Just like the classic Notepad Ctrl+Click RCE behavior, terminals... on Windows, Linux, and macOS also support clickable file/URI handlers. printf "\x1b]8;;file:///C:/windows/system32/calc.exe\x07Click here\x1b]8;;\x1b\\n"show more

PewDiePie just hit 20K GitHub stars in under 24... hours. The project? Odysseus. A self-hosted AI workspace that runs 100% on your machine. • Agents with tools • MCP built in • Persistent memory • File handling • Windows, macOS, Linux Your data never leaves your device. It supports Ollama, llama.cpp, and vLLM locally with OpenAI and OpenRouter support if you want cloud models too. The crazy part? A YouTuber with 110M+ subscribers just out-shipped most AI startups. And he built half of it using AI.show more

Stop...Your AI shouldn’t sit in a sidebar. It should... actually work inside your notes. AnySlate just made that real. Your AI can: • read your markdown files • edit sections • create docs • search your workspace • organize everything No copy paste No switching tabs No proprietary lock-in Every file is a real .md file Your notes stay portable forever And the wild part — it supports MCP natively So Claude, Cursor, or Windsurf can directly work inside your workspace This is the shift from AI that helps you write to AI that manages your knowledge Real-time collaboration Built-in AI assistant PDF + web publishing Version history Works on Mac Windows Linux + browser This is what a modern markdown editor should look likeshow more

Codex CLI Update: Let there be Search Whatup nerds,... back so soon looking or yet ANOTHER update?! I got you. Update 0.121.0 is here! > You can now search through previous user prompts with CTRL+R. Just trigger search and enter your search string, you can easily arrow through all matches. See video below! > 🥔 Support for Spud! Is not here yet. Sorry. Maybe tomorrow. 🫢 > v0.121.0 adds custom marketplace installs in Codex: you can run codex marketplace add to register marketplaces from GitHub shorthand, git URLs, or local directories. Codex validates the marketplace layout and stores it in your user config so it shows up consistently in plugin discovery. > Improved memory features, including a new /memories TUI menu with use/generate toggles, a reset-all-memories action, app-server support for setting thread memory mode and clearing memories, and cleanup of stale memory-extension resources. Note: Not working on Linux for me. /memories command unavailable. > Codex MCP got further upgrades with direct app tool calls, cleaner namespacing, and safe optional parallel execution for faster workflows > Codex realtime got better controls (text/audio + clear “done” signals), easier history syncing, and safer file handling. > Hardened devcontainer setup plus smarter macOS socket allowlists for safer local runtime access. > Dozens of other bug fixes, see repo below. Toodles! ✌️show more

Run Gemma 4 26b MTP on 8 GB VRAM... GPUs at 25+ tokens/second. Flags included! local llm space is moving at terminal velocity. only 3 days ago google released gemma 4 26b a4b qat quants. more efficient than before, ran on 8gb vram at 20 tok/sec. and now just a few hours ago, mainline llama.cpp merged a massive update and we just shattered our own record. decode throughput went 25-40% up on the same 8 GB VRAM setup! Before MTP: 20 tps -> After MTP: 28 tps! llama.cpp just officially merged PR #23398 ("add Gemma4 MTP"), bringing native Multi-Token Prediction (MTP) support to Gemma 4 models. By running speculative drafting on the same 8GB VRAM RTX 4060 setup, my decode throughput on a 64k context instantly leaped to a blistering 25–27 tokens/sec thats 25-30% increase with the same hardware. Here is the architectural catch you need to know: Unlike the Qwen 3.5 and 3.6 series, which bake the MTP heads directly into the base GGUF, the Gemma 4 MTP head is not built in. You must download a separate, specialized MTP drafter GGUF (the assistant model) to act as the speculator. (I've dropped the download link in the replies). copy and try the exact flags: -m gemma-4-26B-A4B-it-qat-UD-Q4_K_XL.gguf --spec-type draft-mtp --spec-draft-n-max 6 --spec-draft-p-min 0.7 --spec-draft-model gemma-4-26b-A4B-it-assistant-Q4_0.gguf -c 64000 -v n-max 4 and p-min 0.7 is also worth checking out. benchmark on your setup and workflow. if you have a single 8 gb vram nvidia rtx 4060, 3060, 3070, 2080, 2070, grab the MTP drafter GGUF link in the comments and try it yourself. Check it out even if you have asmaller or a larger gpu, such as a single rtx 3090, 4090, 3060, 2060. MTP works for all gemma 4 sizes such as gemma 4 12b, gemma 4 31b etc. but remember to grab the correct mtp draft assistant models respectively. what are you benchmarking todayshow more

1️⃣ I did everything I could for Ritual I... made all kinds of contributions. For 8 months, I contributed to Ritual, but I am extremely disappointed with their system. I made 181 contributions on X, bought Canva Premium to design infographics, and sent almost 17,000 messages. Despite all this, I have been permanently ignored for 8 months. Two of my friends, Ashish.Base.eth (❖,❖) (230 contributions) and KUNDAN (150 contributions), are facing the same treatment. Because of this unfair and favoritism based system, I decided to post on Twitter to expose it. When we raised our voice after 6 months of hard work and 150+ contributions, the so called Indian community handlers and famous team members labeled it as FUD and continued ignoring us. 2️⃣ When we asked the Indian community handlers why we were not given roles after 8 months, they said: “We can’t do anything, the team decides.” But when we contacted the team or mods, they said: “Go back to the Indian community handlers, they will help you.” If these handlers and big mods really can’t do anything, then why are they even here? They are here to build connections This system has been making fun of me and other hard working contributors. 3️⃣ Now look at the people who received roles based on “quality contributions.” Everything is available with proof. One person (Yaneul) got a role with only 15 very low quality contributions (mostly food and event screenshots). He even got Ritualist with just 25 contributions. This clearly shows either connections were used or it could be a mod’s second account. Another person (placboeffect) got a role with just 9 contributions. Many people received roles at 35, 40, 50, 55, 60 contributions, while I had 181 contributions. This is what they call “quality contributions.” 4️⃣ Now decide yourself: Who contributed more? Who worked harder? People contributing for 8 months to 1 year are ignored, while others get roles at 9 or 25 contributions. Yes, we may have wasted our time, but we did work hard. 5️⃣ I didn’t want to come to X. but after contributing like a mad person and getting ignored, I don’t want to waste my time anymore. 6️⃣ I don’t know if the Josh (❖,❖) admin is aware of how mods and team are damaging this project Claire (❖,❖) she just help her country mates doesn't care about other people. 7️⃣ What’s the worst that can happen? They will ban me, and my 8 months of contributions will be wasted. There was never respect for real contributors anyway. If I had good connections with mods, I wouldn’t have struggled this much. It’s not like Ritual is my life, or Web3 has only Ritual, or they gave me any role to take back. Now I will openly speak on X without fear. 8️⃣ Some fearful people will read this post. They know everything that’s happening, but they won’t comment because they are scared. I understand their fear. --- 9️⃣ Best of luck to those who keep working blindly despite seeing all this. And special luck to those who have roles because of good connections with mods. --- 🔟 Last The so called Indian community handlers are only mods by name. One of them is a paid event manager, doing the job only for money. Another handler sends one message after 10 days. These are the people leading us in Ritual, who don’t even know what they’re doing. Jez ritual/acc (❖,❖) dunken(ritual/acc) (❖,❖) Stefan | Mad Scientist (❖,❖) Hinata Sir W A R D E N you are mod in Donut and Dhillon saab π² you are mod at data haven, Choudhary (Ø,G) ꧁IP꧂ you are mod at Capx and sir Botsan (capx arc) you are Admin of Capx You should learn from ritual mods how to ignore real contributors and how to promote favourtismshow more

