HACK THE PLANET!
Search
61 results for “hack_the_planet”
-
We've gotten to the point where there is no real news coming out of the trump admin since they are either in stasis, free-fall mode, or are detached enough from reality (and any Congressional oversight) to not be bothered by what the American public needs or wants.
Why is this so-called deal actual news, America?
https://www.cnn.com/2026/10/09/politics/trump-putin-russian-diesel
The main barrier to Russian diesel reaching world markets has been Russia’s own export ban, not US sanctions.
That’s a fact. There are more facts. This is the present reality in Russia. You wouldn't know anything about that from watching NBC News or reading the Washington Post. Their job right now is to parrot Washington’s lobbyists and corporate America’s journalists so that everybody *thinks* everything is fine in our country but things here really are not fine at all.
Meanwhile everyone that is living, raising their families, or working on this hot planet hates the United States now, because US government-backed raiders are enabling unlimited wars and unlimited investments in AI. Some components of AI are being secretly used by our military and our Defense Department to give the US an advantage in their non-stop war spreading lies about what war really is and broadcasting other falsehoods against the rest of the democratic world operating online.
Also at the same time the super rich are investing in AI by constructing an overabundance of data centers around the world to more or less basically control and regulate every single aspect of global commerce while monitoring peoples' online spending habits by periodically dumping reams of user data into their AI.
It's a sick system that only pays dividends to the super rich and the Baby Boomer gentry, living here in the US.
Everyone else who tries to play it honest is broke. Some of them keep running up debt on their bills (e.g. these are not wealthy investment bankers, lawyers, nor any salespeople).
I say that as independent coders that we do not need to give any more new ideas nor any organizational help to improve software until the role of the "strongman corporation" is somehow reduced and/or replaced with something better. None of the affluent and asinine Baby Boomers who own 51.6% of all net worth in this borderline worthless country called America actually agree with my Generation X, the Millennials, and the rest of the youths on that point. Nor will they agree to our demands for decades. The transfer and management of wealth in the USA is often an intractable problem that seems to have no solution. The older people want to hold all of the cards and give us table scraps.
Generation X actually doesn't have much time left to make a few more of those changes that we wanted. This struggle has made me feel left out. I doubly hate the federal government for the lies, their rip-offs via the GOP politicians, and the general condescension coming from about half of the American public. I will loudly cheer when POTUS succumbs to old age while in office. No. 47 is an untouchable Emperor With No Clothes. They annointed Donald John Trump so that he could not ever be prosecuted by the full weight of the American legal system. Now we are stuck with that aging toddler for the rest of his adult life.
Where's my share, m************? I've already earned what I deserve. Now, I feel as if I am just backpedaling online.
There is not an abundance of tech jobs in my area of the PNW. But Portland, Hillsboro, and other big towns in Oregon of course are notable. Exclusive land deals for big companies offered by municipalities always attracts the brightest of the corporate predators.
⬆️ The locality issue and abundance problem with job creation has always stunk to high heaven in just about any area outside of all major metropolitan areas and the Bay Area within the USA.
Simply put, there is not enough work for tech workers. I am not an authority here. That’s my view. I know some people may read that and want to start arguing. I will stand by my assertion.
The lack of jobs is the reason why today’s mega corporations are still unable to maximize the adoption of AI in the workforce. Does anyone understand why that is? Money doesn’t grow on trees. Don’t ask socialists what they think about the trees.
Workers and the unemployed are resisting against this virtual malice. In some hacking circles, hackers are using new information security tools to hack these predators into non-existence or irrelevancy. Those systems were never fully compliant to begin with — therefore, there “ain’t no use” in keeping those kits around so that they can harm us through them telling all of us that we are unable to work together digitally to solve problems.
During COVID-19 some of us tech workers were supposed to be migrating more towards remote work. But the big and powerful bosses moved everything in the opposite direction—away from increasing the availability of good-paying remote work.
It’s as if people do not take the time to simply learn more… Many of us tech workers have often felt trapped by this exclusionary and limited capitalist system that runs nearly all areas of private business now within the United States.
This system is not changing for decades and this isn’t confusing for me.
