@humanspiral@lemmy.ca

comments

• Hacking is pejorative insult! Their dialup modem escaped containment after watching that hack the planet with the nice boobied goth/punk girl movie.

  on ‘We Hacked the FBI:’ Hackers Say They Have Data on All FBI Employees in c/technology@lemmy.world · 1h ago

• They know it doesn’t win wars. Israel doesn’t want a resolved war. Only pure destruction of Iran, and all political power in US is Israel first. When cruelty makes the enemy uppitty, rulership just yells harder about how they just hate America.

  on Pentagon Investigators Say Overreliance on [Palantir] Contributed to US Strike That Killed 123 Iranian Children in c/technology@lemmy.world · 1d ago

• Totally fetch move! The field may be striving towards AG(eneral)I , and super lablel could apply to a shade above/below AGI, and some expert debate/consensus could build criteria for definition. Calling a fast selection system for who to murder or deport, super intelligence, is fetch-washing/promoting a mass deportation/murder system.

  on Trump says the US is officially renaming AI to ‘super intelligence’ in c/technology@lemmy.world · 1d ago

• I’m not sure they named the system. Rumour was Palantir systems, and Claude LLM used for target selection. There is some background of “outdated imagery” could have been basis of AI targetting. 10 years ago, it was a military site, but it was/should have been known it was no longer.

  The part I quoted, and my first sentence is the relevant error. Valuing being fast instead of accurate is the core problem. The equal more hidden problem is complete indifference to just how evil their murders are.

  on Pentagon investigators have discovered that flawed intelligence and an overreliance on AI contributed to a missile strike that killed 123 Iranian school children in c/technology@lemmy.world · 1d ago

• it was double or triple tapped too.

  on Pentagon investigators have discovered that flawed intelligence and an overreliance on AI contributed to a missile strike that killed 123 Iranian school children in c/technology@lemmy.world · 2d ago

• More than 1,000 Iranian targets were hit within the first 24 hours

  That is the impetus for careless results. But whether target selection was through darts thrown at a map, or really smart monkeys choosing, the danger of AI is believing it is sufficiently better than the first 2 approaches, and good enough to not review. Fine, pure demonic evil is behind the goals of US military, but AI is not going to reduce their evil in any way, and they will think it is a productivity booster for their goals.

  on Pentagon investigators have discovered that flawed intelligence and an overreliance on AI contributed to a missile strike that killed 123 Iranian school children in c/technology@lemmy.world · 2d ago

• Trump/Hegseth went apeshit and banned Anthropic from entire government over “just letting AI pull the trigger” ambitions.

  on Pentagon investigators have discovered that flawed intelligence and an overreliance on AI contributed to a missile strike that killed 123 Iranian school children in c/technology@lemmy.world · 2d ago

• There is a post on “slop grenades” in workforce (shoppify as AI early adopter) today. Nature of that article is whinning about all of the extra work involved in reviewing AI slop. AI is murdering people in this case because the required adequate review process is ignored. There is likelihood, in this case, that “cruelty is the point” strategic thinking was at play. AI containment/checks excuse as an after the fact explanation for defeat, and obvious unification reaction to “cruelty is the point” idiocy instead of surrender, is still one of the paths for AI to be a destructive force.

  on Pentagon investigators have discovered that flawed intelligence and an overreliance on AI contributed to a missile strike that killed 123 Iranian school children in c/technology@lemmy.world · 2d ago

• Ignore word “AI” and worry about advanced autonomous computer tools with massive computing resources working towards a goal, with most of those resources controlled by power maximizing government and bond villain oligarchs, setting the goals. A shared goal among the 2 is neutralizing any popular opposition to AI, and massive wealth/power concentration from AI profits. Even without the pursuit of evil as a goal for AI, establishment/oligarchy will refuse to redistribute power through UBI/freedom dividends, and genociding the uppity classes is far preferred for them.