Another WTF moment. A developer just open-sourced a coding... agent harness that boots 245x faster than Claude Code. It's called jcode. You launch it and the first frame renders in 14 milliseconds. Claude Code takes 3,436. One active session uses 27.8 MB of RAM. Claude Code uses 386.6. Run ten sessions in parallel and jcode holds at 117 MB while OpenCode swells to 3.2 GB. Each agent has a semantic memory graph instead of a scratchpad. Every turn gets embedded as a vector. The graph is queried on every turn for related memories, and a sideagent verifies the hits before injecting them into context. Consolidation runs in the background to check for stale or conflicting facts. No manual /remember calls. No token burn on lookup tools. The provider list is 30+ deep. Claude, ChatGPT, Gemini, GitHub Copilot, Azure, OpenRouter, DeepSeek, Groq, Mistral, Perplexity, Fireworks, Ollama, LM Studio, and any OpenAI-compatible endpoint you point it at. Ran out of tokens on your first ChatGPT Pro sub? /account swaps to the second. Then there's Swarm. Spawn two agents in the same repo and the server manages them. When agent A edits a file agent B has been reading, agent B gets pinged and can check the diff. Agents can DM each other, broadcast to the room, or spawn their own worker teams for parallel tasks. Groups, channels, and completion statuses are handled automatically. The UI has live side panels that render mermaid diagrams inline. To make it fast, the author wrote a Rust mermaid renderer 1800x faster than the JavaScript one, then wrote a custom terminal called Handterm because no existing terminal could do smooth partial-line scrolling. Self-dev mode is where it gets wild. Tell your agent to enter self-dev and it starts editing jcode's own source code, rebuilds the binary, reloads it live, and keeps working across your existing sessions. You can also resume broken sessions from Claude Code, Codex, OpenCode, or pi directly inside jcode. Anthropic's cache goes cold at the 5-minute mark and you're staring down a big cache miss on your next turn? The UI warns you before you spend the tokens. Written in Rust. MIT licensed. Runs on macOS, Windows, Linux, and Termux. Sitting at 11.2k stars with a native iOS app coming.show more

I told you to claim your free 16GB NVIDIA... GPU for learning Local LLMs. Now I’m going to show you how to double its inference speed without touching the hardware. Google Colab gives you an enterprise grade NVIDIA Tesla T4 GPU for free, roughly 4 hours every single day. It is the absolute perfect sandbox for learning AI engineering, testing inference flags, and pushing massive context windows. The local AI timeline is moving way too fast. If you aren't using Multi Token Prediction (MTP) yet, you are leaving massive performance on the table. I just pushed DeepMind’s Gemma 4 26B to 64.9 t/s on this exact free tier. Let's look at the raw benchmark data running on an Ubuntu Linux environment with the latest compiled llama.cpp binaries and quantized GGUFs from Unsloth via HuggingFace: # Qwen 3.5 9B (Dense): Base: [ Prompt: 626.7 t/s | Generation: 21.0 t/s ] With MTP: [ Prompt: 539.1 t/s | Generation: 24.8 t/s ] # Gemma 4 26B QAT (MoE): Base: [ Prompt: 634.2 t/s | Generation: 48.3 t/s ] With MTP: [ Prompt: 572.1 t/s | Generation: 64.9 t/s ] If you are paying attention, this single Colab notebook reveals 3 massive observations about the current state of local LLMs: # 1. The MTP Speedup (Software Overclocking) Standard autoregressive decoding guesses one token at a time. MTP acts like a highly optimized, built in speculative decoder. It predicts multiple future tokens at once and the main model verifies them in parallel. The result? Zero accuracy loss and a massive throughput increase. Gemma jumped from 48 to 65 t/s just by flipping a flag. # 2. The MoE Paradox (Bigger is Faster) How does a 26B parameter model absolutely destroy a 9B model in raw speed on the exact same hardware? Architecture. Qwen 3.5 9B is a dense model. it activates all 9 billion parameters for every single token. Gemma 4 26B is a Mixture of Experts (MoE) model. It routes data efficiently, activating only 4B parameters per token. You get the reasoning capabilities of a 26B model with the compute cost of a 4B model. 3. Thinking Efficiency When I ran the exact same complex prompt on both models, the larger MoE spent significantly fewer "thinking" tokens to arrive at the correct answer. A smarter model doesn't just give better answers; it gets to the point faster, saving you compute cycles and preserving your context window. # Want to run this yourself? Here are the exact llama.cpp CLI commands. For Qwen (MTP is baked into the main model): ./llama-cli -m Qwen3.5-9B-UD-Q4_K_XL.gguf -p "Explain quantum computing." -n 2000 -c 8000 -ngl 99 -fa on --spec-type draft-mtp --spec-draft-n-max 4 --spec-draft-p-min 0.7 For Gemma (Using a separate lightweight draft model): ./llama-cli -m gemma-4-26B-A4B-it-qat-UD-Q4_K_XL.gguf --model-draft mtp-gemma-4-26B-A4B-it.gguf -p "Explain quantum computing." -n 2000 -c 8000 -ngl 99 -fa on --spec-type draft-mtp --spec-draft-n-max 4 --spec-draft-p-min 0.7 Stop waiting for a $3,000 rig. Boot up Colab, pull these models, and start building your stack. I’ve put together a completely free, cell by cell Google Colab notebook that automates this entire workflow so you can test it yourself in 5 minutes and learn. Link to the notebook is in the comments below. Experiemt with different MTP parameters, context windows and post your results in the comments.show more

