as expensive as the approach. Protoss mercenaries pursued. Scourge continued attacking. One battlecruiser that had survived the firing run lost engines before reaching the mountain cover and was abandoned after its crew transferred to nearby ships. Another made it back with enough structural damage that the engineers eventually declared repair economically irrational. The fleet that returned to StarCraft Co.'s forward base was substantially smaller than the fleet that left. The emitter was gone. That was enough.

Without endless reinforcement, the Zerg surrounding the mountain became finite. Finite problems were StarCraft Co.'s specialty. The counteroffensive began two days later. Siege tanks moved out from the mountain base beneath science-vessel detection. Marines and medics protected the artillery from zerglings while Goliaths handled the remaining air threat. The first hive complex received almost the same treatment as the northern hive on Khamsin. Sunken colonies were destroyed from outside their range. Spore colonies were shelled until Wraiths could operate safely overhead. Hydralisks attempting to burrow around the siege lines appeared on scanner sweeps and died before surfacing. The difference was scale. Niflheim contained enormous numbers of Zerg accumulated during the weeks the emitter had operated. Destroying them took months.

StarCraft Co. did not rush. The company established permanent forward positions throughout the mountains and gradually expanded outward. Each captured mineral field supported additional SCVs and fortifications. Siege lines advanced kilometers at a time. When resistance became too dense, the army stopped, built another base and resumed bombardment from there. The remaining renegade Terrans fought alongside the swarm. Their siege tanks forced StarCraft Co. to conduct counterbattery operations. Ghosts infiltrated defensive positions and locked down armored units before artillery advances. Renegade bunkers were isolated and destroyed. The Protoss mercenaries withdrew gradually once it became clear the psi emitter had been lost and the renegades could no longer guarantee victory. StarCraft Co. did not pursue them. The contract concerned Niflheim, not settling every score in the system. Winter became irrelevant on a planet where winter never ended. Weeks became months. One hive disappeared, then another. The frozen plains became covered in blackened creep and crater fields. Entire canyon networks were burned clean. Tens of thousands of Zerg died under artillery fire. Thousands more were exterminated by infantry advancing behind the siege lines. StarCraft Co. replaced losses. The UIC mostly sent paperwork.

Government reinforcements eventually arrived, but never in the numbers promised. Some formations fought effectively once integrated into StarCraft Co.'s operational structure. Others spent weeks securing areas that had already been cleared or refusing operations until orders arrived from commands several systems away. More than once, Central Command attempted to redirect StarCraft Co. units already engaged in the campaign because another department had decided the UIC retainer entitled it to priority access. The commander began rejecting those requests without discussion. General Voss intervened where he could, but his ability to protect the campaign from his own government was plainly limited. By the fourth month, StarCraft Co. was technically fighting under UIC strategic authority while operationally making almost every important decision itself.

Six months after the disastrous UIC landing, the final major hive on Niflheim was destroyed. The last spawning pools burned beneath a snowstorm. Science vessels scanned the region repeatedly and found only scattered biological signatures, remnants too small to constitute an organized force. Local teams could eliminate those over time. The planet no longer represented a strategic threat. The cost was enormous. The UIC had lost the majority of the forces committed during the initial assault. StarCraft Co. had sacrificed battlecruisers worth fortunes to destroy the emitter and then spent months feeding ammunition, replacement vehicles and soldiers into the ground campaign. Entire veteran formations had been rebuilt more than once. But the alternative had been worse. Recovered data from the emitter facility showed that the device's effective range encompassed multiple star systems simultaneously. Had it remained operational, the renegades could have continued drawing wild Zerg toward Niflheim until the swarm numbered in the millions. Once gathered, enthrallment systems could have redirected portions toward populated worlds throughout the core. New Carthage would not have been an exceptional siege. It would have been one of dozens.

The UIC paid accordingly. The final settlement was larger than any single transfer StarCraft Co. had ever received. Lost battlecruisers were compensated at replacement value. Combat bonuses, casualty payments, equipment losses and six months of operational expenses were added on top of the original contract. The government also increased StarCraft Co.'s standing retainer, ostensibly recognizing that the company now represented a strategic military asset rather than an ordinary PMC. On paper, the relationship had never been closer. In practice, both sides were beginning to understand that money was becoming the easiest part of it.

General Voss arrived in person for the final briefing. He walked through the company's mountain base past memorial plaques listing ships lost during the emitter run and soldiers killed during the campaign. The commander met him inside an operations room whose walls still displayed six months of shifting battle lines. Voss studied the final map. “All clear?” “Clear enough.” The general nodded. For a while neither man spoke. Then Voss said, “Central Command considers the operation a strategic victory.” The commander looked at him. “The operation where they landed an army in a bowl and lost most of it?” Voss winced. “That portion is being reviewed separately.” “I'm sure it is.” The general almost objected, then stopped. They had moved past pretending.

