@FaustoR0 Espectacular!
Que resolviste diferente a que si lo hacen directo en vercel cooo lo venía haciendo con el mío?
Español
Gaston Krasny
3.7K posts
Gaston Krasny
@gukras
✏️https://t.co/aNgDn1iaDT 🤖Snack AI newsletter (https://t.co/AN6bXbgVxK) 📱https://t.co/4BoNQ7NTcR (acquired by https://t.co/e3iF1rZ5tL)
Katılım Şubat 2010
81 Takip Edilen732 Takipçiler
Tengo ganas de hacer un taller de vibecoding para chicos (8-12 años aprox) en CABA.
Juntamos quorum? hay padres interesados en que hagamos algo asi?
Fausto@FaustoR0
For the past few weeks I’ve been teaching 50+ kids how to create their own games with AI. The results have been incredible 🤯 I started with my 5 y/o nephew because I was worried about him spending hours doomscrolling and consuming brainrot content. He built his first game using @v0, though I had to walk him through each step. Watching kids interact with these tools taught me so much about how they think, where they get stuck, and what keeps them motivated. The moment that really struck me came right after he finished that first game. Every time I saw him afterward, he’d run up and ask, “Did you bring your laptop so I can keep playing?”. For him, “playing” meant creating. That kind of excitement is organic pull. Don’t get me wrong, I love @v0 and @Lovable. They’re insanely powerful. Their UIs just aren’t designed for kids, and that’s totally fine. Kids still need guidance, structure, and a bit of tutoring to turn ideas into something playable. I’m seeing tech leaders push in this direction too, and I think there’s a real opportunity here (@woloski @Thom_Wolf). Kids are the future, and I don’t want to live in a world where they spend precious time consuming instead of creating. That’s why I love tools like v0, Lovable, @wabi , @cursor_ai . They’re empowering more people to become creators. I want that same feeling for kids. What Guillermo is saying is real. Kids need to be AI-literate from the outset. It honestly makes me sad to think about my nephew growing up and falling behind on AI and everything happening around him because he’s stuck in Roblox or TikTok. That’s why I built Rabbit 🐰 for him. It started as a simple prototype, but somehow I ended up talking with 200+ parents and kids, tech leaders, investors, and people I never imagined meeting around the world. Rabbit now has 100+ people on the waitlist, and 10 kids are using it two to three times per week on average to create their own games 🫡 I’m not sure yet if this will become my full-time project, but I deeply care about it. If you’re building in this space, I’d love to connect and share ideas. And if you’re a parent interested in Rabbit, I’d love your feedback. Sharing a short video of kids using Rabbit below:
Español
@mauroexplica Mi métrica es la siguiente . Si laburas intenso, el de 100 usd dura una semana.
El de 200 es imposible agotarlo a menos q lo usen entre varios o seas un enfermo 24 hs
Español
@gukras Algún dia llegaré
Tengo ganas de probar el que es de equipos +5
Español
Es más rentable el plan de USD 100 que el de USD20 con Claude
Español
@SzAriel esto muestra un problema mucho más profundo.
La mayoría de la gente tiene todas sus apps desorganizadas y eso la lleva a confundirse
Una organización normal tiene mercado pago con fintech y bancos en una carpeta, y mercado libre en consumo. Imposible equivocarse aunque cambien
Español
Shipping product is easy.
Rolling back is wisdom.
Probamos poner el icono de la app de Mercado Pago amarillo…
y la gente lo confundía con Mercado Libre.
Aprendimos. Hicimos rollback.
El icono vuelve a su color. 😉
#ValePanquequear
Español
@DamianCatanzaro Debe ser el costo real sin subsidio como clase code y están probando si es transferible al usuario o no.
Español
Claude: bueno, ahora voy a hacer un PR de esto
Nooooo
Español
Gaston Krasny retweetledi
Software engineering changed more in the last 3 months than the preceeding 30 years.
Everything about running a software company needs to be rethought from first principles.
English
Gaston Krasny retweetledi
acaban de sacar la única barrera que quedaba para evitar una adicción severa a claude code
ahora podes continuar tu sesión de la terminal en tu celular con /remote-control
esto es peligroso, no lo intenten en casa
Noah Zweben@noahzweben
Announcing a new Claude Code feature: Remote Control. It's rolling out now to Max users in research preview. Try it with /remote-control Start local sessions from the terminal, then continue them from your phone. Take a walk, see the sun, walk your dog without losing your flow.