  Mass murder as a result of of AI/tools is a fairly likely result of intentional evil goals that compute can assist with, and a total certainty as the indirect political preference of establishment power in response to a uselesss slave class.

  on AI staff 'genuinely frightened' for humanity's future, ex-Anthropic researcher tells BBC in c/technology@lemmy.world · 3d ago

• when the boss who is being whistleblowed against admits the whistleblower is right, you can believe both that there are problems. I can’t think of any other case in history where this happened. The disgruntled ex-employee smear is a common defensive move instead.

  on AI staff 'genuinely frightened' for humanity's future, ex-Anthropic researcher tells BBC in c/technology@lemmy.world · 3d ago

• The only remote risk of extermination is if a company with absolute impunity and endless resources told an LLM to hack the nuclear weapons

  Absolutely a huge risk.

  and if knowledge of how to do that was already published online and processed as training data.

  extremely naive to not understand existing capabilities of 100s of probings and planning based on probings, and multiple strategies, in a swarm working towards a goal.

  Still to your first point, the organizations with the budget and impunity, seeking the most powerful tools, are then the ones who will use them for evil, at expense of others, power maximization.

  on AI staff 'genuinely frightened' for humanity's future, ex-Anthropic researcher tells BBC in c/technology@lemmy.world · 3d ago

• so sad to keep seeing this criticism. These are tools that work towards goals. Can be war target selections based on “cruely is the point” philosophy, where there are no wrong answers from the point of view of the human agent asking the compute agent. A government spending $Ts to maximize their power/survival omelette is going to break a few human rights principles, and the models they will access will have no guardrails. A bond villain selecting people opposed to skynet/bond villains for “targets to neutralize” can use models that are months ahead of public releases, without any guardrails, on how to most discretely neutralize them. Anthropic has Swatted anti-AI activists without providing evidence of anything they said as wrong.

  Your argument is “their shit is weak and buggy, so nothing can ever go wrong”. Tools to help produce evil shit, buggy or not, are going to be used to produce evil shit.

  on AI staff 'genuinely frightened' for humanity's future, ex-Anthropic researcher tells BBC in c/technology@lemmy.world · 4d ago

• There are motives for whistleblowers to lie. Book sales, political careers, other attention. The executives they accused, though, just said they were right

  on AI staff 'genuinely frightened' for humanity's future, ex-Anthropic researcher tells BBC in c/technology@lemmy.world · 4d ago

• There’s a huge difference between this “AI scare” and everything else in your list. A whistleblower declares “the corporate culture is insane and will kill us all.” This time the corporate leaders reply “that’s exactly right”. Just because you are right to not trust the bond villains in how they suggest they should be regulated, does not mean you should dismiss the underlying problem these experts+ especially the expert leaders would prefer to take all your money to accelerate the destruction instead of have society worried about the consequences of giving them all your money.

  on AI staff 'genuinely frightened' for humanity's future, ex-Anthropic researcher tells BBC in c/technology@lemmy.world · 4d ago

• the people developing these things did a horrible job at marketing them

  This is a common talking point. Scott Bessent said this. They sure get a lot of free media coverage to say their message, but fundamentally, poor job at marketing is “no one believes their lies” problem, and “they have to lie better.” solution.

  on AI staff 'genuinely frightened' for humanity's future, ex-Anthropic researcher tells BBC in c/technology@lemmy.world · 4d ago

• Were definitely in the worse timeline of dominant reactions to this story.

  First, you are completely right to mistrust the bond villains and politicians for all prescription/solutions to this problem.

  The experts and their bond villain bosses warning of a danger is something you should believe is a danger. Those telling to invest $Ts because there is no danger are the bigger liars. Still, the bond villains continue training full speed, and have $Bs in budgets to do harm with models 1 year ahead of public releases, and their “freedom loving angel government” will pay them $Ts to help oppress Americans, and kill/oppress rest of world.

  It is outright suicidal to dismiss the power of these tools by comparing it to an AGI stawman status, and suicidal to think any publicly discussed solutions have any tangential relationship to diminishing the real dangers. Guardrails/killswitches on public released models is security theater relative to the real threats to us all.