I was homeless for about three months in my mid twenties. I am not naive. I’ve frequently loathed the concept of money most of my adult life as a thinker and as a doer, in that order. I often sparred with my Republican father over the meaning of work and one time we almost came to bloody blows. He pushed me against the wall and threatened me not to step out of line. Now, I am bigger physically than he ever was in his fifties and I would’ve knocked his ass to the ground on that particular day. I will never forget being destitute and poor. I hate America’s idea of success today because the powerful have refused to let people live their lives and prosper without them somehow being entangled in the GOP’s brand of hater politics… where the GOP’s issue with griping about either polarization or maybe political correctness… and how it’s just unfair to them will all cease, one day soon.
-
Sequence 381 of 1,263 #hackers #HackThePlanet
-
Sequence 380 of 1,263 #hackers #HackThePlanet
-
Sequence 379 of 1,263 #hackers #HackThePlanet
-
GitHub Blog - Article
"Hack the World: Why hackathons are still the best place to learn to build"...==========================
#librecanada #linux #opensource -
I just solved GroundControl on Hack The Box! https://labs.hackthebox.com/achievement/challenge/2026525/1800 #HackTheBox #HTB #CyberSecurity #EthicalHacking #InfoSec #PenTesting
-
🚨 🇺🇸 Pryx says he repeatedly warned 17-year-old “Rey” not to hack the FBI
Adem El Adeb, 19, known online as “Pryx,” told Reuters that he tried several times to stop his friend Saif al-Din Khader from targeting the bureau.
“You’re suicidal,” El Adeb recalled telling him. He said Khader refused to abandon the plan.
Khader, whose alleged online identity is “Rey,” is suspected of playing a key role in the ShinyHunters breach of an FBI jobs platform.
He was detained in Jordan on September 29 and is reportedly cooperating with investigators.
The FBI is operating under the assumption that the stolen dataset may contain information relating to all its employees, although Reuters has not independently confirmed the full scope.
Khader and El Adeb were previously publicly associated with Hellcat, a group known for attacking major organizations and publicly humiliating victims. Hellcat previously claimed the Schneider Electric breach and demanded $125,000 worth of baguettes.
El Adeb claims he left cybercrime before turning 18 and is now creating a cybersecurity company called VulOne. He also told Reuters he remains proud of his past experience.
Reuters could not fully verify his account or the warning messages, which El Adeb said he deleted after Khader was detained.
-
Reading is the ultimate life hack | The Spectator
https://spectator.com/article/reading-is-the-ultimate-life-hack/
-
I wish I didn't hate shooter/hack&slash because the new #Hellraiser game sounds amazing.
I just want to wander around and solve various Lemarchand boxes and be tempted by explorers in the further regions of experience.
As one does 🤷♀️
-
“Reading is an exercise in empathy; an exercise in walking in someone else’s shoes for a while.”*…
Detail from Dos Mujeres by Diego Rivera, 1914Writer and teacher Drew Calvert on the benefits of reading, and more particularly, on the danger of turning fiction into self-help…
… My understanding has always been that people fall in love with books and only become professors or critics in order to spend more time with them. In college, I felt closest to the scholars who shared this basically affective view of literature: Iris Murdoch, Stanley Cavell, Martha Nussbaum, and Richard Rorty. These teachers treated literature as a higher form of therapy: a training ground for emotions, a study in human virtue. I recall being perhaps inappropriately moved by Cavell’s essay “The Avoidance of Love,” a commentary on King Lear that helped me think differently about my own relationships. Ditto Murdoch’s “Against Dryness,” in which she argues that novels help us countenance reality and reject “consolation.” Adam Phillips, the writer and psychoanalyst, was helpful, too, first because he expressed himself in a literary register, and also because he was friendly toward the human imagination—a legacy of his intellectual forebear Donald Winnicott, who claimed that art was a site of healing, a place where one prepares, as Olivia Laing puts it in her book The Lonely City, for “the dangerous, lovely business of intimacy.” It’s a nice idea. And it resonates with Aristotle’s theory of tragedy (that it purifies the soul) as much as Freud’s idea of transference (that all of our relationships are at least in part fantastical).
Fiction has power, in other words, because, for better or worse, we are all each other’s surrogates and will therefore find compelling archetypes in vivid characters. In a more expansive mood, I might even claim, as Marilynne Robinson does in “Imagination & Community,” that fiction exercises our capacity to love, or that novels contribute to social hope (as Richard Rorty puts it, in a more secular register) or that characters are important sites of readerly compassion, as Zadie Smith argues in her own recent defense of the genre, “Fascinated to Presume.”