AgentLinter is here! Is your agent sharp & secure?... I built AgentLinter, a linter for and agent config files. Here's why. Whether you're vibe-coding or agent-coding, your AI's output quality comes down to one thing: how well you wrote your But managing these files properly? Way harder than it looks. 🎯 The Silent Failure Problem Vague instructions like "write good code" let the agent interpret however it wants. Output gets inconsistent, but nothing throws an error. The failure is silent. Anthropic's own docs say write "Use 2-space indentation" not "Format code properly." But as the file grows, spotting these with your eyes alone is nearly impossible. 🔐 The Security Problem People hard-code API keys and tokens directly into or and commit them, way more often than you'd think. AgentLinter stats show 1 in 5 workspaces has exposed credentials. .gitignore doesn't catch secrets buried inside markdown files. 💥 The Consistency Problem Multiple config files = contradictions. says "be a friendly assistant," says "concise, direct tone." The agent gets confused. references files that don't exist. Past 5 files, these conflicts triple. So I thought: is code. Code has ESLint. Why doesn't this have a linter? 🔍 What AgentLinter Does It diagnoses your agent config across 8 categories: 1) Structure: file organization 2) Clarity: instruction specificity 3) Completeness: missing definitions 4) Security: exposed secrets 5) Consistency: cross-file contradictions 6) Memory: session handoff 7) Runtime Config: gateway/auth settings 8) Skill Safety: dangerous shell commands & injection patterns Each scored 0–100 with concrete fix suggestions. Write "be helpful" and it tells you to specify response length, tone, and format. Find an API key? Instant CRITICAL alert to rotate. 🔒 Privacy-First & 100% Local Everything runs on your machine. Files never leave. Only the results are shared, and you can turn that off in settings. This matters — these files can contain system prompts, security rules, and personal context. Fully open source, MIT license, 100% free. 🛠️ Multi-Tool Support Works with Claude Code, Cursor, Windsurf, and Clawdbot. Detects for project mode, or clawdbot.json for agent mode and adjusts diagnostics automatically. 🚀 Get Started with one line npx agentlinter Node.js 18+, no config needed. Run it, check your score, fix what needs fixing. Happy vibe-coding & happy agent life! 🤙 Website: Github:show more

Release: LichtFeld Studio v0.5.3 is out! With 316 commits... merged into master, this release is a huge step forward for LichtFeld Studio. What's new in v0.5.3 • Vulkan viewer/rendering migration: New Vulkan viewport pipeline, pass graph, VkSplat renderer, Vulkan point-cloud renderer, 3DGUT/VkSplat support, improved alpha/depth composition, tighter CUDA/Vulkan interoperability, and device matching on multi-GPU systems. • RAD + LOD workflow: Added RAD file export/import, RAD LOD viewer, Spark-style GPU LOD selection, GPU-driven page prefetching, a bounded VRAM pool, out-of-core PLY-to-RAD LOD conversion, and RAD import/export speedups of approximately 3–5×. • HiGS / macro-tile inference: Added a macro-tile inference path for the Vulkan viewer, including macro sorting, batched rasterization, composition, and capacity management. • Asset Manager: Added and significantly enhanced the Asset Manager with thumbnails, SH information, faster synchronization, import-from-URL support, docked mode, data-loading popup integration, and general UI cleanup. • Viewport export: Integrated viewport export directly into the application as a toolbar/overlay tool, added fast render_view_u8-style readback paths, fixed high-resolution clipping issues, improved orthographic export parity, resolved 32K image/video export problems, and added post-export GPU resource cleanup. • Selection and tooling: Added and reworked selection toolbar controls, the Select menu, ring selection, color eyedropper, distance-from-center selection, faster point-cloud and zoomed-out selection paths, Vulkan measurement tool fixes, and drag-and-drop scene graph improvements. • UI/RmlUi platform work: Major RmlUi redesign efforts, hot reloading for RML/RCSS/Python UI files, reactive UI/store integration, viewport toolbar flyouts, improved histogram interactions, input settings enhancements, custom TRS gizmos, and numerous panel, tooltip, and localization fixes. • Windowing and UX: Added borderless window support, title bar drag/maximize/restore behavior, work-area-aware maximize functionality, resize responsiveness and performance improvements, and DPI/UI scaling fixes. • Training and data features: Added adaptive depth loss and depth gradients for the EWA rasterizer, mask loading/application fixes, a new combined Ignore+Segment mask mode, --add-splat, --freeze, improved checkpoint and training state handling, and training speed and VRAM optimizations. • COLMAP/equirectangular support: Added SPHERICAL/equirectangular camera model support and canonical EQUIRECTANGULAR handling, along with fixes for undistortion and camera export. This release will be available to all supporters as a Windows binary via approximately in about an hour. At the same time, LichtFeld Studio remains committed to being free and open source under GPLv3 and can also be built directly from source. Please consider supporting the ongoing development of LichtFeld Studio through a donation via the portal or the supporters page. Thank you to everyone who supports this project financially, contributes code, reports bugs, provides datasets, helps with the website, and contributes in countless other ways. A special thank you to our foundational sponsor Core11 and our Gold Sponsor Volinga, whose support has helped make the current state of the software possible. Thank you as well to every donor and to all of our new Bronze Sponsors. Looking ahead to v0.6 For the next major release, work will focus primarily on stability and user experience. This includes improved cleanup workflows and the ability to modify training parameters while training is in progress. I would also like to introduce a native .licht project format that allows users to save and restore their complete editor state. You can find links to our main sponsors below. Please also visit our website to discover all our Bronze Sponsors. Hint: We do not yet have a Silver Sponsor or Platinum 😉show more