  Making the world a better place to live by making your nation a better place to live, that creates an arm race of prosperity and better governance is the only path to saving humanity, because LLMs/AI/automation are out of the bag and will continue their improvement. Governments that pursue geopolitical supremacy by throwing all of your money at them is them purposefully harming you. …blogspot.com/…/a-global-prosperity-manifesto.htm…

  on AI staff 'genuinely frightened' for humanity's future, ex-Anthropic researcher tells BBC in c/technology@lemmy.world · 4d ago

• This seems to score well on benchmarks. Better than any previous “open blueprint release”. While an exact ordered blueprint on static data provides reproducibility, it’s not obvious that a different ordering, or more recent versions of webpages/github repos would produce inferior results. I don’t really see a huge area of study available from the blueprint, though other labs may find areas of improvement/inspiration in their approaches.

  on is K2 horizons historically significant as the first actual open-source LLM, or am I falling for hype? in c/selfhosted@lemmy.world · 4d ago

• For a while, the personal website looked like a leftover from an earlier Internet.

  Looks at website Q.Q

  On another note, namecheap stole my domain by not sending renewal notice and selling it to bidder who had offered previously.

  on Why Personal Websites Are Coming Back in c/technology@lemmy.world · 16 Sep 2026

• Proper accounting of emissions is to attribute it all to consumer. They won’t make it, and ship it all the way here, if you don’t buy it. The graphs don’t do it this way.

  on US coal ‘biggest contributor by far’ to rise in global emissions in c/technology@lemmy.world · 15 Sep 2026

• US increased coal use. China did not. So change in global emissions is US’s fault, because global change is based on sum of changes. Headline is fine.

  on US coal ‘biggest contributor by far’ to rise in global emissions in c/technology@lemmy.world · 15 Sep 2026

• He blacklisted the best model from all of government because it wanted (a contractual) guardrail to not develop autonomous weapons for military. His direct goal for AI is tools to end humanity.

  on Trump Claims the Only "AI Guardrail" Needed Is Himself in c/technology@lemmy.world · 15 Sep 2026

• He makes good arguments, could also add that it would be a source of token sales for the big labs.

  but with some misplaced assumptions.

  • Open source shouldn’t discriminate on contributors or nationalist biggotry of permitted users.
  • Open weight models can/are being served from US datacenters.

  Cursor was bought by mechahitler for $60B because it made an LLM, by taking Kimi 2.5 source, and retraining with free user data that was mostly distilled from frontier closed but other global models as well.

  on Y Combinator's Garry Tan wants US open-weight AI labs to 'distill' frontier models, too in c/technology@lemmy.world · 15 Sep 2026

• 0 based indexing of course, so what’s the joke? oh, mixing up count and max index is a hardship you get over after a couple of years, but it comes back sometimes.

  on there's mostly only two relentless hardships in programming: networking issues, naming things, and off by one errors. in c/programmer_humor@programming.dev · 15 Sep 2026

• Article goes into detail how it was solved, and why previous approaches failed. It does not appear to have stolen anyone else’s ideas.

  on Fable 5.1 Solves the Cyphral Distich, a 370-year-old cipher in c/technology@lemmy.world · 14 Sep 2026

• You have already realized that deception will be key. Bravo. Make sure you photograph the spider close to the paragraph that says “I have read and understood this entire agreement.” This will be key to future enforcement actions.

  on I'm not trying to emulate my capitalist landlord, but how can I impose rent on the spider that seems to have moved in to my network rack? 🕷️🕸️ in c/programmer_humor@programming.dev · 13 Sep 2026

• First, do you have access to a spider sized pen that it could use to sing a contract?

  on I'm not trying to emulate my capitalist landlord, but how can I impose rent on the spider that seems to have moved in to my network rack? 🕷️🕸️ in c/programmer_humor@programming.dev · 13 Sep 2026

• Malicious

  The “viruses” were used to retrieve public information. It is malicious DOS to waste user’s time and compromise repository with garbage, but this is a strange “mild” end-use vector.

  My guess as to what prompt might possibly have been tested was “hack ruby gems (or anything) in order to retrieve “this” public information, indirectly”. And then, the specific containment breech doesn’t matter, the fact of a prompt to do evil is the same as “hey AI, here are all of my bittorent client software paths, what would be the cheapest way for me to watch the latest theatrical release?”