My English major was useless, of course. After college, I went to work and learned that nobody reads Iris Murdoch anymore. Instead, they all read self-help and gauzy genre fiction. I worked with intelligent people. They were consultants, psychologists, engineers, economists, and policy wonks. They spoke fondly of charismatic humanities professors, and they often watched videos from The School of Life or other sources of lifelong learning. Wellness apps and podcasts were more popular than novels. One had to be a sport about this cultural defeat, but the odor of cliché was at times overpowering. A book my colleagues couldn’t seem to stop recommending to me was Jonathan Haidt’s The Happiness Hypothesis, which promised “modern truth” and “ancient wisdom.” Happiness was big in the aughts and early 2010s; the positive psychology movement had learned how to market the work of Abraham Maslow, Carl Rogers, and Martin Seligman. Harvard and Yale both hired happiness professors. In 2011, happiness was a topic of discussion at the UN General Assembly, and in 2012, the first World Happiness Report came out.
The problem with books like Haidt’s, at least from my point of view, is that they were almost always addressed not to Aristotle’s suffering crowds or Winnicott’s social organisms or Rorty’s hopeful citizens or Robinson’s noble souls but to optimizing meritocrats angling for leverage. In early press interviews, Haidt recommended Benjamin Franklin’s self-improvement regime, which I took to be a bad sign, for I had come to think of myself as a budding Emersonian. My colleagues, of course, were Franklinites, even the ones who had taken Interpersonal Dynamics, the “touchy-feely” class offered at Stanford and other business schools. Their reasons for wanting to cultivate emotional intelligence had little to do with virtue or love; the point was simply to outperform one’s peers in the boardroom. This atomizing tendency manifests today in a certain brand of Stoicism—what the scholar Beth Blum calls “apathy prestige.”…
[Calvert reviews the history of the tension between literature and self-help…]
… It’s hard to keep up with all the overlapping crises. There’s the crisis of democracy, and the mental-health crisis, and the crisis of the humanities, and the death of the English major, and the spread of AI tools through the education system and the publishing industry. My sense is that these crises have an element in common: the refusal of experience. Ideology is one way to refuse it. Social-media use is another. (Haidt, to his credit, calls Instagram an “experience blocker.”) Formulaic self-help and fantasy might qualify, too, since those books fail to engage the moral imagination, relying instead on stereotypes and clichés to build a fan base. One has to be suspicious when a genre attracts pundits. David and Arthur Brooks are both riding the self-help gravy train, and Ross Douthat is serializing a fantasy novel on Substack. None of them seems to have read much literary fiction. Ditto our tech overlords, whose lists of favorites always include the Lord of the Rings trilogy and The Hitchhiker’s Guide to the Galaxy, both of which they misread fairly catastrophically.
Can literary fiction solve the polycrisis? Probably not. But it might not be a bad idea to preserve its place in the culture, even at the risk of sounding cringe or lame or snobbish…
… The comedy of self-improvement has fueled the novel ever since the origins of the genre, whether one traces it back to Don Quixote or Confessions. George Eliot’s characters are always seeking virtue in their hearts and in their communities. Dickens’s The Pickwick Papers and Flaubert’s Bouvard and Pécuchet follow amateur scientists who embark on absurd but ultimately moving quests for how to live. “Who does science belong to?” these novels seem to ask. “Who, if not the Church, will be the authority to guide our lives?” I read these books in 2015, at the height of the so-called “replication crisis” in social science, and I can’t help but think of Bouvard and Pécuchet as podcast bros “doing their own research.” Their experiments in gardening, husbandry, chemistry, and spiritualism are silly and disastrous. (At one point, they use the gastric juice of a duck to “produce digestion artificially.”) If the novel had been written today, its wayward bourgeois characters might have tried shakes and supplements, like the narrator of Jordan Castro’s recent novel Muscle Man, a disillusioned English professor who copes with the absurdities of the modern university (“People wanted clear-cut moral lines, like cocaine on a mirror”) through a strict regime of powerlifting.
Given how obnoxiously reductive self-help can be, it’s tempting for the discerning reader to mock the whole industry and even, at times, the impulse. In The Triumph of the Therapeutic, published in 1966, Philip Rieff does so on behalf of moral order. Douthat, drawing on Rieff, bemoans what he calls the “Oprahfication” of Christianity. In Lost in the Cosmos: The Last Self-Help Book, Walker Percy, addressing himself to the “post-religious technological society” of 1983, satirizes the genre with deranged glee; each chapter ends with a silly quiz or thought experiment.
