Elliot by alias, security by profession. Focusing on fintech & security research.
/home/user
Joined June 2023
- Tweets6.5K
- Following928
- Followers5.3K
- Likes251K
Pinned Tweet
Los únicos trabajos que no va a poder reemplazar la IA y la robótica son aquellos con mayor conexión humana, no necesariamente porque estas tecnologías no puedan desempeñar esas funciones, sino porque sencillamente la gente prefiere interactuar con otros humanos.
ElliotSecOps retweeted
The only thing that stops a bad guy with an abliterated LLM is a good guy with an abliterated LLM.
ElliotSecOps retweeted
Introducing Kimi K3.1 Pro: our most capable 3T-class open-weight model.
🔹 1M-token context, 2.8T total / 104B active params (MoE), Multimodal.
🔹 Full-project coding: games in Unreal, Unity & Godot.
🔹 Excels at parametric 3D modeling in Blender & CAD.
🔹 More reliable at finance work: Accurate modeling with hallucination down by 65%.
🔹 Matching the strongest proprietary models across coding, agentic and knowledge benchmarks.
Kimi K3.1 Pro and K3.1 Pro Highspeed (2–3× faster) are now live on Kimi.ai, Kimi Work, Kimi Code, and the Kimi API.
Open weights by Nov 4, 2026.
🔗 API: platform.kimi.ai
🔗 Tech blog: kimi.ai/blog/kimi-k3-1-pro
ElliotSecOps retweeted
Today we're releasing abliterated-model-large-v2.
Based on GLM-5.3, which is #3 on Terminal-Bench 4.0 (behind only Opus 5 and Fable), with 2× the cyber exploitation of 5.2.
We abliterated and hosted it so it does the offensive cyber, red teaming, and agent testing work other models refuse to do.
- US-hosted
- FP8
- 1 million context window
- Zero input/output prompt retention
Live now. 🧵
Hay mucha alarma por la posibilidad de que la IA ataque las criptos, pero internet y la banca usan un cifrado todavía más débil. La IA no necesita romper la criptografía de Bitcoin para hacer daño, con los datos filtrados ya se potencia el phishing, la ingeniería social, etc.
Esa vulnerabilidad de PHP que está causando estragos se detectó con GPT-5.6 Sol 👀
OpenAI models helped researchers—operating as part of our Patch the Planet initiative—discover critical security flaws in PHP, an open-source programming language used to run an estimated 70% of websites.
Over the past two months, we led a coalition of maintainers, researchers, and security providers to prepare fixes and protections ahead of disclosure. This duration was necessary because the fix was quite intricate and required a major change to the PHP engine. Every site using PHP can now safely upgrade their version to stay secure, in addition to applying mitigations from various security vendors.
Due to the severity of these flaws and PHP’s widespread use, we consider this bug to be one of the most impactful vulnerabilities in recent memory. It also marks a new phase, where AI can find security bugs with internet-scale impact.
This is an example of the Defender’s Window in action: to our knowledge, frontier AI was able to find this vulnerability ahead of attackers, giving defenders time to prepare.
openai.com/index/coordinated…
Readers added context they thought people might want to know
OpenAI may claim a role in the coordinated response and patching process. But the initial discovery and exploit development, however, occurred several months earlier by an independent researcher.
x.com/lcfr_hax/statu…
Ahora con la IA se genera codigo más rapido que nunca, pero a su vez estamos revisando más codigo de lo que nunca antes lo habíamos hecho.
Se suponía que la IA iba a reducir nuestro trabajo, sin embargo siento que nunca antes había trabajado tanto en mi vida.
ElliotSecOps retweeted
software engineering is so back.
> people are reading books again.
> every line of code is being read carefully.
> fundamentals are cool again.
> architecture and system design are back in focus.
> nobody is fighting over frameworks anymore.
> understanding why something works matters again.
> code reviews matter more than lines of code shipped.
> debugging is becoming a superpower again.
> technical judgment matters more than typing speed.
> engineering is about thinking again, not just coding.
ElliotSecOps retweeted
Writing is a form of thinking.
So don't let AI do it all for you.
Apena hice este post OpenCode sufrió un acceso no autorizado afectando aproximadamente al 1.6% de sus usuarios, que a pesar de ser una cantidad pequeña, sigue siendo grave, pero sería injusto decir que OpenCode ha sido negligente cuando hoy los ataques son mucho más sofisticados
No hace falta ser objetivos de un APT para sufrir un acceso no autorizado o una brecha de datos, incluso un equipo de pocas personas con un enjambre de agentes haciendo reckoning día y noche (y un poco de ingeniería social) basta para encontrar un acceso oculto
Hasta el FBI cayó por un zero day, no por mala configuración, sino por software de terceros sin parche. Ni el equipo más capaz del mundo se salva cuando hasta los grupos más chicos de ciberdelincuentes operan con swarms de IA a costo marginal.
bleepingcomputer.com/news/se…
ElliotSecOps retweeted
nah bro we're in a simulation for sure
This entire ocean is a 61 KB HTML file.
Built with Claude Code + Opus 5.5 in about 30 minutes, according to its creator.
It runs in the browser using WebGPU: FFT waves, whitecaps, sun glitter, a physically based sunset sky with clouds, and a free-flying camera. No game engine or rendering library.
Claude rendered screenshots in headless Chrome, then had separate subagents critique them. The creator gave feedback along the way, mainly on the clouds.
The 61 KB build is minified and self-contained. Download it, open it in Chrome with WebGPU support, and explore with WASD.
ElliotSecOps retweeted
How does an actor have a better epistemological understanding of AI than the people building it?
Ben Affleck says AI is simply matrix multiplication, linear algebra and floating point calculations, and however convincingly it sounds like a person, it has nothing to do with thought, humanity or feeling
“Nor do I believe, and I think it’s important to draw the distinction between what AI does, which is simply predict the next token or the next word in a sentence, then add that word on, run it back to the machine, do a lot of computation and pick the next word."
"That is a very different thing than sentience or thought. And I think we need to be really clear about that because it can be convincingly, in an illusory way, sound like thought or like a person."
"But let’s be real clear. This is matrix multiplication. This is linear algebra. These are floating point calculations. It has nothing to do with thought, humanity, or feeling.”
No importa si estás feliz con Codex, Claude Code, Cursor o lo que sea, debes tener OpenCode así sea solo el plan de $10, es increible todo lo que rinde y puede lograr DeepSeek, nunca un modelo flash se había sentido tan SOTA
Varios micro-skills > un megaskill. Carga solo lo que necesitas, cuando lo necesitas, tus usuarios y tu tarjeta de crédito te lo van a agradecer.
El error más caro construyendo agentes es meter todas las instrucciones en un solo skill gigante "para tener todo en un mismo lugar"
Además del costo y la lentitud, está el déficit de atención del modelo. Entre más largo el texto, más probable es que ignore las reglas del medio.
Un micro-skill de 300 tokens enfocado en una sola tarea se obedece a la perfección; un testamento de 4,000 tokens es una sugerencia.