And here it is, 🚨🚨Documents obtained by investigative journalist... Yehuda Miller through the Freedom of Information Act, CONFIRMS, BEYOND THE SHADOW OF A DOUBT, The Department of Justice and the FBI uncovered a massive 2020 ballot fraud operation based in Michigan and operated in multiple swing states including Washington DC, Chicago, IL, Georgia, Iowa, South Carolina, Pennsylvania, and Florida, funded by Joe Biden’s 2020 presidential campaign. GBI strategies BUSTED for submitting fraudulent voter registrations during 2020 election cycle. Following a raid, Michigan law enforcement discovered caches of pre-paid gift cards, semi-automatic rifles with silencers, four modified pistols with ammunition inside. and disposable burner phones. Throughout the 2020 election period, these Democratic cartel election committees provided more than $4,000,000 to this criminal organization: Biden for President: $450,000 Democratic Senatorial Campaign: $2,117,605 DNC Services Corp: $1,031,856 Democratic Party of Iowa: $493,100 The investigation was initiated following the observation of a Muskegon, Michigan, clerk who noticed an individual depositing 8,000 to 10,000 completed voter registration applications at the city office on October 8, 2020. This same individual returned multiple times, registering an additional 2,500 voters. Thousands of these registrations displayed identical handwriting with fraudulent addresses and phone numbers. Additionally signatures did not match those on file with the criminal Secretary of State Jocelyn Benson. 🚨🚨And the icing on the cake. + 100,000 ballots cast in the Nov 3, 2020 election were also falsified at Jocelyn Benson’s own election headquarters. FOIA: “I am following up with some information on one of the most egregious situations that I encountered. There are many examples of significant election integrity issues, but this example involves actions of GOVERNMENT OFFICIALS from both the local and state levels that are shocking.” Dawn N. Ison Department of Justice Assistant U.S. Attorney | public corruption unit, “We received this complaint. Guess what the criminals do, when the FBI and the DOJ got hold of this investigation? Passed it right back to this criminal who were running it, 📝who the fuck was she planning to shoot?show more

Top 25 AWS services explained EC2 – Your server.... But in the cloud. You pay even when it’s doing nothing. (Sound familiar?) Lambda – EC2’s lazy cousin. Only wakes up when there’s work. No work, no bill. ECS – “I want containers but Kubernetes gives me anxiety.” EKS – Kubernetes. For people who enjoy suffering professionally. Auto Scaling – Your app gets famous overnight. This makes sure it doesn’t die from the attention. S3 – A bucket that never fills up. Jeff Bezos’s gift to humanity. EBS – A hard drive for your EC2. Loyal. But only to one instance at a time. EFS – EBS but for people who like sharing. Multiple instances, one file system. FSx – EFS but for enterprises who need Windows compatibility and a bigger invoice. Snowball – When your internet is too slow to upload data, AWS ships you a literal box. VPC – Your private neighborhood inside AWS. Strangers not allowed. Route 53 – The GPS of your app. Tells traffic where to go. ELB – The bouncer at the club. Splits traffic so no one server gets overwhelmed. CloudFront – Your content, cached globally. Because nobody likes a slow website. Direct Connect – A private highway between your office and AWS. No public internet drama. RDS – A managed database. AWS handles backups so you can sleep at night. DynamoDB – NoSQL at insane speed. Schema? We don’t do that here. Aurora – RDS on steroids. Faster, smarter, slightly more expensive. Redshift – A warehouse for your data. Not clothes. Petabytes of analytics data. ElastiCache – RAM for your app. Because hitting the database every time is embarrassing. IAM – The bouncer for your entire AWS account. Get this wrong and you’re headlines. KMS – Locks your secrets in a vault. AWS holds the key. You trust them. Mostly. Cognito – “Login with Google” but you built it on AWS. GuardDuty – The security camera that never blinks. Watches for sketchy behavior 24/7. WAF – Stops hackers at the door before they touch your app. Bookmark it.show more

my 8 GB VRAM gaming laptop is absolutely going... to hate me for this. but I still did it. ran a 31b dense model (Gemma 4 31b Q4) with only 8 GB VRAM last week I ran Gemma 4 26B A4B a mixture of experts model on my RTX 4060 and hit 25–28 tokens/sec using llama.cpp's new MTP support. smooth. snappy. but MoE has a secret: it only activates 4B parameters per token despite having 26B total. that's why it flies. so the real question started haunting me. what if I throw a full, no tricks, every parameter fires on every token, 31B DENSE model at the same machine? # Hardware: GPU: NVIDIA RTX 4060, 8 GB VRAM RAM: 16 GB CPU: Intel Core i7 H Laptop. Gaming. Modest. The model: gemma-4-31B-it-qat-UD-Q4_K_XL.gguf (model's unsloth huggingface link in the comments) This is Google DeepMind's flagship dense model in the Gemma 4 family that can run on single consumer GPU. It packs a hybrid attention architecture, supports up to 256K context natively, and is QAT (Quantization Aware Training) optimized, meaning it retains far more quality than standard post training quants at the same bit depth. This is NOT the MoE. This is 31 BILLION dense parameters, every single one of them loaded. # the flags I used: -m gemma-4-31B-it-qat-UD-Q4_K_XL.gguf -cnv --spec-type draft-mtp --spec-draft-model mtp-gemma-4-31B-it.gguf --spec-draft-n-max 8 --spec-draft-p-min 0.6 -c 6000 -v Multi Token Prediction (MTP) is still active here. Separate draft GGUF required, same as the 26B setup. # Results: → Decode: ~3 tokens/sec → Prefill: ~2 tokens/sec → Context: 6000 tokens → Hardware crying quietly in the corner: yes so is 3 tps actually usable? For real time back and forth chat? Not ideal. You're not having a fluid conversation at 3 tps. but slow ≠ useless. And this is where it gets genuinely interesting. think about how senior devs actually work in a real team. But when something is architectural, deeply complex, or needs serious reasoning? they walk down the hall and escalate to the senior. That's exactly the local AI agent architecture this unlocks: → Fast orchestrator model (Gemma 4 26B MoE at 25+ tps) handles routing, simple queries, tool calls, memory. The junior dev. → Gemma 4 31B dense is the senior, called only when the fast model genuinely hits a wall. Hard multi step reasoning. Complex code generation. Deep architectural decisions. The agentic loop stays fast. Only the hard hops touch the 31B. That's a legitimate production grade local AI architecture on a budget hardware. (requires 2 8gb gpus) other workflows where 3 tps is completely fine: - overnight batch jobs. summarize documents, extract structured data, review code. Fire it off. Sleep. wake up to results. - One shot deep reasoning - Silent code audit loops, you write and test, the 31B reviews diffs and flags issues in the background between your sprints - Any workflow where output quality > output speed A few weeks ago, nobody was running a 30B+ dense model on a single consumer GPU with 8 GB VRAM. At all. Now we're doing it on an Intel i7-H gaming laptop with a NVIDIA RTX 4060, thanks to llama.cpp + QAT quants + MTP speculative drafting. Google DeepMind said the Gemma 4 31B targets "consumer GPUs and workstations." They were not exaggerating. The hardware bar to run serious frontier class models locally keeps dropping. the tools are here. the models are here. you just have to be willing to abuse your laptop a little. what workflows would you actually run on a local 3 tps 31B dense model? genuinely curious. drop it below.show more

