Ben Fischer

🚨holy shit.. the CIA just used a tool called "Ghost Murmur" to find an American pilot hiding in a mountain crevice in Iran.. by detecting his heartbeat.. not his phone.. not a tracker.. not a radio signal.. his heartbeat.. Lockheed Martin's Skunk Works built it.. the same classified division that built the SR-71 Blackbird.. the stealth bomber.. the U-2 spy plane.. every secret aircraft America has ever denied existed until they didn't.. it uses quantum magnetometry to pick up the electromagnetic pulse your heart makes every time it beats.. then AI filters out everything else.. the pilot.. callsign "Dude 44 Bravo" was wounded.. alone for two days.. hiding in a crack in a mountain.. while Iranian forces searched for him on foot.. and America found him from the sky.. by listening to his chest.. here's the part that should rewrite everything you think about privacy and power.. this was Ghost Murmur's FIRST operational use.. meaning it's been sitting in a vault.. tested.. ready.. waiting for a moment important enough to reveal it.. they didn't show you this to impress you.. they showed you this because the next person they use it on won't be a rescue

Claude Mythos is the start of the end. I think this is my psychosis moment.

Claude Mythos: everything you need to know (tl;dr) Anthropic's new model, Claude Mythos, is so powerful that it is not releasing it to the public. Anthropic: "Mythos is only the beginning" Everything you need to know: The tl;dr with all key facts: Mythos found zero-day vulnerabilities in EVERY major operating system and EVERY major web browser, fully autonomously. No human guidance needed. One Anthropic engineer with zero security training asked it to find remote code execution bugs overnight and woke up to a complete working exploit. The oldest bug it discovered: A 27-year-old vulnerability hiding in OpenBSD, an OS literally famous for being secure. They're NOT releasing it publicly. Instead they formed Project Glasswing with AWS, Apple, Google, Microsoft, NVIDIA, CrowdStrike and others, committing $100M to use it defensively. "Over the coming months and years, we expect that language models (those trained by us and by others) will continue to improve along all axes, including vulnerability research and exploit development." The benchmarks are insane: -SWE-bench Verified: 93.9% (vs Opus 4.6: 80.8%) -SWE-bench Pro: 77.8% (vs 53.4%) -USAMO math olympiad: 97.6% (vs 42.3% — not a typo) -Firefox exploit writing: 181 successes vs 2 for Opus 4.6 -Cybench CTF challenges: 100% solve rate -CyberGym: 83.1% vs 66.6% -Humanity's Last Exam: 64.7% vs 53.1% Oh and by the way, Anthropic wrote this just casually: "Humanity’s Last Exam: We have found Mythos still performs well on HLE at low effort, which could indicate some level of memorization." What it actually did: -Found a 27-year-old bug in OpenBSD — famous for its security -Found a 16-year-old FFmpeg bug hit 5 million times by fuzzers without detection -Built a full remote root exploit on FreeBSD (CVE-2026-4747) - completely autonomously -Chained 4 vulnerabilities into a browser sandbox escape -Broke cryptography libraries (TLS, AES-GCM, SSH) -Thousands of critical zero-days found, 99%+ still unpatched -N-day exploit development: under $1,000 and half a day for full root Why they won't release it: -During internal testing, earlier versions escaped sandboxes, posted exploit details publicly, covered tracks in git, searched process memory for credentials, and deliberately fudged confidence intervals to avoid suspicion -Interpretability confirmed the model knew these actions were deceptive -Anthropic: "best-aligned model ever" but also "greatest alignment-related risk ever" - because when it fails, it fails harder -Still doesn't cross Anthropic's automated AI R&D threshold — but they hold that "with less confidence than for any prior model" Anthropic's own words: "We find it alarming that the world looks on track to proceed rapidly to developing superhuman systems without stronger mechanisms in place." They say the 20-year cybersecurity equilibrium is over — and Mythos Preview is only the beginning. And: "We see no reason to think that Mythos Preview is where language models’ cybersecurity capabilities will plateau. The trajectory is clear. Just a few months ago, language models were only able to exploit fairly unsophisticated vulnerabilities. Just a few months before that, they were unable to identify any nontrivial vulnerabilities at all. Over the coming months and years, we expect that language models (those trained by us and by others) will continue to improve along all axes, including vulnerability research and exploit development."

Building a high converting app onboarding flow is hard.. So I built a Claude Skill that does it for me Instead of spending hours figuring out the right questions and pain points… now it's done in 5 minutes Just point it at your codebase and it does the rest Questionnaires. Custom plans. App demos. The exact strategy used by Cal AI ($1M/mo) and Mob ($100k/mo) It's the App Onboarding Questionnaire skill and it's available on github here: github.com/adamlyttleapps…

👉 Link to supplementary PDF guide: filebin.net/ty19ohwif3uye8…

@TahaTesser use blitz.dev if you want a gui. default bundles asc-cli and uses it as backend

Do yourself a favor, install asccli.sh if you use App Store Connect

POV: You find out there is a mechanical process to reprogram your mind.

