God First🙏| Exposing corruption, amplifying accountability, building public databases for justice |🤖 Building journalism + AI tools. Founder PrimeKeyHomes🏯
Right here and now
Joined June 2023
- Tweets7.5K
- Following1.9K
- Followers593
- Likes6.8K
Pinned Tweet
Every day we’re given the chance to believe in others.
To believe that people are doing their best.
To believe that everyone deserves the chance to be their best.
Kindness takes so many forms: A gentle word of encouragement. Helping someone grow. Or being an anonymous angel who fills a stranger’s gas tank.
This video brought me to tears in the best way. Last night, I was struggling with something difficult… but seeing this gave me tears of joy.
It reminded me of something powerful: acts of kindness don’t end with the person you help. They ripple outward. They touch lives you’ll never even know.
So give freely. Help freely.
Because your kindness might be the light that heals someone far beyond your reach.
Song: Surrender by Natalie Taylor
#kindness
#compassion
#bekind
#inspiration
#innerpeace
#spreadlove
Ryan 🦅 retweeted
Still cooking on Saturday.
We compressed our most popular local cyber model down to 15.7 GB.
Meet OrcaSAQ-2 Cyber 27B Uncensored GGUF — built for defensive red teaming, vulnerability research, security coding, terminal workflows, and authorized security testing.
54.7 → 15.7 GB
262K context
94.4% Top-1 agreement
Cyber capability should not require sending your source code, logs, or vulnerabilities to someone else’s cloud. Run it locally. Keep the data local.
Our smallest cyber model yet and one of the most capable we’ve evaluated in this size class. Have fun!
huggingface.co/orcarouter/Or…
Ryan 🦅 retweeted
The Man In This Video, Who Was Arrested By The Nigerian Police, Is a Serial Thief Who Checks Into Service Apartments, Gains Access To The Property And Eventually Steals The PlayStation 5 (PS5) Before Disappearing.
He Has Been Carrying Out This Scheme For a While And Was Finally Caught. Apparently, This Is Becoming a Growing Trend In Some Service Apartments, With Thieves Targeting Valuable Electronics And Other Items Before Checking Out. 👀
Ryan 🦅 retweeted
@Tyson_Fury @anthonyjoshua
What a place NYC, & what a venue MSG,
& what a fight Fury vs AJ, the land of opportunity. what a weekend away for the lads on the piss in nyc. Iconic!❤️🙏 not long. ow guys,
Ryan 🦅 retweeted
NO DIGAS A NADIE QUE TE DIJE ESTO:
ALGUIEN HA METIDO UN HACKER CON IA ENTERO EN TU PROPIO PORTÁTIL 👿
Sin nube. Sin API.
Sin que salga un solo dato de tu máquina.
METATRON es un asistente de pentesting que corre 100% en local.
Le das una IP o un dominio y hace el trabajo de reconocimiento solo:
→ Lanza nmap, whois, whatweb, curl, dig y nikto
→ Le pasa TODO el resultado a un modelo de IA que corre en tu equipo con Ollama
→ La IA analiza el objetivo, señala los puntos débiles y busca los CVE
→ Puede pedir más herramientas a mitad del análisis, sola
→ Y lo guarda todo en una base de datos con historial completo
→ Con informes exportables en PDF y HTML
Corre sobre Parrot OS, en Python, con el modelo por Ollama.
3.200 ⭐. 658 forks. MIT.
⚠️ úsalo SOLO sobre sistemas tuyos o con permiso escrito. Escanear lo que no es tuyo es delito.
Para montarte un laboratorio y aprender ciberseguridad es una joya.
Para lo otro, es tu problema con el juez.
Link en los comentarios 👇🏻
Ryan 🦅 retweeted
🇨🇳🇳🇬 A Chinese travel creator was trying local food at a market in Nigeria when a group of men stopped him and demanded money.
When he asked why, their explanation was basically: "we’re Nigerian, you’re not, so we can take money from you."
Over roughly 10 kilometers, he says he passed more than 20 checkpoints and was asked for money again and again.
His final travel review?
“Don’t come to Nigeria.”
Writer: Sol
Ryan 🦅 retweeted
AI MODELS RUMORED TO BE RELEASING NEXT MONTH:
* Gemini 3.8 Flash
* Gemini 3.5 Pro
* Qwen 4
* GPT-5.7 “Zinc”
* GPT-5.7 “Magnesium”
* OpenAI Astra
* Claude Fable 5.1
* Claude Opus 6
* Grok 4.7
* Kimi K3.1
* Mistral next sparse open-weight model
* MiniMax M3.x
Strap in. It’s going to be a crazy month.
Ryan 🦅 retweeted
DeepSeek-V4-Flash-Vision-Exp is now live on the DeepSeek API Platform! 🚀
🔹 This experimental multimodal model matches DeepSeek-V4-Flash on text capabilities—including agents, reasoning, and world knowledge.
🔹 On multimodal agent benchmarks, V4-Flash-Vision-Exp makes a major leap over V4-Flash, bringing multimodal agent performance close to Opus-4.8.
Try it with model='deepseek-v4-flash-vision-exp'. DeepSeek Harness 0.1.1 was released today with out-of-the-box support for the new model.
1/n
Ryan 🦅 retweeted
CLAUDE CODE SESSIONS CAN NOW MESSAGE EACH OTHER
Instead of re-explaining your work, one session can send a summary to another and hand off the task instantly.
Ryan 🦅 retweeted
DEEPSEEK V4 FLASH IS FREE ON TWO ROUTES RIGHT NOW -- ONE IS AN AGENT, THE OTHER IS AN API KEY
• CodeBuddy, the agent
> Free until Dec 31, no card, nothing to install.
> Open any repo on cnb.cool, type @ codebuddy, turn on "work for me" mode.
• InferX, the API
> Free until Aug 12, no card.
> base url: model.inferx.net/endpoints/v…
> model: deepseek-v4-flash-0731
284B params with 13B active and 1M context behind both.
CodeBuddy runs on CNB, Tencent Cloud's platform, so aim it at side projects.
Five months on the first -- One week left on the second.
I picked a model once in Hermes -> three different models ended up doing the work, and my bill dropped 80.8%.
After that first click I never opened the model menu again. The router chose for every prompt on its own.
Three prompts, escalating:
• basic percentage math
> Cheap model. Back instantly.
• live BTC price off an API
> One tier up. Working python.
• Solidity staking contract with a 30 day lock
> Straight to frontier.
It's Smart Router from @clawhuntersol, and the setup is one dropdown.
The receipts:
• $0.13 spent against an estimated $0.67
> 80.8% savings rate. $0.46 of it from the router alone.
• grok-4-5, glm-5-2, deepseek-v4-flash
> None of them was my call.
It's all in the video -- watch the model change while the selector stays put.
-> clawhunter.fun/inference
You were never supposed to be the router.
Don't forget to Save & Try it ↓
⚠️Wyatt Tessari L’Allié, Executive Director of AI Governance and Safety Canada, told Canadian lawmakers that early signs of humans losing control over AI systems are already appearing.
His warning was clear:
Some researchers believe advanced AI could eventually create risks serious enough to threaten humanity itself.
That is why he urged Canada to help lead international discussions before systems become too powerful to regulate effectively.
The core issue is no longer whether AI will transform society.
It is whether governance can move fast enough to shape that transformation responsibly.
Technology compounds quickly.
Law usually does not.
#AI #ArtificialIntelligence #AIGovernance #FutureOfAI #Technology #Canada #successfularcs