FACT CHECK: Here at the first trial, the Commonwealth’s... own expert witness, Ian Whiffin, confirms the necessity & importance of hash values for the sake of “hash verification”, a necessary step in authenticating the data & being able to verify that it hasn’t been altered or manipulated. In fact, Whiffin actually gives this testimony in response to a question about when the data have been altered or tampered with, if there’s a way for the forensic examiner (him) to detect it, and/or verify its authenticity and integrity. Remarkably, despite the DFIR industry standard methodology of hash verifying a digital forensic extraction, like that of Jen McCabe’s iPhone, prior to conducting any analysis on it with any forensic tools, Ian Whiffin testified that notably, for his work on this case, not only did he abandon this standard methodology, but he also admitted that the forensic extraction of Jen McCabe’s iPhone, which he received from the Commonwealth, was stripped of its hash value. Perhaps more remarkably, this stunning fact apparently didn’t raise any red flags for Ian Whiffin when conducting his analysis in this case, where he’s providing testimony in a murder trial. One must ask themselves why that is? However, defense expert Richard Green, in his affidavit, states that: “Typically, forensic examiners are provided with the raw image file and the associated: hash value documentation together. After validating the hash value, I would then accept that the data has not been manipulated. Here, however, the hash documentation was not provided with the raw image of the cell phone. Instead, it was withheld from the defense. As a forensic examiner having received hundreds of imaged phones over the course of my decades-long career, this was unprecedented.” Contrary to Mr. Whiffin’s approach, upon initially receiving a purported extraction of Jen McCabe’s iPhone without a hash value to authenticate and verify the integrity of the data, Mr. Green promptly requested the hash value and corresponding GrayKey supplemental files from the Commonwealth in order to conduct his analysis. After making this demand, and when the Commonwealth had to produce the hash verification data for Jen McCabe’s iPhone, remarkably, the Commonwealth also produced—for the first time, and over a year later on February 8, 2023—the Full File System Extraction of Jen McCabe’s iPhone (see “Notice of Discovery VIII,” attached). Unlike the initial purported “extraction” produced by Trooper Nicholas Guarino, this one contained Jen McCabe’s incriminating 2:27am Google search and all of the manual deletions of her communications, among other incriminating evidence, surrounding the murder of Officer John O’Keefe (see defense’s Rule 17 motion from April 12, 2023, attached). So, this begs the question: If Ian Whiffin knows the importance of hash verification in validating the authenticity of the data he’s working with in the first place, then why didn’t he take the same actions as defense expert Richard Green did to responsibly and reliably provide analysis in this case? If Whiffin ought to be deemed an expert, qualified to provide analysis and testimony at trial, then why did he abandon his industry’s standard methodology of hash verification in this case? Even Cellebrite knows this is a no-no! What say you? #KarenReadTrial #Cellebrite #DFIRshow more

In July 2014, Lars Mittank, a 28-year-old from Germany... went on a vacation with his five of friends to Varna, Bulgaria. The week went by really fast," said Paul Rohmann, one of Mittank’s friends. However, the day before the group was scheduled to fly back home, Lars got in a fight with other German nationals at a bar. The fight stemmed from a disagreement over football. Lars was absent for most of the evening but reappeared at his hotel the next morning, recounting to his friends that he had been assaulted by a group of four men, resulting in injuries to his jaw and a ruptured eardrum. Afterwards, Lars visited a doctor who recommended against flying due to his injury. Additionally, the doctor prescribed an antibiotic for him. Despite his friends' desire to stay with him, he insisted he could manage alone. He encouraged them to stick to the initial travel schedule and fly home on July 7, which they did. Lars checked out of the hotel with his friends and then checked into a new hotel. But this is where things turn really strange. The day following his friends' departure, Lars displayed odd and paranoid behavior. From his hotel, he contacted his mother, Sandra Mittank, speaking in hushed tones, expressing fears that individuals were attempting to harm or rob him. He also urged her to cancel his credit cards. The CCTV in the hotel captured him pacing up and down the halls, looking out windows, and hiding in an elevator. Around 1 am, he departed from the hotel and came back approximately an hour later. His whereabouts during this time remain unknown. In the morning, he once again called his mother, telling her that the people pursuing him were getting closer. Lars arrived at Varna Airport on July 8, 2014, the day he intended to board his flight back to Germany. He sought advice from the airport doctor, Dr. Kosta Kostov, who later characterized his demeanor as "nervous and erratic." According to Kostov, he told Mittank that he was fine and could return home. Lars also expressed concerns about the medication he was taking. Shortly after, a construction worker entered the office, causing Lars to exclaim, "I don't want to die here. I have to get out of here." He abruptly fled the office, abandoning his luggage containing his wallet, cell phone, and passport. These final moments of Lars Mittank can be observed in the CCTV footage provided below. Upon exiting the airport, he can be observed in the footage jogging away from the airport, scaling a fence, and bolting into a meadow. Following this, he completely disappeared.show more

🚨12 HOUR NEWS RECAP 1.⁠ Trump said his support... among Republicans is at record highs, touting 90–95% approval within MAGA ranks - even as the Epstein file controversy brews inside his own base. He called the renewed outrage a “hoax” pushed by “radical left troublemakers.” 2.⁠ Israeli jets hit Houthi “military infrastructure” at Yemen’s Hodeida port, targeting vessels, fuel tanks, and engineering gear used to rebuild the site. Israel accused the Houthis of using the port for Iranian-supplied weapons and maritime attacks against Israel. 3.⁠ Tesla opens its Hollywood diner to the public today. Elon promised: “If our retro-futuristic diner turns out well, which I think it will, Tesla will establish these in major cities around the world, as well as at Supercharger sites on long distance routes.” 4.⁠ Trump is pushing an executive order to block federal funding for AI companies whose models show “liberal bias,” demanding political neutrality in exchange for government contracts. With firms like OpenAI, Google, Anthropic, and xAI already landing $200M in Pentagon deals, it’s a clear warning: comply or lose the cash. 5.⁠ The Liberal Democratic Party might be about to eat dirt in Japan’s upper house election - and the conservative Sanseito is licking its chops. High inflation, rising immigration, and voters tired of PM Ishiba’s “steady hand” have turned this into a full-blown referendum. If Sanseito grabs enough seats, it’s not just the coalition that cracks - it’s possibly Ishiba’s job. 6.⁠ Republicans claimed Biden’s team used an auto-pen to rubber-stamp decisions while he was in “deep mental decline.” They think a bunch of his executive orders, judge picks, and pardons could be totally illegal. 7.⁠ New York state agreed to pay $450,000 to settle a lawsuit from Brittany Commisso, who alleged former Gov. Andrew Cuomo sexually harassed and groped her while in office. Cuomo's attorneys called the allegations "false" and opposed dismissing the lawsuit. 8.⁠ A zero-day flaw in Microsoft’s SharePoint servers just cracked open U.S agencies, energy firms, and universities. Hackers are inside - stealing keys, lifting docs, and maybe worse. No fix from Microsoft. No ID on the attackers. No way to know if they’re still in your systems. 9.⁠ The UK is finally fast-tracking self-driving legislation - but only for taxi- and bus-style trials starting in 2026. No private use. Just app-booked robo-shuttles. While the U.S and China already have self-driving cars on public roads, Britain’s playing it safe with a “consultation phase” and 2 more years of red tape. 10.⁠ Bo’s new Turbo e-scooter clocked 85 mph at Goodwood, with eyes on triple digits. Developed by engineers from the Bloodhound land-speed team and Williams F1, this isn’t last-mile mobility. It’s a $29,500 bullet on wheels - limited-run, built-to-order.show more