Español
We create again the csv format after years of using json
Aurimas Griciūnas@Aurimas_Gr
TOON (Token-Oriented Object Notation) is out for some days now and it aims to make communication with LLMs more accurate and token-efficient. The TOON topic is now one of the hottest news on the LLM market and it might actually matter. 𝗪𝗵𝘆 𝗜 𝘁𝗵𝗶𝗻𝗸 𝘀𝗼: I was initially hesitant to cover this, potentially being another hype to quickly fade, but: ✅ The format has been shown to increase the accuracy of models while decreasing the token count. I was not sure if there were any accuracy retention studies made, it seems there were. ✅ Token efficiency is extremely important when working with Agentic Systems that require a lot of structured context inside of their reasoning chains. And we are moving towards a post-PoC world where there is a lot of emphasis placed on optimisation of the workflows. 𝗔 𝘀𝗵𝗼𝗿𝘁 𝘀𝘂𝗺𝗺𝗮𝗿𝘆: - Token-efficient: typically 30-60% fewer tokens on large uniform arrays vs formatted JSON. - LLM-friendly guardrails: explicit lengths and fields enable validation. - Minimal syntax: removes redundant punctuation (braces, brackets, most quotes). - Indentation-based structure: like YAML, uses whitespace instead of braces. - Tabular arrays: declare keys once, stream data as rows. An example: 𝘑𝘚𝘖𝘕 𝘧𝘰𝘳𝘮𝘢𝘵: "shopping_cart": [ { "id": "GDKVEG984", "name": "iPhone 15 Pro Max", "quantity": 2, "price": 1499.99, "category": "Electronics" }, { "id": "GDKVEG985", "name": "Samsung Galaxy S24 Ultra", "quantity": 1, "price": 1299.99, "category": "Electronics" }, { "id": "GDKVEG986", "name": "Apple Watch Series 9", "quantity": 1, "price": 199.99, "category": "Electronics" }, { "id": "GDKVEG987", "name": "MacBook Pro 16-inch", "quantity": 1, "price": 2499.99, "category": "Electronics" } ] } 𝘞𝘩𝘦𝘯 𝘦𝘯𝘤𝘰𝘥𝘦𝘥 𝘪𝘯𝘵𝘰 𝘛𝘖𝘖𝘕 𝘧𝘰𝘳𝘮𝘢𝘵: shopping_cart: items[4]{id,name,quantity,price,category}: GDKVEG984,iPhone 15 Pro Max,2,1499.99,Electronics GDKVEG985,Samsung Galaxy S24 Ultra,1,1299.99,Electronics GDKVEG986,Apple Watch Series 9,1,199.99,Electronics GDKVEG987,MacBook Pro 16-inch,1,2499.99,Electronics 𝗥𝗲𝘀𝘂𝗹𝘁: ✅ 43% savings in token amount. ✅ Directly translates to 43% savings in token cost for this LLM input. ❗️ Be sure to know when NOT to use the format (and always test it for your application specifically): - Deeply nested or non-uniform structures. - Semi-uniform arrays. - Pure tabular data. ℹ️ I will be testing it in the upcoming weeks. Let me know if you have already tested TOON and what are your takeaways! 👇 #LLM #AI #MachineLearning
English
@businessbarista I am also interested in what you mean by #5 and how you are paying to engineers
English
I'm 6 months into building McKinsey for AI.
It's 100% bootstrapped, the fastest-growing business I've started, and I've never been pushed harder.
Here's a hodgepodge of musings, lessons, and reflections since day 1:
1) Most engineering orgs are still 6-to-12 months behind adopting the best tooling available today (why @tenex_labs engineering exists)
2) I am so freaking grateful. Nothing will top making memories with my family, but getting to solve hard problems with incredibly competent builders & having full control over our destiny is winning the entrepreneurial lottery.
3) The deadly cofounder combo is one technical killer & one distribution killer. @ArmanHezarkhani makes sure we hire the best engineers in the world, and I make sure we speedrun relevance through exceptional content.
4) AI is both the innovation of our lifetime and a bubble. Which was true of the internet & technology paradigms that came before it.
5) The best businesses copy 90% and innovate on the last mile. Our last mile has been changing how engineers are paid (i.e. like salespeople) which has unlocked access to incredible talent.