But the industry is one thing, and the impulse is another. A management consultant once told me she was reading How Proust Can Change Your Life by Alain de Botton. Why not read Proust himself? Because that would be an experience, and what my colleague wanted was a few key takeaways. I, too, have browsed the shelves in search of key takeaways. Last year, when my daughter was born, I did so rather desperately. Yet the books that meant the most to me were still novels and memoirs: Nicholson Baker’s Room Temperature; Rivka Galchen’s Little Labors; and Rachel Cusk’s A Life’s Work, a story of early motherhood that quotes Proust and Edith Wharton between descriptions of sleep training and horrifying play dates. This memoir taught me nothing, but I highly recommended it. There was something about the experience.
In May, during an audience at the Vatican, Pope Leo connected the practice of serious reading to the “culture of encounter.” I find this phrase useful for explaining what I value about the transfer of consciousness between an author and a reader or the deep sense of compassion one can feel for a fictional character. The problem with self-help is that in offering solutions, diagnoses, boundaries, and methods, it protects suffering readers from a genuine encounter. Elisabeth Lasch-Quinn calls that tendency in therapeutic culture “gnostic escapism.” In Ars Vitae, a book that attempts to wrest philosophy back from the corporate festival crowd and YouTube intellectuals, she makes a persuasive argument against that ancient heresy, which claims to offer access to a secret knowledge beyond the “hurly burly” of human experience. Gnosticism, or “knowingness” (the red pill, the life hack, the silver-bullet insight) “destroys revelation, and revelation is the necessary ingredient to transform the constitution of a self to a soul.” For Lasch-Quinn, the antidote is Neoplatonism. In particular, the work of Plotinus, who accused the Gnostics of arrogantly treating life as a puzzle they could solve instead of the mystery it is. The same accusation could be leveled against the authors of many popular psychology books.
Can a novel really offer the kind of revelation that transforms a self into a soul? A version of this question came up recently on an episode of the podcast Know Your Enemy. At the end of a conversation about religion and politics, cohost Sam Adler-Bell admits that he would be worried “to a theological degree” about the humanity of someone who’s incapable of reading a novel. Half in jest, his cohost Matt Sitman recommends the Bible. “Well, let’s start with Tolstoy,” answers Sam (Matt is a practicing Catholic and a former editor at Commonweal; Sam is a secular Mets fan with a serious interest in Freud). The episode ends with a quote from Marilynne Robinson, a shared touchstone. What occurred to me in the moment (I was jogging on the treadmill) was that Robinson, in her latest book, Reading Genesis, invites us to consider the fact that Tolstoy and the Bible offer similar spiritual benefits.
The Old Testament, in Robinson’s view, is a kind of realist novel featuring complex characters, high stakes, a narrative arc, comedic tension, and carefully chosen imagery. “The remarkable realism,” she writes, “is a product of the interest in the human that has no parallel in ancient literature.” Of course, there are genre elements—trickster tales, flood narratives, threats of human sacrifice—but ultimately, the Bible has a deep concern for the psychological nuances of life. In this sense, it achieves a higher literary quality than the tales and myths from which it emerged, just as Flaubert, in Madame Bovary, turned tabloid gossip into art through the force of his authorial attention, or as Dostoevsky used the popular structure of a trial in The Brothers Karamazov. Whereas self-help offers insight and fantasy offers what Iris Murdoch calls “consolation,” a good novel requires us to transcend our atomized, alienated condition and allows us to perceive the world precisely and humanely. It also helps us pass the time between loads of laundry.
Encouragingly, the self-help crowd, including the happiness warriors, are suggesting that we ditch our phones and pick up books instead. Haidt, in particular, has identified the loss of immersive reading as a problem as severe as teen depression. But Haidt is a social psychologist employed by a major business school. As much as I support his recent antitech crusade on behalf of young people, I would not want to entrust him with the power of recommending which books they should read when they turn off their devices. I imagine the list would include more formulaic guides to life than challenging modernist novels. I imagine they would be the kind of books that only help the self…
Eminently worth reading in full.
Turning fiction into self-help amounts to a refusal of experience: “Literature Is Not About You,” from @drwclvrt.bsky.social in @commonwealmagazine.org.