This Chinese guy built a Second Brain in Obsidian... and every morning gets 3 trading ideas that brought him $180,000 in 6 months. Inside he runs a pipeline of 6 workflows on N8N that automatically pulls every read article, listened podcast, and voice note into a shared Obsidian vault, and a neural network analyst every morning at 6:00 finds connections between the fresh and the old and puts the 3 strongest trading ideas for the day into the inbox. No analytics desk, no Bloomberg terminal, no Telegram chats with traders. Just a Mac Mini by the wall, an iPhone in the pocket, and 1 local Obsidian vault. And traditional quant funds keep entire teams of 8 people on salary for the same flow of insights, while his expenses are only subscriptions to Readwise, Whisper API, and N8N hosting. 6 pipelines process about 200 sources a day and close the monthly API bill at about $120. The Mac Mini itself stores the entire vault and keeps the neural network analyst running 24/7, and from the iPhone the owner drops any idea he hears on the go into a Telegram bot, and it lands in the vault inbox in just 30 seconds. The starting instruction that sits in the VAULT.md file at the root of his vault looks like this: "you are the AI analyst of a solo trader. you read his vault every morning at 6:00, find connections between fresh and old notes, and deliver 3 trading ideas he can verify in the hour before the market opens. pipelines: // Reader (pulls every article and highlight from Readwise, Twitter bookmarks, and Kindle into /notes) // Listener (transcribes podcasts through Airr and voice notes through Whisper, puts them in /notes) // Catcher (accepts any message from the Telegram bot and writes it to /inbox with a timestamp) // Connector (every night reads across the entire vault and updates the connection graph between 4,000 notes) // Briefer (at 6:00 AM writes a brief: 3 trading ideas for today plus the emerging thesis of the week, puts it in /inbox) // Mobile (lives in the iPhone, answers any question about the vault by voice, and confirms alerts while the owner is on the go). you wake the owner with a push notification only when a fresh note contradicts his active thesis or when 1 of the 3 morning ideas has a confidence score above 90%." This instruction immediately sets the role for the system and the limits of its autonomy. It knows it is supposed to connect new with old on its own. It knows it is supposed to prepare 3 trading ideas every morning on its own. It knows it connects the live trader only when a thesis is contradicted or an ultra-confident idea appears. → Reader pulls about 80 articles and highlights a day from Readwise, Twitter, and Kindle → Listener transcribes 4 to 6 podcasts a week through Airr and Whisper → Catcher intercepts all voice and text ideas through the Telegram bot, averaging 15 to 20 a day → Connector updates the connection graph between 4,000 notes every night, adding 25 to 30 new edges → Briefer puts a fresh brief with 3 trading ideas and the emerging thesis into the inbox at exactly 6:00 → Mobile answers any question about the vault by voice and confirms alerts right from the iPhone And only when a new note contradicts his active thesis or 1 of the ideas breaks 90% confidence does the orchestrator raise the owner with a push notification. And when the trader at that moment is driving to the gym or eating breakfast, the Mobile agent in his iPhone answers any quick question about the vault by voice: what he wrote about this ticker last week, which 3 sources support the idea of long NVDA, and what counter-thesis already sits in his notes. The trader makes the decision and sends the order before New York opens. The fresh brief from last Monday looks like this: "reader: 78 materials added over the weekend, 11 of them about semiconductors, 4 about energy, 3 about biotech. passing to connector." "connector: 27 new connections found between fresh materials and the vault, the strongest one is that the Goldman report from Wednesday matches the NVDA thesis you wrote 3 weeks ago." "briefer: 3 trading ideas for today: long NVDA (confidence 0.84), short Tesla at the close of the quarterly report (0.71), watch URI (0.62). emerging thesis of the week: the market is underpricing capex on data centers." "alert: your fresh note about long-term risk in semis contradicts the NVDA thesis. sending for review." In his work setup there is no cloud server, no team of analysts, and not even a Bloomberg subscription. At home sits a Mac Mini with a local Obsidian vault, on top run 6 N8N pipelines and a neural network analyst, and the same vault mirrors to a secure terminal on the iPhone. Out of everything I have seen this year, this is the cleanest solo trading setup on a second brain: $120 a month on the API, about $30,000 a month into the account, and between them 6 pipelines, 4,000 connected notes, and 1 iPhone in the pocket.show more

