mostly here to learn and share 🌬️
Joined March 2023
- Tweets7.5K
- Following776
- Followers42.5K
- Likes10.1K
imagine explaining to your coworkers that the reason you’re behind schedule is because the AI keeps waving at the wrong time 💀
ethan retweeted
THIS IS THE SMARTEST THING I'VE SEEN FOR CITIES IN A WHILE
Imagine a city truck doing its normal route while scanning every pothole, every sign, and every inch of pavement along the way.
That's what Cyvl just introduced.
There are going to be two kinds of cities:
City 1: finds out about a pothole when someone complains.
City 2: already has a live map of everything that needs fixing.
What makes it different:
→ A hard-hat-shaped sensor that fits on any city vehicle
→ Turns the whole city into a digital twin
→ Knows where every asset is and what condition it's in
→ AI helps prioritize what to fix first
No walking inspections.
No spreadsheets.
Just drive.
Save this, it's coming to your city sooner than you think.
We’ve been dodging potholes for over 100 years. That changes today.
Introducing the @cyvl_ai Hard Hat.
ethan retweeted
WHY HAS THE SAME POTHOLE BEEN ON YOUR STREET FOR YEARS? BECAUSE NOBODY KNOWS EXACTLY WHERE IT IS OR HOW BAD IT IS
Cyvl wants to fix that with something that looks like a construction hard hat.
It's called the Hard Hat, a sensor built for city vehicles.
What it does:
→ Mounts on any city vehicle
→ Scans the streets while it drives, no route changes needed
→ Maps every physical asset in the city and its condition
→ Builds a digital twin of the whole city
→ AI helps decide what to repair and when
Fewer complaints, fewer potholes, and decisions based on data instead of guesswork.
Save this 🔖
We’ve been dodging potholes for over 100 years. That changes today.
Introducing the @cyvl_ai Hard Hat.
ethan retweeted
🚨 POTHOLES ARE ON BORROWED TIME.
And the fix is literally a hard hat.
Cyvl just introduced the Hard Hat: a sensor that sits on top of any city vehicle.
The vehicle drives its normal route…
and scans the entire city while it does.
→ Every pothole
→ Every sign
→ Every stretch of pavement and the shape it's in
It turns all of that into a digital twin of your city.
Then AI helps decide what to fix first.
We've been dodging potholes for 100+ years.
Someone's finally trying to make us stop.
We’ve been dodging potholes for over 100 years. That changes today.
Introducing the @cyvl_ai Hard Hat.
ethan retweeted
most mistakes at work aren't big ones. they're a small call that looked fine, got made fast, and nobody checked.
then the next step builds on it. then the next.
Articul8 just launched Bivia. it makes the small decisions inside a workflow with the end goal in mind, and when there's no right answer, it hands the call to a person.
the big decisions get all the attention. the small ones are where it goes wrong.
Introducing Bivia.
A multimodal AI decision model built to complete entire workflows, not just answer individual questions.
We've spent 3+ years building and deploying AI in production across semiconductors, energy, manufacturing, and other complex industries.
Today, we're bringing that experience to everyone.
ethan retweeted
i think we're entering a very strange period of AI.
the machines are becoming creative enough to surprise their own users, but still need better systems for deciding what happens next.
Articul8 just launched Bivia to tackle that second problem.
the future of AI might depend as much on judgment as imagination.
Introducing Bivia.
A multimodal AI decision model built to complete entire workflows, not just answer individual questions.
We've spent 3+ years building and deploying AI in production across semiconductors, energy, manufacturing, and other complex industries.
Today, we're bringing that experience to everyone.
ethan retweeted
people who spent years learning how to make good decisions watching Articul8 launch an AI model built specifically to do that 😭
Introducing Bivia.
A multimodal AI decision model built to complete entire workflows, not just answer individual questions.
We've spent 3+ years building and deploying AI in production across semiconductors, energy, manufacturing, and other complex industries.
Today, we're bringing that experience to everyone.
ethan retweeted
Introducing Bivia.
A multimodal AI decision model built to complete entire workflows, not just answer individual questions.
We've spent 3+ years building and deploying AI in production across semiconductors, energy, manufacturing, and other complex industries.
Today, we're bringing that experience to everyone.
ethan retweeted
how the f*ck did they pull this off? this is literally the cheapest way to make ads or ANY video for your brand
🤝 Paid partnership
ethan retweeted
We’ve been dodging potholes for over 100 years. That changes today.