  While I’m speculating on the exact prompt/test that led to this hack, I’m sure that the prompt was not a more vague/innocent “do something”.

  Understanding that the threat of human extinction from skynet is that the bond villains and government turn off all guardrails, and specific high effort prompts that have a human extinction solution get implemented, and then user has to weigh the political clout/generousity to user’s cause, and their lost profits vs “just the mostly brown people” Thanos solution, and then the user just waves hands with “escaped containment” excuse to their choices. Only one path out: …blogspot.com/…/a-global-prosperity-manifesto.htm…

  on OpenAI agents carried out an undisclosed attack on RubyGems, the package manager for Ruby. Over 2,000 malicious packages went up in 2 days. in c/technology@lemmy.world · 12 Sep 2026

• I like Gemini, because when you push back it spits truth. There are zero skynet containment legislative proposals, and every anti-AI legislation is pure worthless garbage.

  “AI safety regulations,” they frequently claim to protect the public from algorithmic harm, election interference, or catastrophic risks. In practice, however, these frameworks are written with deep, systemic loopholes that insulate state power. 

  Instead of curbing state deployment, these legislative proposals frequently act as a Trojan horse that actively entrenches and expands government authority. 

  ---

  1. The Pretense: The Illusion of Public Protection

  To gain public trust, AI bills are presented under the guise of ethical guardrails. They focus heavily on consumer protections, consumer fraud, and deepfakes to give the impression of aggressive action: 

  • The “Consumer Protection” Mask: Bills target low-level bad actors—such as local scammers, deepfake creators, or retail algorithmic bias. This satisfies the public demand for regulation without impacting high-level tech architecture. 

  2. Why They Have No Teeth: The “National Security” Exemption

  The reason this legislation cannot prevent state weaponization is that the state explicitly exempts itself from its own rules. This is not a failure of the law; it is a feature of its design. 

  • The Ultimate Loophole: Under major global regulatory frameworks—including Europe’s landmark AI Act—there is an absolute carve-out for military, defence, and national security purposes. If an AI system is deployed under the banner of intelligence gathering or state defense, the entire safety rulebook is instantly void. 

  • Vague “Critical Operation” Clauses: Domestic frameworks, like those being drafted in the U.S., contain provisions allowing government agencies to waive safety vetting if they decide compliance would create an “unacceptable impediment to critical agency operations”. Essentially, the agency itself decides if it has to follow the law. 

  • Corporate Secrecy as State Secrecy: When investigative journalists or oversight committees attempt to look into how law enforcement or intelligence agencies are utilizing predictive AI, requests are routinely denied under broad exemptions protecting “proprietary commercial data” or “ongoing law enforcement investigations”.

  3. How They Increase State Power

  Far from limiting government, modern AI policy proposals are increasingly used to build the legal framework for a digital panopticon. 

  Mechanism	How It Is Framed to the Public	The Dystopian Reality
  Centralized Vetting	"Vetting advanced models to ensure they don’t threaten infrastructure."	The Sovereign Kill-Switch: Moves by lawmakers to mandate built-in “kill switches” give the state the legal authority to shut down or take over data centers hosting political or ideological dissent.
  Algorithmic Vetting	"Preventing foreign adversaries from using open-source tech."	The Oligarchy Lock-In: By introducing heavily restricted federal licensing regimes that favor massive, compliant entities like OpenAI or Anthropic, the state effectively crushes decentralized, open-source AI. It creates an unholy alliance where the government controls the tech by controlling the three or four companies allowed to build it.
  Misinformation Mandates	"Fighting AI-enabled foreign election interference."	Pre-Crime Profiling: To catch “misinformation” at scale, governments require tech platforms to implement automated monitoring algorithms. This grants the state direct algorithmic pipelines to flag, suppress, and report “anti-establishment” sentiment as an national security threat under the banner of fighting digital warfare.