“There is going to be a formal inquiry,” Voss said. “Senior command wants testimony concerning the opening assault.” The commander looked at him. “Mine?” “Among others.” “Are they investigating who approved the landing zone?” Voss's hesitation answered before he did. “The inquiry has a broad mandate.” “That's not what I asked.” “No.” The commander leaned against the operations table. “Then I'm not especially interested in helping them determine which dead colonel should have objected more forcefully.” Voss's expression hardened slightly. “Careful.” It was the first time the general had spoken to him that way. The commander met his eyes. “I am being careful.”

The silence that followed was different from their earlier disagreements. Voss was still an ally, perhaps the closest thing StarCraft Co. had inside the UIC hierarchy, but he was also a general of that hierarchy. The distinction mattered more every month. “There are people in Central Command,” Voss said eventually, “who think your company is becoming too independent.” The commander almost smiled. “They hired us because we were independent.” “They hired you because you could do things quickly.” “Those are related.” “Not to everyone.” Voss glanced toward the tactical map. “You've ignored deployment orders, conducted operations without authorization and established direct relationships with foreign powers. You now possess capital ships, nuclear weapons, experimental infantry and a private intelligence network. Some people are asking where a contractor stops and an army begins.”

The commander considered the question. “Did any of those people make it to Niflheim before the shooting stopped?” Voss's jaw tightened. “That's exactly the kind of answer that worries them.” “Then they should worry more about why the answer is no.” Neither man raised his voice. They didn't need to. For years, StarCraft Co. had needed the UIC far more than the UIC needed StarCraft Co. The retainer, procurement access and government contracts had financed its transformation from a handful of mercenaries into a military organization. Niflheim made it increasingly difficult to say which side now possessed the greater leverage.

Voss looked again at the casualty figures and some of the tension left his expression. “I know what you think of us.” “I don't think that's a productive conversation.” “Probably not.” He paused. “But I also know we'd have lost this planet without you.” The commander folded his arms. “You might have lost more than this planet.” “Yes.” That admission mattered. Voss turned toward him. “This isn't going away. The renegades are stronger than we believed. They're coordinating with forces outside Terran space. They can breed Zerg, modify them, control them and attract them from systems away. And my military…” He stopped. The commander waited. Voss eventually finished. “…is not adapting quickly enough.” It was the first time a senior UIC officer had said it plainly. “No,” the commander replied.

Voss nodded once, then slid a datapad across the table. It contained the authorization for StarCraft Co.'s payment and increased retainer. Beneath it were several new conditions Central Command wanted added to the agreement: greater reporting requirements, tighter controls on independent deployments, mandatory notification before acquiring strategic weapons and a provision allowing the UIC to place liaison officers aboard StarCraft Co. capital ships. The commander read them without comment. “Those weren't part of the contract,” he said. “They're proposed revisions.” “We won't accept them.” “I assumed you wouldn't.” “Then why bring them?” Voss looked genuinely tired. “Because if I didn't, somebody less reasonable would.”

The commander pushed the datapad back. “They can keep proposing.” Voss took it. “And you can keep refusing. Eventually somebody is going to decide this arrangement needs to be clarified.” The commander looked through the observation window toward the surviving StarCraft Co. fleet above the mountains. New Minotaurs had already been ordered to replace the ships lost over the emitter. Veterans of the UIC's destroyed army were submitting applications to join the company in numbers large enough that recruiting had begun turning people away. Engineers wanted to work on StarCraft Co.'s weapons programs. Colonial governments increasingly contacted the company directly before bothering with Central Command. Even some UIC officers had started routing intelligence through the commander because doing so produced a response faster than sending it through their own hierarchy.

“Maybe it does,” the commander said.

Voss studied him for a moment, as though trying to decide whether there was more in the answer than the words themselves. If he found anything, he didn't say so. They shook hands before the general departed. It remained cordial. Technically, they were still on the same side.

Outside, StarCraft Co. battlecruisers moved above the mountains while SCVs dismantled forward positions no longer needed. The company had spent six months paying in blood for mistakes made during the first hours of the campaign, but it had emerged larger, richer, harder and considerably less willing to accept orders simply because the people giving them carried government insignia. The UIC had money, fleets, ministries and more soldiers than StarCraft Co. could yet hope to field. It also possessed layers of leadership that had repeatedly demonstrated an extraordinary talent for converting overwhelming advantages into disasters.

Open war had begun with the UIC attempting to demonstrate that it could still solve the crisis through overwhelming conventional force. Niflheim had demonstrated something else. The UIC had an army. StarCraft Co. knew how to fight a war.

And, increasingly, StarCraft Co. was beginning to wonder exactly who should be deciding how that war was fought.

Thanks Brits!

Thanks to the British, we now have data that should shut down American Football, but it won’t.

When you are poor and athletic, nothing brings the promise of a better life like two-a-day football practice in high school. Like musicians, athletes are dangled in front of wealth, but rarely ever experience it. We’ve known for a while about CTE. Now we have numbers. (Again, thanks to the British.) We put kids in meat grinders here in the USA, tell them they’ll be rich. I have met kids wanting to play football to buy their parents a house. What do you think the odds are of that?