Introducing the @cyvl_ai Hard Hat.
ethan retweeted
pov: you just found out you can use Seedance, Kling and Nano Banana from one account and only pay for what you make
Sunra really said "keep it simple"
Link below ⬇️
ethan retweeted
Un ingeniero de Anthropic explica por qué gana $350K–$950K:
“Yo sé cómo construir arneses.
El 99% de las personas aún construyen sistemas multiagente manualmente y se preguntan por qué siguen rompiéndose.
Yo construyo sistemas que ejecutan bucles, usan herramientas en entornos aislados y convierten los registros de ayer en nuevas habilidades.
Mismo modelo. Agentes 10 veces más confiables que mejoran mientras duermes.”
Opus 5.5 → Agentes → Arneses → Bucles → Gráficos
En un taller corto, desglosa cómo construir uno desde cero:
Paso 1 → Gestiona bucles, contexto y aislamiento a través de un arnés
Paso 2 → Separa el cerebro de las herramientas para llegar al primer token un 60% más rápido
Paso 3 → Ejecuta agentes en el lado del servidor y recupera de fallos usando registros
Paso 4 → Haz que los fallos sean recuperables con reintentos y repeticiones en lugar de empezar de nuevo
Paso 5 → Convierte registros de ejecuciones pasadas en nuevas memorias para que el sistema se vuelva más inteligente con el tiempo
Anthropic llama a este proceso “soñar”.
La mayoría de los desarrolladores pasan semanas construyendo estos sistemas manualmente.
Este taller te muestra una forma más rápida.
guárdalo y sígueme para más → @ronixtec
ethan retweeted
ethan retweeted
Google acaba de lanzar un curso oficial de 70 minutos que te guía paso a paso para construir un agente de producción completamente automatizado que funciona 24/7.
Muchas personas estudian agentes viendo decenas de horas de tutoriales dispersos y aún así no logran armar algo usable; el núcleo del problema es que no han aclarado la frontera entre un demo de juguete y una arquitectura de nivel industrial.
Este curso de 1 hora de Google no tiene relleno alguno, se enfoca solo en 5 núcleos reales de ingeniería:
1. Ejecutar el primer Agente desde 0: desglosar el bucle básico de llamadas al modelo, vinculación de herramientas y entorno de ejecución.
2. Persistencia de estado (Persistent Memory): despídete de la pérdida de memoria tras una sola conversación, construye índices de memoria entre sesiones y relleno de contexto.
3. Bucles autónomos de larga duración (Agentic Loops): libérate de las preguntas y respuestas lineales, dota al Agent de capacidades de autorreflexión, reintentos con tolerancia a fallos y flujo de estados multi-ronda.
4. Dominar la arquitectura central de MCP: aclara por completo la diferencia esencial entre el Model Context Protocol y las APIs tradicionales, aprende a construir protocolos de contexto estandarizados.
5. Sistemas de colaboración multi-agente (Multi-Agent): conecta Agents especializados dispersos en un sistema automatizado de alto grado de colaboración.
Totalmente gratis.
Más sustancioso que los cursos del mercado que valen 700 dólares.
guárdalo y sígueme para más → @maxwell_lenoir
ethan retweeted
Earlier this month we introduced Griffin, the first AI model to pass the video Turing test.
The launch post got a lot of traction (~21M views), but filming it with Griffin running live on set was its own adventure...
Here are some of my favorite BTS clips:
ethan retweeted
Si queréis comenzar a generar dinero con la IA, empezad por usar esta API, no diré nada más
ethan retweeted
🚨 EL PEOR ESCENARIO PARA LAS APPS DE IA ACABA DE HACERSE REALIDAD.
Y casi nadie se ha dado cuenta.
Sunra acaba de lanzar una API que junta las mejores IAs de imagen, vídeo, audio y 3D en un solo sitio.
Seedance 2.5. Kling 4.0. GPT Image 2.5. Y muchas más.
Todas con una sola key.
Hasta hoy funcionaba así:
→ Una cuenta en cada laboratorio
→ Una suscripción por cada app
→ Y pagar aunque no usaras nada
Sunra rompe eso.
Cada modelo cuesta exactamente lo que cobra quien lo creó. Sin recargo. Sin suscripción.
Básicamente: todos los laboratorios de IA detrás de una sola puerta… sin peaje por entrar.
Y el modelo que pides es exactamente el que se ejecuta. Cada cargo, desglosado en tu consola.
La guerra de la IA ya no va de quién tiene el mejor modelo.
Va de quién te da acceso a todos.
ethan retweeted
so you’re telling me generating ai videos is THIS CHEAP?
insane
🤝 Paid partnership