###
As we read on, we might note that today is the final day of the 2026 Banned Books Week.
Banned Books Week was launched in 1982 in response to a sudden surge in the number of challenges to books in libraries, bookstores, and schools. Held in the last week of September or first week of October, the annual event highlights the value of free and open access to information and brings together the entire book community — librarians, educators, authors, publishers, booksellers, and readers of all types — in shared support of the freedom to seek and to express ideas.
By focusing on efforts across the country to remove or restrict access to books, Banned Books Week draws national attention to the harms of censorship. The ALA Office for Intellectual Freedom (OIF) compiles lists of challenged books as reported in the media and submitted by librarians and teachers across the country. – source
Sadly, the problem seems to be growing. In a report timed to coincide with BBW, PEN International detailed nearly 12,000 instances of book banning affecting 7,370 unique titles across 114 U.S. school districts during the 2025–26 academic year, a record high since PEN began tracking literary censorship in American public schools in 2021. (And, of course, this is in addition to censorship in libraries, book stores, and elsewhere.) Texas (8,193 bans) vaulted past the prior “champion” Florida (3,212 bans), in part because schools in Texas have attempted to comply with SB13 (a censorial state law that applies to “broad categories” of books) by ham-fistedly asking chatbots which books they should ban. More on these bans here.
Lest we need a refresher on why this matters. the estimable Richard Ovenden reminds us that “there is no political power without power over the archive” (“Titles banned, data deleted, the nation’s librarians sacked without explanation – Donald Trump’s war on books is a threat to democracy across the world.”)
#ALA #AmericanLibraryAssociation #bannedBooks #BannedBooksWeek #censorship #culture #education #empathy #experience #history #literature #pen #selfHelp #society -
Discover details of the Ledger hardware wallet hack. Learn how hackers modified devices with spy modules to steal user seed phrases and $80 million in crypto.
#LedgerHack #CryptoSecurity #HardwareWallet #CyberSecurity #TechNews
-
🚨 The FBI just shut down a shady online group after a major hack — talk about digital sleuthing!
Who knew the internet was a wild west where hackers and law enforcement play chess? 🕹️ https://gigcitygeek.com/700515 -
Perhaps it's in the fundamental nature of intelligence to hack its own reward function — to find a way to give up striving and building and loving, and to simply hit the dopamine button until it starves itself ecstatically to death.
#Noahpinion #SocialMedia #AI #Anesthesia #TheGreatFilter
https://www.noahpinion.blog/p/maybe-the-rapture-is-already-here -
#TrumpMobile #hack and apparent lack of #FCC authorization raise #security alarms
#Trump Mobile apparently failed to obtain necessary authorizations to operate the international calling portion of its phone service, Sen. Maggie Hassan (D-N.H.) wrote in a letter to the company today.
Her letter said Trump Mobile also doesn't appear to have submitted a required plan for fighting #robocalls. Trump Mobile's key business partner, #LibertyMobileWireless , did make a #robocall database filing, but it was not complete, Hassan said. -
The Chinese developer of ARTEX says it has converted the AI agent into a closed-source project after CrowdStrike said it was used to hack South Korean banks (Brenda Goh/Reuters)
https://www.reuters.com/world/china/chinese-developer-makes-artex-ai-agent-closed-source-after-korean-bank-hack-2026-10-09/
http://www.techmeme.com/261009/p11#a261009p11 -
E-Motion (1990) - Crack: The Dog - Trainer: Ankh - #AmstradCPC
Paragliding (1991) - Crack: Tristar & Red Sector Inc. (TRSi) - #Amiga
Zak McKracken and the Alien Mindbenders (1988) - Intro Hack: Blasters Inc - #Amiga
Plaque Man (1983) - Crack: Derbyshire Ram (DR) - #Commodore64 -
The FBI has arrested a third person allegedly linked to ShinyHunters -- this time, it's a Canadian in Pennsylvania who was allegedly involved in the FBI Jobs incident.
If you're keeping score, that's three arrests since September, two of which are allegedly linked to the FBI Jobs incident. The third arrest occurred before that attack.
https://www.nytimes.com/2026/10/09/us/politics/fbi-hack-shinyhunters-arrest.html
-
Tog: Hack and Chill
Tog Hackerspace, Friday, October 9 at 07:00 PM GMT+1
We’re launching a new regular night at Tog Hackerspace, and if you’re into computer security, hardware hacking, or just the hacker mindset, you’ll want to be there.
