the whole thing from one prompt, including the b-roll, is the claim i'd poke at first.
b-roll is where every AI video i've made stopped looking like us
🤝 Paid partnership
Syren Video is now live and FREE to try!
ChatGPT moment for agentic video.
Prompt → agency-quality AI video → chat to edit.
Powered by Opus 5.5 + Syren renderer w/ 3D support.
Browser or Claude MCP.
Insane examples in thread 🤯🤯🤯
Try it free here: shorturl.at/r57qy
yo dándome cuenta de que acaban de lanzar un ordenador pensado para ser usado por la IA y no por personas
🤝 Paid partnership
AI is already smart enough.
Real-world tasks are still slow, expensive and unreliable, because we hand AI a computer built for humans, then wrap it in a heavy harness.
AI doesn’t need to get smarter. It needs a computer built for it.
Today we’re releasing Pine Computer
Readers added context they thought people might want to know
Pine's highlighted SaaS-Bench v1.1 results show higher checkpoint score (78.3%) but lower resolved full-task completion (27.4%) than comparators (29.2%, 31.1%). Pine Computer is in private beta, not a full public release.
unipat.ai/benchmarks/Saa…
pinecomputer.io/blog/introduci…
scraf retweeted
> encuentras un sofá de segunda mano a buen precio
> le pides las medidas al vendedor
> subes la foto del anuncio a maket
> lo colocas en tu plano
> sabes si cabe antes de ir a buscarlo
y yo aún seguía usando el metro.
🤝 Paid partnership
ACABAN DE LANZAR UN ORDENADOR PARA IA QUE GASTA 20 VECES MENOS TOKENS QUE CODEX
Y no es una exageración.
Estos son los resultados publicados en SaaS-Bench v1.1:
→ 78,3% de puntuación
→ La más alta de su tabla
→ 1/20 del coste de Codex
→ 1/26 del de Claude Code
Datos publicados por Pine.
Link abajo de la beta privada👇🏼
🤝 Paid partnership
AI is already smart enough.
Real-world tasks are still slow, expensive and unreliable, because we hand AI a computer built for humans, then wrap it in a heavy harness.
AI doesn’t need to get smarter. It needs a computer built for it.
Today we’re releasing Pine Computer
Readers added context they thought people might want to know
Pine's highlighted SaaS-Bench v1.1 results show higher checkpoint score (78.3%) but lower resolved full-task completion (27.4%) than comparators (29.2%, 31.1%). Pine Computer is in private beta, not a full public release.
unipat.ai/benchmarks/Saa…
pinecomputer.io/blog/introduci…
scraf retweeted
how many of my videos would Syren need before it gets the rhythm of my edits right?
want to try it on my channel
🤝 Paid partnership
Syren Video is now live and FREE to try!
ChatGPT moment for agentic video.
Prompt → agency-quality AI video → chat to edit.
Powered by Opus 5.5 + Syren renderer w/ 3D support.
Browser or Claude MCP.
Insane examples in thread 🤯🤯🤯
Try it free here: shorturl.at/r57qy
scraf retweeted
Same shopping agent. 20 decisions in a row. One already bought the item
→ Drex: 20/20 done
→ Microsoft Decision-1: 18/20
→ OpenAI GPT-6 Sol: 3/20
Drex placed the order in 10.4s. GPT-6 Sol was still stuck on decision 4.
Drex is the fastest decision model on @OpenRouter right now.
By @NaceAI 👇🏼
scraf retweeted
GPT-6 ASTRA ES EL MEJOR MODELO DE DISEÑO 3D AHORA MISMO
guía completa para usarlo desde codex
→ cómo meterte en codex para diseñar
→ cómo sacarle partido a gpt-6 en 3d
→ todo lo que necesitas para empezar
guárdalo antes de que se te pierda 🔖
scraf retweeted
DEJA DE PEDIRLE TEXTO A LA IA. PÍDELE TRABAJO.
llamar al modelo es lo fácil, lo que viene después es lo que duele
→ antes: modelo + tools + tu loop
→ ahora: un SDK y mandas un job
→ te llega trabajo hecho, no texto
prompt dentro, trabajo fuera
si construyes con IA, guárdalo y mira el lanzamiento 🔖
#PineComputer #ComputerUse #DeveloperTools
🤝 Paid partnership
AI is already smart enough.