(Google says…)

“How likely is a highschooler in Texas in football to make enough in football to buy parents a house with football money?”

To buy a house, a player generally needs to secure a massive Name, Image, and Likeness (NIL) deal or reach the NFL. 

• The High School Population: There are roughly 170,000 high school football players in Texas at any given time.
• The College Jump: Only about 7% to 9.5% of high school players make it to any college roster, and only about 3% make it to a Division I FBS school, where the major money resides.
• The NFL Jump: Nationwide, out of more than 1 million high school players, only about 250–300 get drafted into the NFL each year. That is a 0.025% chance (roughly 1 in 4,000) of going from high school to the NFL. 

Because Texas produces a disproportionate amount of elite talent, a Texas player's individual odds might be slightly higher than the national average, but they still face a fraction of a percent chance.

What’s the deal with brain damage in America?

In the USA, we do not think much about brain damage, even if you are rich! If you start out poor, we bless you with wealth for taking on such brain damage. Not that you can enjoy it.

I am reminded of Romans and gladiators. I know this news this morning from Britain will fall on deaf ears because folks like football. They like music too! But that doesn’t mean they like the players involved. So athletes and musicians get taken advantage of with lottery-type career systems and concussions.

Are “earning” and “winning” the same thing? What about the losers? Didn’t they earn a living too, doing the same thing? Okay, who out there does not deserve to live?

A musician or athlete does not “earn” a living. They “win” it, often taking it from someone else. Folks do not “work” their way up, no matter how hard they work. It’s a lie. I am not saying they do not work. Many work circles around me. But to get the good life, luck is involved. .5 percent make it to NFL who start out in high school. Now we know 1 in 4 in the NFL get CTE. All that work and luck just to wind up brain damaged. CTE is the second form of brain damage these kids incur too. The first was joining football to begin with, those kind of odds. But they sure do! Also, they’re in high school. Not impressionable or anything ..

Let’s talk about lottery for a moment…

“What are popular thoughts against lottery systems, Google?”

(Google says..)

Popular thoughts against state-run financial lottery systems focus on their mathematical design, economic impact on low-income groups, and targeting of vulnerable communities. 

Mathematical Odds

• Extremely low chances: Winning a major jackpot like Powerball has odds of roughly 1 in 292 million.
• Negative expected return: Payouts average only about 60% of total revenue, meaning habitual players lose a vast majority of their spent money over time. 

Economic and Regressive Impact

• Regressive structure: Lower-income households spend a significantly higher percentage of their disposable income on tickets compared to wealthier individuals.
• Opportunity cost: Money spent on lottery tickets diverts funds from practical savings, investments, or basic household needs. 

Social and Demographic Concerns

• Targeted marketing: Critics note that ticket retailers and advertisements are often concentrated in lower-income and minority neighborhoods. 
• Exploitation of hope: Advocacy groups argue that state-sponsored gambling preys on people looking for financial escape from systemic poverty. 

Government Reliance

• Flawed revenue source: Critics argue that using lotteries to fund public services like education allows governments to avoid raising standard, progressive taxes or providing consistent budgetary support.

Who fixes this?

“Targeting vulnerable communities..” I see it says that. The flawed revenue source is good too. Instead of taxing millionaires and billionaires for civic upgrades and fair policy, we’ll just have a lottery. The plebs can do it themselves. At least they’re right about us doing it ourselves. But we have to get rid of team owners.

I want nice things too. But I also don’t want your kids to be brain damaged so I can watch a damn game.

One thing you get when folks think you’re rich is suck ups. Sycophants. Many people daydream about having folks concerned about their life, achievements, aspirations. Many folks want a team to work on. Many folks like money. Me! I am folks!

I don’t feel I am “owed” any of that stuff. I don’t even really like the sycophant thing since I quit social media and realized how fake everything is outside of it, (inside too, wooboy). I don’t feel like I have “earned” it, either. I am kinda’ lazy, do whatever art projects I want. I love my life, don’t get me wrong, but it isn’t because I am making bank. I am making decisions. For me, that’s just as good.

What do you want money for?

I create mainly because it gives me things to do, think about. That’s what a lot of folks want money for. Guess I found a softer way on accident. (It totally was not on accident, fyi. My reason for music is simply getting to do it. Happy place.)

Music and NFL fans, your sign.

Seriously NFL fans, the burden is on your shoulder pads . Music fans too. Do not reward slimy team owners with more money for 1 in 4 cases of CTE. That’s gladiator numbers. Do not reward greedy entertainment companies and managers using micro-targeting for pop-stars practically vacuuming the money and attention from your very life. That’s brain damaged too. So, I know you won’t do it. I just don’t see it getting better for these kids in high school until you do.

Kelly Kintner

The Horse’s Mouth magazine.