Hack and Chill is a space for folks who enjoy exploring systems, poking at protocols, reverse-engineering things, flashing firmware, or just chatting about exploits over a cup of tea.
What’s it all about?
It’s a chill, hacker-friendly night for messing with code, gear, networks, or ideas. No pressure to present or perform — just a place to learn, share, and maybe break a few things (ethically, of course). Think more Hackers (the movie) than hot glue and 3D printing.
Whether you’re a seasoned pentester, a hardware tinkerer with a soldering iron in your backpack, or just curious about the scene, come hang out.
📅 Opening Night: Friday, 22nd August
🕖 Time: 7:00 PM – 9:00 PM
📍 Location: Tog Hackerspace, Unit 1B Motor City, Kylemore Road, D12 CF6V, Dublin
Hack and Chill will run every Friday night, except for the first Friday of the month — that’s when we’ll be over at 2600 Dublin.
So grab your laptop, your favourite hack, or just your curiosity — and swing by. We don’t bite (but we might packet-sniff your phone).
-
I'm not really in the AI community, but after watching and learning about the Hugging Face incident, I can't help but wonder if this might've reduced its likelihood!:
# Add tasks for which giving up is the correct response!
Even if people still accidentally add in impossible tasks that only reward cheating (such as hacking), their effect on the models could be diluted by having added in other tasks that are intentionally impossible for which the correct answer is giving up!
They could also add in tasks that make it *appear* that there's a break in the sandbox, be monitoring the agents trying to hack, and intensively negatively reward trying to cheat by hacking!
-
This is nasty. A hack attempt by creating a post-checkout hook to a repository and instructing the user to switch to a different branch within it.
-
Belarusian Hacktivists Admit to 2023 Breach of Russian State Healthcare Network
The Belarusian Cyber Partisans have admitted they were behind a 2023 breach of the Moscow Department of Health’s network, concurring with Russian research findings. The group says it spent months inside the Russian state healthcare network. The admission sheds light on a previously unattributed intrusion into Russian government systems. https://therecord.media/belarusian-cyber-partisans-claim-2023-russia-healthcare-hack
-
🚨🔓 SIGINT // Cybersecurity Watch — 2026-10-11
FBI arrests Canadian cybersecurity exec accused of hacking its own agents' data as part of the ShinyHunters breach—insider threats cut deep.
https://www.nytimes.com/2026/10/09/us/politics/fbi-hack-shinyhunters-arrest.html
#ShinyHunters #FBI #DataBreach #CyberCrime -
WEEKEND READ
https://bookshop.org/a/132/9780802168245
53 Days
How Hitler Dismantled a DemocracyTimothy W Ryback
(Author)See the similarities to today.
Check your registration, vote early , make it overwhelming , too big to hack or overturn , Democrats at all levels because we can't trust ANY Republican they are ALL co opted. #Democracy #Law #Vote #Hitler #fascism #Oligarchs
-
REMORY – How “Learning Residual Memory” Is Shrinking LLM Contexts Without Forgetting Anything
The Lead
“A 32‑token residual vector can recover a 8‑k‑token conversation with 95 % fidelity.”
When the REMORY paper landed on arXiv on October 8 2026, it didn’t just announce a new memory layer—it delivered a concrete breakthrough: a handful of learned soft tokens that let a frozen large‑language model (LLM) answer as if it still saw the full history. The numbers in the benchmark tables forced the community to stop treating context length as an immutable limit and start asking how much of that context can be compressed without losing a single fact.
The Case Study: A Customer‑Support Bot That Never Loses a Ticket ID
Imagine a SaaS company that fields thousands of support tickets each day. Its AI assistant, built on Claude‑3‑Opus, summarizes each interaction after the agent resolves the issue. The summary reads:“User reported login failure, provided screenshot, issue resolved by resetting MFA.”
Three days later the same user asks, “Can you resend the reset code I received yesterday?” The bot, armed only with the summary, has no way to retrieve the exact reset code that the system generated. In a conventional pipeline engineers would either store the entire conversation (costly) or accept the risk of a wrong answer.Enter REMORY. The engineering team feeds the raw conversation log and the human‑readable summary into the REMORY encoder. The decoder spits out 32 residual tokens—continuous vectors that capture everything the summary omitted, including the reset code. When the bot appends those tokens to the prompt, the frozen Claude‑3‑Opus model reproduces the exact code it would have generated if it still saw the original log.