Real-world tasks are still slow, expensive and unreliable, because we hand AI a computer built for humans, then wrap it in a heavy harness.
AI doesn’t need to get smarter. It needs a computer built for it.
Today we’re releasing Pine Computer
Readers added context they thought people might want to know
Pine's highlighted SaaS-Bench v1.1 results show higher checkpoint score (78.3%) but lower resolved full-task completion (27.4%) than comparators (29.2%, 31.1%). Pine Computer is in private beta, not a full public release.
unipat.ai/benchmarks/Saa…
pinecomputer.io/blog/introduci…
scraf retweeted
🚨 ANTES DE COMPRAR UN SOFÁ, HAZ ESTO (O TE VAS A ARREPENTIR)
Casi todo el mundo compra muebles a ojo y luego descubre que no caben en el salón.
Maket acaba de sacar una función que lo soluciona: haces una captura de cualquier mueble y lo metes en el plano de tu casa con sus medidas exactas.
Te dejo el paso a paso 👇🏼
1. Abre un proyecto: sube el plano de tu casa, genéralo con IA o empieza de cero. Luego pasa el ratón por "Objects" y pulsa "Create your own".
2. Sube una imagen: puede ser una captura de cualquier web o una foto que hayas hecho en la tienda. Ojo, solo funciona desde ordenador, no desde el móvil.
3. Selecciona el mueble: si la foto tiene varios, puedes elegir más de uno a la vez. Si solo hay uno, haz clic y listo.
4. Pon las medidas exactas: abre la ficha del producto en la web de la tienda y copia el alto, el ancho y el fondo en Maket.
5. Colócalo en tu plano: entra en "Objects", luego en "My Objects", selecciona tu mueble y arrástralo donde quieras.
En 5 pasos sabes si ese sofá cabe de verdad antes de gastarte un euro.
🔖 Guárdalo, esto te puede ahorrar mucho dinero.
🤝 Paid partnership
scraf retweeted
same task, start to done: 2-5x faster and about 1/25 the model api cost than running ai on a conventional computer, according to pine computer's early tests.
plus every customer gets their own computer, all running in parallel. still in private beta
#PineComputer #ComputerUse
🤝 Paid partnership
AI is already smart enough.
Real-world tasks are still slow, expensive and unreliable, because we hand AI a computer built for humans, then wrap it in a heavy harness.
AI doesn’t need to get smarter. It needs a computer built for it.
Today we’re releasing Pine Computer
Readers added context they thought people might want to know
Pine's highlighted SaaS-Bench v1.1 results show higher checkpoint score (78.3%) but lower resolved full-task completion (27.4%) than comparators (29.2%, 31.1%). Pine Computer is in private beta, not a full public release.
unipat.ai/benchmarks/Saa…
pinecomputer.io/blog/introduci…
scraf retweeted
SI QUIERES QUE CHATGPT TE RECOMIENDE, MIRA ESTE ANUNCIO, NO DIRÉ NADA MÁS...
🤝 Paid partnership
BREAKING: Today, we witnessed the future of marketing.
Introducing Citation Outreach
A profitable company can add +$3M/month just by getting recommended by ChatGPT. 12 months ago nobody believed this was possible.
Here's what changed:
People don't Google anymore. They ask ChatGPT.
"What's the best tool for [Category]?"
"Which [Category] tool/service should I use?"
ChatGPT gives them 3 or 4 names. Those brands get the customer. Everyone else doesn't exist.
So how does ChatGPT pick those names?
It reads what other websites say. Listicles, comparisons, reviews, how-to guides. The brands that keep getting mentioned there are the ones it recommends.