I’m here only because I have a visceral reaction... to misinformation….Jessica Machado Apparently there’s some confusion as to why people are pissed. People aren’t pissed because J7 trashed Baker on your show. People aren’t pissed because you said you don’t like Baker. People aren’t pissed because you made jokes about Baker sticking his dick into a fleshlight. People are pissed about how you responded to Baker’s 4Loco live where he told the real story behind the sex toy gift package. Baker explained that he and J7 had a sexting flirtation for a couple weeks but after he started seeing some red flags in her behavior he cut things off. J7 doesn’t handle rejection very well so she went off the rails and started blowing up his phone constantly. She even tracked down his danish phone number after he had blocked her on everything else. It was then that she sent him that package of sex toys. That is scary and weird as fuck. We all know damn well that if a woman had cut off contact and blocked a man who wouldn’t leave her alone and that man then sent her a box of sex toys we all would be horrified. Baker also talked about the abuse he went through with his ex-wife and it was fucking heartbreaking. In response to alla this you did not say something like “Fuck man I’m sorry. I didn’t know the whole story so while I still think you’re an asshole I apologize for making fun of the sex toy stuff.” Instead you doubled down by saying that he “preys on women”, he has “severe issues with women”, he “played the victim”, implied that he is lying about all of this because you’ve “seen all of the text messages” and you “know things that we don’t”. That is why people are pissed. And yes, you are taking way more heat from Baker and everyone else than J7 is but that’s because we all agree that J7 is crazy and you’re not crazy so we expect more from you than we do from the crazy lady. Also, nobody thought that you were talking about Baker being a predator when you were actually talking about J7 and Proctor being predators. That’s why none of us were posting that clip in an effort to try prove that you called Baker a predator. It seems that a couple of the bunnies just found you saying the word “predator” on a livestream and assumed that we’re all idiots who lack basic listening comprehension skills. Then you took it and ran with it even tho that explanation made absolutely no sense. And yes, I fucking hate alla this. There are only like 8 people who are thrilled that this whole shitshow happened. The rest of us are pissed, disappointed, and sick about it.show more

Free NVIDIA GPU with 16 GB VRAM GPU for... Running Local LLMs! If you want to master local LLMs but you're waiting until you can afford a $1,500 GPU, you're honestly not going to make it. The open source AI ecosystem is moving way too fast for you to wait on your budget to catch up. Especially when you can build a bleeding edge inference engine from scratch right now, completely for free. You don't need a heavy local rig to start. Google is literally letting you use an enterprise grade NVIDIA Tesla T4 GPU for $0/hour. At standard cloud computing rates (~$0.20/hr), Google Colab’s 4 hour daily free tier hands you roughly $24 worth of data center tier GPU compute every single month. And most people just waste it. Let’s talk about the hardware you get access to for free. The NVIDIA Tesla T4 is an absolute workhorse: - Architecture: NVIDIA Turing (TU104) - VRAM: 16GB GDDR6 (320 GB/s bandwidth) - Compute: 320 Tensor Cores | 2560 CUDA Cores - Performance: 130 TOPS INT8 | 8.1 TFLOPS FP32 - Power: Sipping energy at a max 70W TDP This is the exact same hardware I used to run DeepMind's Gemma 4 26B A4B QAT MoE at a 250,000 context window without a single Out Of Memory (OOM) crash. If you have a web browser and 10 minutes, you have everything you need. I’ve put together a fully documented, cell by cell Google Colab notebook that teaches you exactly how to do this. Here is what the notebook actually teaches you: - How to provision an Ubuntu Linux environment with CUDA 13.0 and verify your driver stack. - How to pull the source code and compile the latest llama.cpp C++ binaries from scratch, specifically optimizing the build for your exact GPU using the -DCMAKE_CUDA_ARCHITECTURES=native flag. - How to directly download quantized local LLMs (GGUF format) straight from HuggingFace using the CLI. - How to manage 16GB VRAM limits, offload neural network layers to the GPU, and push massive context windows. Compile raw llama.cpp, ollama run a model, or spin up the LM Studio CLI. Pick whatever stack you are comfortable with. just start building. No hardware. No credit card. No excuses. Bookmark this post right now so you don't lose the tutorial. Even if you don't have time to run it today, you are going to want this workflow in your engineering toolkit. The link to the free Colab Notebook is in the comments below. Lemme know if you need more tutorials like this.show more

[Discrete Fourier Transform] by Hand ✍️ In signal processing,... the Discrete Fourier Transform (DFT) is no doubt the most important method. But the math involved is extremely complex, literally, involving a summation over a complex number term e^(-iwt). I developed this exercise to demonstrate that underneath such complexity, DFT is just a series of matrix multiplications you can calculate by hand. ✍️ Once you see that, it should not surprise you that a deep neural network, which is also a series of matrix multiplications, with activation functions in-between, can learn to perform DFT to process and analyze signals so effectively. How does DFT work? [1] Given ↳ Signals A, B, and C in the 🟧 frequency domain: ◦ A = cos(w) + 2cos(2w) ◦ B = cos(w) + cos(3w) + cos(4w) ◦ C = -cos(2w) + cos(3w) ◦ Each signal is a weighed sum of four cosine waves at frequencies 1w, 2w, 3w, and 4w. ◦ We will apply Inverse DFT to convert the signals to time domain representations, and then demonstrate DFT can convert back to their original frequency domain representations. ↳ Signal X in the 🟩 time domain. X is sampled at 10 time points 1t, 2t, …, 10t: ◦ X = [-2.5, -1.8, 3, -0.7, -1.0, -0.7, 3, -1.8, -2.5, 5] ◦ Suppose X is also a weighted sum of the same four cosine waves, but we don’t already know their weights. We will apply DFT to discover them. [2] 🟧 Frequency Matrix (F) ↳ Write the coefficients of A, B, C as a matrix F. Each signal is a row. Each frequency is a column. ↳ A → [1, 2, 0, 0] ↳ B → [1, 0, 1, 1] ↳ C → [0, 1-, 1, 0] [3] Cosine → Discrete ↳ Sample from the continuous cosine waves at discrete time points 1t, 2t, 3t, to 10t. [4] Cosine Matrix (W) ↳ Write the samples as a matrix, Each frequency is a row. Each time point is a column. [5] Inverse DFT: 🟧 Frequency → 🟩 Time ↳ Multiply the frequency matrix F and the cosine matrix W. ↳ The meaning of this multiplication is to linearly combine the four cosine waves (rows in W) into time-domain signals (rows in T) using the weights specified in F. ↳ The result is matrix T, which are signals A, B, C converted to the time domain. Each signal is a row. Each time point is a column. [6] Transpose ↳ Transpose T, converting each signal’s time domain representation from a row to a column. [7] DFT: 🟩 Time → 🟧 Frequency ↳ Multiply the cosine matrix W with the transpose of matrix T. ↳ The purpose of this multiplication is to take a dot-product between each time-domain signal (columns in the transpose of T) and each cosine wave (rows in W), which has the effect of projecting the signal onto a cosine wave to determine how much they are correlated. Zero means not correlated at all. ↳ The result is an intermediate version of the “recovered” frequency matrix where each column corresponds to a signal and each row corresponds to a frequency. ↳ Compared to the original frequency matrix F, this intermediate matrix has non-zero weights in the correct places, but scaled up by a factor of 5 (n/2, n=10). For example, signal A, originally [1,2,0,0], is recovered at [5,10,0,0]. [8] Scale ↳ Multiply each value by 2/n = 1/5 to scale down the intermediate matrix to match the magnitude of the original frequency matrix F. [9] Transpose ↳ Transpose the recovered frequency matrix back to the same orientation of the original frequency matrix F. ↳ Like magic 🪄, the result is identical to the original F, which means DFT successfully recovered the frequency components of signals A, B, C. [10] Apply DFT to X: 🟩 Time → 🟧 Frequency ↳ Now that we have some confidence in DFT’s ability to recover frequency components, we apply DFT to X’s time-domain representation by multiplying W with X. ↳ The result is the an intermediate matrix. [11] Scale ↳ Similarly, we scale down by a factor of 5 to obtain the recovered frequency components of X (a column). [12] Transpose ↳ Similarly, we transpose the recovered column to row to match the orientation of the frequency matrix. ↳ Using the coefficients [0,0,3,2], we can write the equation of X as 3cos(3w) + 2cos(4w). Notes: I hope this by hand exercise helps you understand the essence of DFT. But there is more technical details, such as: • Sine: The complete DFT math also includes sine waves that follow a similar calculation process. • Phase: Here, we assume all the cosine waves are aligned at the origin, namely, phase is 0. If a phase p is added, for example, cos(w+p), we will need to calculate the sine component and use their ratio to figure out what p is. • Magnitude: If phase is not zero, the magnitude will need to be calculated by combining both cosine and sine terms.show more