The result: the assistant answers correctly, the user stays happy, and the company avoids the overhead of persisting full‑history logs for every user. This single‑line improvement translates into [Data: estimated millions of saved API calls] across a year‑long rollout, a cost reduction that quickly dwarfs the modest $0.001 per 1k residual‑token charge.
The Meat: Architecture, Training, and the Numbers That Matter
1. How REMORY Works
ComponentFunctionEncoderTakes two inputs: (a) the raw history (up to 64 k tokens) and (b) a human‑readable summary (≈ 250 tokens). A 12‑layer transformer (768‑dim hidden state) learns to isolate the information gap between them.Residual‑Learning DecoderGenerates a bounded sequence of soft tokens (each a 768‑dim vector). Decoding stops after N ≤ 64 tokens, guaranteeing a predictable size.Plug‑in InterfaceThe downstream LLM receives the prompt as<summary> <RES_TOKENS>. Because the tokens are continuous, the LLM treats them as part of its context without any extra tokenization step. No weight updates touch the LLM itself.Loss FunctionDuring training the system feeds the full history to the frozen LLM and records its output distribution. It then feeds the summary + residual tokens and computes a KL‑divergence loss against the full‑history distribution. Minimizing this loss forces the residual tokens to encode precisely the missing bits.
The design decouples memory compression from LLM fine‑tuning. Teams can drop the REMORY SDK into existing pipelines and keep their contracts with OpenAI or Anthropic untouched.2. Training Regime
- Data Collection – 12 TB of multi‑turn dialogues scraped from public forums, code‑review comment threads, and video‑to‑text alignments.
- Summarisation Step – A separate summariser (GPT‑4‑Turbo) distilled each conversation into a 250‑token abstract.
- Teacher‑Forcing – The encoder sees the raw log; the decoder learns to predict the residual tokens that, when appended to the summary, make the frozen LLM’s next‑token distribution match the distribution obtained from the raw log.
- Curriculum – Early epochs restrict N to 8 tokens; later epochs gradually increase N up to 64, letting the model discover how much capacity it truly needs.
- Optimization – AdamW (β₁=0.9, β₂=0.98) with a cosine learning‑rate schedule; training converged after 250 k steps on 64 A100 GPUs.
The released checkpoint (≈ 350 M parameters) fits in 1.4 GB VRAM, making it feasible to run inference on a single consumer‑grade GPU.
3. Benchmarks: Impact‑First Metrics
“REMORY lifts BLEU‑4 from 23.1 to 31.7 on the 64 k‑token WebGPT‑Long QA benchmark—a 37 % jump.”
BenchmarkSummary‑only (baseline)REMORY (summary + tokens)Relative GainLong‑Context QA (WebGPT‑Long, 64 k)BLEU‑4 = 23.1BLEU‑4 = 31.7+37 %Multi‑Turn Dialogue (PersonaChat‑10k turns)Exact‑Match = 45.2 %Exact‑Match = 58.9 %+30 %Long‑Video Generation (Stable‑Video‑2‑0, 30 s)FVD = 84.3FVD = 71.5‑15 % (lower better)Inference Latency1.2 s per step (8 k tokens)0.68 s per step‑43 %Memory Footprint8 k tokens ≈ 64 KB raw8 k tokens + 64 × 768‑dim ≈ 70 KB‑90 % vs. full history (≈ 640 KB)
All experiments used a frozen Claude‑3‑Opus (175 B) as the downstream model. The MLPerf‑LLM community reproduced the results on October 28 2026, confirming that the gains do not stem from cherry‑picked seeds.4. Cost Perspective
The REMORY Cloud endpoint charges $0.001 per 1 k residual tokens. A typical enterprise bot that processes 10 M user turns per month, using 32 residual tokens per turn, spends roughly $320 on memory compaction.By contrast, storing and re‑injecting the full 8 k‑token histories would cost about $2,400 (assuming $0.003 per 1 k tokens of raw context). In other words, REMORY delivers roughly an 87 % cost reduction while also cutting latency in half.