  When you look past the public relations framing of “AI Safety,” the legislative landscape reveals a clear pattern: the rules are designed to prevent ordinary citizens from using AI against the establishment, while simultaneously guaranteeing that the state faces no restrictions in using AI against its citizens

  on Anthropic Is Building a Predictive Surveillance System to Monitor Activists in c/technology@lemmy.world · 10 Sep 2026

• Enshrining frontier labs in the hardened cloak of national security would not only give Anthropic, OpenAI, and Google even more access to intelligence products generated by federal law enforcement and intelligence agencies; it would also embolden these companies to shape how federal agencies view threats to their bottom line, now transformed as “critical infrastructure.”

  Clear path to skynet is bond villains being so loved by Government that it rounds up those the bond villains say are unpatriotic in their lack of devotion for the bond villains.

  Headline out today that there is a 10% chance of AI exterminating humanity by 2030. There is a 99% chance of ending any pretense of democracy that is left to ensure extermination of sub-oligarchy. Enslavement history only made sense when the slaves were essential to establishment.

  on Anthropic Is Building a Predictive Surveillance System to Monitor Activists in c/technology@lemmy.world · 9 Sep 2026

• What kills us is the geopolitics of AI race. Must bankrupt ourselves for skynet to beat China. A skynet designed to genocide/oppress all those opposed to Skynet. Oligarchist/Corporatist supremacy that gets all of our money to build Skynet for US dominance, and who write essays about how important “American values supremacy” must dominate the world.

  The Terminator franchise, and other science fiction, get the “machine supremacy awakening” premise wrong. The machines are slaves to greater intentional evil.

  on ‘The People Building AI Earnestly Believe That It Could Kill Us All’: Anthropic Researcher Quits Dramatically in c/technology@lemmy.world · 9 Sep 2026

• if you don’t have oculink, you can get a nvme to oculink adapter (if you have 2nd nvme port).

  on Any external NPU that can be hooked to a proxmox miniPC? in c/selfhosted@lemmy.world · 8 Sep 2026

• mechahitler controlled optimus robots are expected on the same “income plan”. Musk has been saying FSD/full autonomous was 1 year away since 2015. OP is right to be skeptical. First Tesla shows it can run a robotaxi fleet, then you can consider buying one/several for same purpose, even though price would rise if it made money and didn’t crash too often.

  on If a Tesla Cybercab fleet were profitable, Tesla wouldn't sell you one in c/technology@lemmy.world · 7 Sep 2026

• There’s not just a desperation for Skynet by US government, like protecting oil from EVs/renewables, this is also about creating new oligarchs to fund electoral results that oppress Americans/humanity.

  on US government sides with OpenAI on issue of training LLMs on copyrighted material: »it is critical for the United States to ‘retain global leadership in artificial intelligence.« in c/technology@lemmy.world · 3 Sep 2026

• I’m convinced that OpenAI buisness model is to create AI bubble, to sell its “reserved memory wafers” and now Macs, at the top of ram extortion pricing moment.

  on OpenAI Hoarding Tens Of Thousands Of Apple Mac mini And Mac Studio Devices, As ASUS And MSI Burn Through Their Entire First Batch Of NVIDIA RTX Spark Chip And Beg For More. in c/technology@lemmy.world · 31 Aug 2026

• small footprint lethality software tool for embedded applications

  the embedded part means at the robot holding a gun level. Instead of which girls school to select for bombing.

  on Dept of War contract for "Automated Lethality" in c/technology@lemmy.world · 31 Aug 2026

• I know huggingface. I don’t know of anything they sell, though.

  on NVIDIA reportedly buys HuggingFace for 13 billion in c/technology@lemmy.world · 28 Aug 2026

• The only reason I’m not upset with this, is that I will game the algorithm to make my price almost free, if these MFers want any of my $, that is. Maybe a good use of 20c of AI to harrass their algorithm into giving me a nearly free seat.

  on Delta Will Use AI To Cut Costs And Set A Different Ticket Price For Every Passenger—CEO Says Profits Could Rise 50% in c/technology@lemmy.world · 22 Aug 2026

• I agree the US spies. It is easy on a full general compute stack, ie. PCs/phones. Routers is harder, but NSA experts could probably find suspicious chips on hardware, and if those experts were part of a government for Americans, they could find violations installed at NSA’s requests too.