DIRECTED ENERGY WEAPONS (DEWs) While typical civilian access to... DEW technology is hyper limited, AI gives more insight into the reality of Directed Energy Weapons as used in combat. However, this is not a very comprehensive snapshot of what DEW munitions really are. First, they are not lasers, but rather known as sasers (sound outside of human audible range) that delivers more than just excited photons as per lasers, but envelopes microwave radiation delivered through linear particle accelerators and amplified by square-wave (jackhammer) ultrasonic concussion exceeding 20,000 pulses per second, where each impact doubles in energy through secondary emissions for every pulse. Secondary emissions are created on-the-spot through Neutrino Events that harvests brand-new energy directly out of atmospheric neutrinos, turning them into new ions. Hence why saser beams deliver more energy to the target than anything short of a nuclear blast. DEWs also actively draw energy out of surrounding capacitors at the scene of the target as well as from local sources in the path of the beam. Microwaves excite the electrons within the target structure, such as car batteries and the electric wires inside homes, that causes additional increase in amplitude. Since the energy being siphoned off is focused in the zero point of the beam, that area creates dual opposing twin vortices at zero target forming a hyperbola in the center of the affected area where the energy collapses in on itself, sucking oxygen out of the air which then amplifies the thermal radiation only within that toroidal vortex area, achieving crucible-level temperatures in an open-air setting. This is hyper-accelerated by the coupling of the SBX-1 mobile Vortex generator that delivers oxygen to the strike zone with hyper-focused accuracy at up to 70+ MPH winds. Which is why target zones see hurricane winds out of nowhere and from clear skies without any clouds. DEWs generate temperatures that are vastly higher than normal house or forest fires; 1200F and 1500F respectively. A self-imploding DEW saser strike can generate thermal signatures high enough to burn terracotta roofing tiles (2100F), melt glass (2900F), and vaporize stucco walls that are rated to hold up for 1 HOUR under direct torch flame, rendering it to tiny traces of powder. Even the concrete slab foundations of homes in the Santa Rosa fire of 2017 had been incinerated and literally gone as if evaporated into the air. Since the air surrounding the strike zone becomes immediately depleted, fire will not be readily sustained anywhere outside of the hyperbola area (just the home or automobile), leaving brush, trees, plastic, within just a few feet away barely seared and, in many cases, totally unscathed altogether. Houses that incinerate all the way down to ash within minutes will leave no soot, or flame marks on white-painted homes as little as ten feet away as if there was no heat or fire present of any kind. It is important to note that a normal house fire takes 3-4 hours to burn down, leaving large portions of the home and contents behind unconsumed. DEW crucible vortex fires consumed homes down to nothing but white ash in Santa Rosa, incinerated within 20 minutes that I observed after pounding on doors in one residential neighborhood of 2 story homes at 2:30 in the morning screaming to get out, then returning to that same location less than a half hour later, with nothing left but smoldering ash. There is nothing ‘normal’ about DEW remains. Kitchen stoves, pots and pans, washers, dryers, water heaters simply vanished. The cast-iron engines in cars just gone with their aluminum alloy wheels and windows melted on the street. Since DEW sasers work through the delivery of microwaves (that are radioactive) through encapsulation within tachyons (like protective bubbles), the resulting debris fields of target areas are also left contaminated, similar to the aftermath of nuclear detonation sites. The sites where homes once stood in the Santa Rosa fire for instance, had to have the soil under the vaporized concrete slabs excavated and stored in radiation containment casks before any new construction could resume, as reported by locals there that were involved in the cleanup. The Lahaina target homes in Maui are still blocked off with absolutely no entrance for any reason for this same reason almost a year and a half later. DEWs have been seen pulling massive arcs from lightning strikes, obviously generated by the extreme excitation of atmospheric particles of the beam, as well as from overhead electrical wires that supercharge the phonon shafts that are invisible to the naked eye unless backlit from flame or sparks. This technology has been shown to cut full size military ships in half from high altitude delivery systems in seconds, rendering all previous forms of explosive armaments, including nuclear warheads, obsolete. All the military weapons delivered from the US and other countries to Ukraine in recent years were outdated and considered unusable in a genuine modern conflict and were therefore literally being disposed of to make room for other, newer forms of weaponry in armories here at home. In other words, they were merely junk. US dark and black ops have had functioning DEW assault systems in place for many decades already. While AI admits to Turkey using them in combat in 2019, the real date of deployment of DEWs goes back more than just centuries, but prior to humans' arrival to earth in 560m BC. A recent use of DEWs shown here from the Great Chicago Fire of 1871. An even more recent DEW ‘Tara Cleansing’ as they’re called, was the Great San Francisco Fire of 1906 that was blamed on a 7.9 earthquake. And yes, the HAARP SBX-1 hurricane generator is also an earthquake generator as well. None of this is new tech, just new to you. Satellite and antigravity drone delivery of such weapons are entirely remotely-controlled, placing no mil soldiers' lives on the line to mount a siege, making military battlefield loss of life virtually a thing of the past at this time. VIDEO: DEW ATTACK PACIFIC PALISADES 1/8/25 gratis Kyle Zinkshow more