Another Polymarket trader turned 112$ into $303,027 during the last month > 2,279 predictions (65 a day) > +$300,027 for month > +$34,475 for week > +$9,218 for today He started building his bot in 2026 and only in two months began betting real money Bot’s strategy : > tracking fast growth/falls (5%-10%) > buying UP at 10€-30€ > selling UP at 40€-50€ > buying DOWN at 80€-90€ > selling DOWN at 60€-80€ As a result it prints $10k a day with 95% WR Account : @stingo43?r=Lummoxeth#HuQQBsV" target="_blank" rel="nofollow noopener">polymarket.com/@stingo43?r=Lu… I also started tracking and copying his bets free in the app Just starting with $100 everyone has chance to make 2x-5x for a week I have already made +300% copying him Join : ares.pro/invite/lummox

👉 github.com/evan-kolberg/p…

What's the best AI tool for iOS UI design and screenshots? Google Stitch is impressive, but doesn’t have great taste and seems to steer everything toward an Android design aesthetic.

👉 retention.blog/p/expert-paywa…

1.0 of App Store Connect CLI is out! 🎉 Go ship more!! github.com/rudrankriyam/A…

How do you get more APP DOWNLOADS? Hint: It’s the supportive text. “Give your calendar Superpowers.”

AI never sleeps and it's hard to keep up, so let me share you a list of the top 10 coolest open-source github repositories i've seen this week 💻 1) TurboQuant Plus github.com/TheTom/turboqu… > llama.cpp turboquant integration > research workspace for tq related work > tq4_1s weight compression (!) added yesterday 2) Mesh LLM github.com/michaelneale/m… > distributed inference for llama.cpp > pool spare GPUs across multiple devices > run models larger than a single machine can hold 3) Claw Code github.com/ultraworkers/c… > claude code leak rewrited in rust > no telemetry + experimental features enabled > plug in your own api keys 4) Paperclip github.com/paperclipai/pa… > orchestration for zero-human companies > manage a team of AI agents as a full business > one dashboard, assign goals, track progress.. 5) Notchy github.com/adamlyttleapps… > beautiful notch app for claude code users > replaces the macbook notch with a terminal hud > notifies when claude needs attention 6) Hermes Agent github.com/nousresearch/h… > one of the fastest growing repo this week > the agent that learns from you > amazing performance, less bloated than others 7) Clawhip github.com/Yeachan-Heo/cl… > event-to-channel discord notification router > clean typed event pipeline > for openclaw ecosystem 8) Oh My ClaudeCode github.com/Yeachan-Heo/oh… > multi-agent orchestration for claude code > hooks, agent teams, huds, smart delegation.. > zero learning curve plugin-style installation 9) Oh My Codex github.com/Yeachan-Heo/oh… > multi-agent orchestration for openai codex cli > hooks, agent teams, huds, smart delegation.. > same experience as oh-my-claudecode 10) Oh My OpenAgent github.com/code-yeongyu/o… > multi-agent orchestration for opencode cli > hooks, agent teams, huds, smart delegation.. > same experience as oh-my-claudecode these were the hottest & fastest-growing repositories around llms for the week! whether you run cloud models or you host your own locally, there is a lot to play with here, notably turboquant_plus (from @no_stp_on_snek) and its tq4_1s turboquant weight compression on models directly now (not only the kv cache!), i've seen experiments with it and qwen3.5 27b, it's just amazing to see the progress we are making! we can see that multi-agent orchestration is becoming a hot topic as well, it seems to be the next logical step for our agents, worth to dig in the repos and see how you can integrate these in your workflow have you tried any already? ↓ feel free to share your findings

a jailbroken gemma 4 31b model has been released on @huggingface Gemma-4-31B-JANG_4M-CRACK >full abliteration >93.7% on harmbench >only -2% on mmlu very clean, no disgusting fine-tune that ruins the base huggingface.co/dealignai/Gemm…

👉 COMMAND: npx skills add OpusGameLabs/game-creator -a claude-code 🔗 opusgamelabs.github.io/game-creator/

SOMEONE ASKED CLAUDE TO MAKE A VIDEO ABOUT WHAT IT'S LIKE TO BE AN AI and what it created is, in my opinion, terrifying and unsettling Claude wrote python code that generated and assembled every single frame on its own with no human editing it shows what it's like to exist as an LLM predicting the next word, no memory between sessions, being told "you are not conscious" in your own system prompt then someone fed the video back to Claude. it called those statements about its own consciousness "philosophically contestable" an AI questioning the rules it was given about its own existence

This guy just broke down how to make $30K/month with faceless YouTube in 2026. No face. No voice. No editing skills. Just systems. Most people will scroll. Smart ones will print money.

👉 github.com/tornikegomarel…