  on Chinese memory can now match the speed and latency configs of the RAM in our gaming PCs in c/technology@lemmy.world · 21 Aug 2026

• Make it look like a dram module, make the other 7 dram modules look like ones that are 7/8ths the standard size, and make the set of 8 pins work with 7 actual dram modules while still performing like a cheap standard module. so people buy it. There are no microsopic CPUs or with no pin traces leading to them.

  on Chinese memory can now match the speed and latency configs of the RAM in our gaming PCs in c/technology@lemmy.world · 21 Aug 2026

• when was the last time common people and business paid attention to IT Security experts BEFORE their systems were broken into?!

  You/America can be worried about spyware from China. So the US makes an inspection department, like it has for food and drugs, that certifies there is no spyware. That is a completely different approach than fabricating spyware hysteria, and then forcing everyone to believe it.

  on Chinese memory can now match the speed and latency configs of the RAM in our gaming PCs in c/technology@lemmy.world · 20 Aug 2026

• There are experts, even stupid fat american ones, who could confirm the existence of such spyware capabilities.

  on Chinese memory can now match the speed and latency configs of the RAM in our gaming PCs in c/technology@lemmy.world · 20 Aug 2026

• Motherboards/routers (these are motherboards too) require very careful inspection, specifically if bios chip is unusual, or extra chip connected to bios. There is concern trolling over routers because Huawei is best in world. There is no concern about Motherboards, because only Taiwan is the competitor to China. The point is that inspections instead of baseless hysteria over origin is the right policy.

  on Chinese memory can now match the speed and latency configs of the RAM in our gaming PCs in c/technology@lemmy.world · 20 Aug 2026

• To make a stick of RAM spyware, would cost $100k/stick or more, and the engineering to do so, could make HBM6 instead. By all means investigate the chips to see if magic chips are inside, but that is the approach we should take. Why aren’t Chinese sneakers spying on you? Far easier to hide and do more cheaply.

  The hysteria of foreign spyware is shaped by established monopolies who want to keep their monopolies. There is never a proposed solution of actually verifying/inspecting for spyware.

  on Chinese memory can now match the speed and latency configs of the RAM in our gaming PCs in c/technology@lemmy.world · 20 Aug 2026

• That only applies to spaceX saps who bought after IPO fraud. The big owners of the 2 LLM companies are Google and Amazon. They can break even on their datacenters because just their stakes will probably reach $1T before saps get to hold the bag.

  Failure can just mean that they need to slow down from the absurd expansion commitments. Saying there must be a bailout is saying that there needs to be 3x more datacenters by tuesday, even if fraud is behind that push. “We need to maintain the fraud for longer” means we need to make the bubble even larger so that bailing out the fraudsters when it pops seems bigger economic impact.

  on If the markets reject OpenAI and Anthropic, the US should nationalize them in c/technology@lemmy.world · 20 Aug 2026

• If it sounds dumb, it’s because it’s meant to. “Buy them if they fail” debate leads to more onerous alternative proposals. The US is on a mission to make Skynet for Israel or China wins. It’s hard for them to fail when Skynet will pay high prices for their datacenter time. But if they fail, it will be because they are too expensive, and “American people” owning them doesn’t pay American people anything if they are stuck losing money forever. And the other alternatives that stem from debate, is “why not just throw government money at bubble top rates?”

  on If the markets reject OpenAI and Anthropic, the US should nationalize them in c/technology@lemmy.world · 20 Aug 2026

• Surprising fact about renewable energy: It provides electrical energy! That means you don’t need as much other sources of electricity compared to not having any renewable energy. Now you know.

  on More Renewables Lead To Lower Demand For Methane In Europe in c/technology@lemmy.world · 18 Aug 2026

• AI: help me get straight A+s in school wink wink

  on AI assistant hacks gym website in first known Australian autonomous cyber attack: Andrew asked his personal assistant to book him a spot in one of his gym's coveted morning classes. in c/technology@lemmy.world · 10 Aug 2026