Citation Outreach gets your brand mentioned on those exact websites.
We find them, reach out to the people who run them, and get your brand offsite mentions. Fully done for you.
You just watch your brand start showing up in the answers.
This is absolutely insane.
Okey, esto empieza a dar miedo...
🤝 Paid partnership
BREAKING: Today, we witnessed the future of marketing.
Introducing Citation Outreach
A profitable company can add +$3M/month just by getting recommended by ChatGPT. 12 months ago nobody believed this was possible.
Here's what changed:
People don't Google anymore. They ask ChatGPT.
"What's the best tool for [Category]?"
"Which [Category] tool/service should I use?"
ChatGPT gives them 3 or 4 names. Those brands get the customer. Everyone else doesn't exist.
So how does ChatGPT pick those names?
It reads what other websites say. Listicles, comparisons, reviews, how-to guides. The brands that keep getting mentioned there are the ones it recommends.
Citation Outreach gets your brand mentioned on those exact websites.
We find them, reach out to the people who run them, and get your brand offsite mentions. Fully done for you.
You just watch your brand start showing up in the answers.
This is absolutely insane.
scraf retweeted
> ves un salón perfecto en Pinterest
> haces captura
> la subes a Maket
> seleccionas sofá, mesa y lámpara de golpe
> aparecen todos en el plano de tu casa
Pinterest acaba de dejar de ser solo inspiración.
🤝 Paid partnership
scraf retweeted
3 AI meeting assistants. 1 research assignment. A 5-step production plan.
Our content team was deciding which AI meeting assistant to feature in our next tutorial.
Normally, we'd research each product separately, compare features across multiple tabs, and spend more time turning those findings into a usable content brief.
This time, we assigned the research to Viktor inside Slack. I wasn't sure it would help. Every AI comparison I'd seen was a generic feature list that still needed hours of checking before we could use it.
We asked him to compare Fireflies, Otter, and Fathom across their features, practical workflows, limitations, and suitability for creators and small teams.
He organized the comparison, prepared a recommendation, and outlined a tutorial brief with potential on-screen demonstrations.
Then he mapped out five production steps covering research, scripting, recording, editing, and publishing. He also flagged the details that still needed verification and the decisions requiring our approval.
The real value isn't getting more information. It's having the research organized into something a team can review and act on.
Chatbots answer. Copilots assist. Employees own outcomes.
Hire @viktor_com for your team. $100 in credits included, no card. Full link in first comment.
#AIemployee
Paid Partnership
🤝 Paid partnership
scraf retweeted
SIGAN A ESTA CUENTA... NO DIRÉ NADA MÁS
BREAKING: Today, we witnessed the future of marketing.
Introducing Citation Outreach
A profitable company can add +$3M/month just by getting recommended by ChatGPT. 12 months ago nobody believed this was possible.
Here's what changed:
People don't Google anymore. They ask ChatGPT.
"What's the best tool for [Category]?"
"Which [Category] tool/service should I use?"
ChatGPT gives them 3 or 4 names. Those brands get the customer. Everyone else doesn't exist.
So how does ChatGPT pick those names?
It reads what other websites say. Listicles, comparisons, reviews, how-to guides. The brands that keep getting mentioned there are the ones it recommends.
Citation Outreach gets your brand mentioned on those exact websites.
We find them, reach out to the people who run them, and get your brand offsite mentions. Fully done for you.
You just watch your brand start showing up in the answers.
This is absolutely insane.
scraf retweeted
OPENAI RECONSTRUYÓ SU APP EN 30 DÍAS
acabo de ver esta entrevista y me parece una locura
→ rehicieron la app de escritorio
→ la ia ya ayuda a construir ia
→ un plan de chatgpt de 500$
esto va más rápido de lo que pensamos
guárdalo antes de que se te pierda 🔖
scraf retweeted
CÓMO CREAR WEBS QUE GENERAN $100.000 AL MES
este tipo lo explica en 25 minutos, paso a paso.
idea → web → monetización
🔖 guárdalo.