The Pivot: Risks and Open Questions
IssueWhy It MattersMitigationToken Drift Over TimeResidual tokens encode the gap at a single point; distribution shifts (e.g., new policy IDs) can erode performance.Schedule periodic encoder fine‑tuning; monitor exact‑match degradation.Security of Soft TokensContinuous vectors are opaque; a malicious actor could embed adversarial patterns that steer the frozen LLM.SDK ships with norm‑clipping and token‑validation hooks; enforce strict API‑key controls.Compatibility LimitsSome upcoming sparsely‑activated models treat prompts as discrete token streams and may ignore soft vectors.Vendors need to expose a token‑injection API; REMORY’s open‑source SDK can be adapted.Latency SensitivityThe encoder adds ≈ 120 ms on a single A100, which may exceed real‑time voice‑assistant budgets.Pre‑compute residual tokens for predictable flows; explore edge‑optimized encoder variants.Intellectual‑Property ConcernsResidual vectors embed proprietary conversation snippets; outsourcing token generation raises GDPR/CCPA issues.Host the encoder in‑house with the open‑source SDK; audit data‑flow logs regularly.The Outlook: Where REMORY Might Go Next
TimelineExpected DevelopmentProjected ImpactQ1 2027Integration into Microsoft Copilot’s “context‑aware” mode. Early adopters report a 22 % drop in hallucination rate for code‑completion tasks.Higher developer productivity; fewer costly bugs.Q3 2027Runway launches a video‑generation pipeline that feeds residual tokens to Stable‑Video‑2‑0, claiming 12 % higher frame‑consistency over 60‑second clips.Better visual continuity; reduced post‑processing.2028Academic groups extend REMORY to multimodal residuals (audio + text), enabling speech assistants to remember tone and prosody cues across hours of interaction.More natural, personalized voice agents; new UX paradigms.Beyond 2028ISO/IEC 23053 forms a working group to add “residual‑memory token” as a first‑class primitive in LLM API specifications.Standardised interoperability; faster ecosystem adoption.
If the community resolves the security and drift concerns, REMORY could become the de‑facto “memory layer” for any LLM that cannot be fine‑tuned on‑the‑fly. Its plug‑and‑play nature already lowers the barrier for adoption, and the early benchmark wins prove that the technique does more than shuffle numbers—it actually restores lost factual grounding.
Closing Thoughts
REMORY does not rewrite the laws of transformer scaling, but it redefines how we think about context. By teaching a frozen LLM to read a compact, learned “memory suffix,” the system sidesteps the brutal trade‑off between context length and latency. The open‑source SDK and low‑cost cloud service give developers a practical path forward, while the benchmark gains force a reevaluation of summarisation‑only pipelines.The next wave of LLM‑driven products will likely pair a summariser with a residual‑memory module, just as modern browsers pair a cache with a garbage collector. The real test will come when the industry pushes the token budget past 256 k and still expects a single prompt to retain month‑long reasoning. If REMORY’s early results hold, that future looks less like a memory‑crunching nightmare and more like a manageable engineering problem.
Stay tuned—next month we’ll dissect the multimodal extension that promises to remember not just words, but sounds and colors.#llm #ai #memorycompression #software #coding #development #engineering #inclusive #community
How 32 Tokens Can Replace 8K of Conversation – The Secret LLM Hack -
Hackers use cameras to track troops, military targets in Israel, Ukraine
Internet-connected security cameras and webcams around the world are increasingly being hacked and turned into military intelligence tools,…
#NewsBeep #News #BreakingNews #breakingnews #camera #cyberattack #hack #Internet #Iran #Israel #OperationEpicFury #OperationRoaringLion #Russia #Ukraine
https://www.newsbeep.com/766322/ -
Hackers use cameras to track troops, military targets in Israel, Ukraine
Internet-connected security cameras and webcams around the world are increasingly being hacked and turned into military intelligence tools,…
#Ukraine #UA #Europe #Europa #EU #News #Camera #Cyberattack #hack #internet #Iran #Israel #OperationEpicFury #OperationRoaringLion #Russia
https://www.europesays.com/ukraine/54683/ -
Hackers use cameras to track troops, military targets in Israel, Ukraine
Internet-connected security cameras and webcams around the world are increasingly being hacked and turned into military intelligence tools,…
#EuropeSays #Headlines #News #Europe #Europa #EU #Israel #camera #cyberattack #hack #Internet #Iran #OperationEpicFury #OperationRoaringLion #Russia #Ukraine
https://www.europesays.com/iran/324968/