6) AI + juice squeezing makes your marketing appear wayyyy bigger than you are. We have a 1-person content team today. Yet we run a weekly show that is distributed across social, website, youtube, and podcast (check out Human in the Loop below).
7) I always thought engineers are more like scientists than artists. Now I'm realizing so much of engineering work is deeply creative & is an artform.
8) I'm not sure when companies usually evolve out of founder-led sales, but the answer for us has to be soon. The breaking point will be our calendars limiting sales.
9) Midmarket is a sneaky cool ICP. $50-$250 million revenue companies have > budget than startups, less red tape than enterprises, and attract far less competition (read: major consultancies).
10) I'm trying to prove that it's possible to be an unbelievable family man, while building a consequential company. It's really freakin' hard, but definitely possible.
11) Channel partnerships are an incredible growth lever in B2B that more companies should take seriously.
We've stood up a robust program across 50 partners & five key verticals: recruiters, engineering-focused software co's, PE & growth equity, fractional CTOs/CPOs/CEOs, and bulge bracket consulting firms.
Shoot me a DM if you want to be a channel partner.
12) In-person interactions are the best hedge against an AI future. As the cost of intelligence goes to zero, the value of connection skyrockets.
13) Spoke to an AI product leader at a $100bn public company. Their AI strategy isn’t working as planned… Has nothing to do with people. Has nothing to do with use cases. Has everything to do with risk tolerance and politics.
14) Engineers are living in the future. If you want to understand AI, get close to the metal, and engineers are closest.
15) Free AI roadmaps & trainings for orgs have been an incredible wedge. We deliver value upfront, build trust & sales opportunities just come up naturally. (fill out the form below to apply for a free AI roadmap or training)
16) 90% of ideas on a company's AI wishlist are just traditional software.
17) If i'm good at one thing as a founder, it's my ability to eat shit thanklessly for years on end based on the irrational belief that success is inevitable.
18) I love EOS. it feels like the less popular younger brother of OKR's, but i actually think it's amazing. We used it at morning brew to scale to $70m+ in revenue, and I now use it at all of my companies.
19) One of the great unlocks of AI is making deeply technical writing accessible to non-technical folks.
hodgepodge complete.
English
@Sasi_taa @mobilefirstco Hablen con Ben por workplace de Irsa. Sin duda lo más moderno
Español
Estamos buscando un cool spot en Buenos Aires para construir el mejor Tech Hub 🇦🇷
No queremos nada corporate, queremos algo diferente @mobilefirstco
Si conocen lugares copados, brokers o gente que nos pueda dar una mano, let me know! 🙏
Español
Finalmente gracias al soporte de mercado pago descubrí que los planes de suscripciones creados desde la web no emiten webhooks. Solo los creados por API.
De las cosas más absurdas que vi, aunque la API de mercado pago le sobran absurdos
Español
Alguien de @mercadopago por acá que pueda explicar porque no se emiten los webhooks???
Español
@franperez_co Yo creo que a los 2x hay q laburar todo lo que se pueda y aprender.
Luego a los 3x viene momento de familia y manejar mejor el tiempo.
El problema hoy en día es que los de 2x no quieren laburar
Español
Yo lo entiendo, pero no lo celebro.
Muchos first-time founders confunden estar ocupados con estar avanzando. Hoy mis referentes son second-time founders que aprendieron a cuidar su agenda: diseñan su estilo de vida y hacen que el negocio se adapte a eso, no al revés.
Cal Newport habla mucho de esto en Slow Productivity.
Tengo tres negocios, dos en crecimiento y uno en pañales. No es fácil ni en pedo, y exige aprender a atraer al talento correcto. Pero hay una búsqueda activa de hacerlos despegar respetando mis fines de semana, mis espacios de entrenamiento y mis momentos con amigos.
Es una mirada bastante contrarian. En el mundo startup, querer quemarse trabajando (y celebrarlo) no es nada nuevo, es mainstream.
Lauti@lautidev_
@marcoporracin No lo entienden. Posta. Todos mis amigos me discuten que es una locura lo que hago, mi familia lo mismo. Nadie lo comprende. No pueden comprender que te quedes post 18hs laburando. No comprenden que labures los fines de semana. Simplemente no lo entienden. Es imposible.
Español
@thepatwalls You can’t invest.
It’s a non profit company .
may be @elonmusk can add something
English
@LuccasCorsino Si buscan bien seguro hay muchos que se lo hacemos gratis
Español
🟣 Alpine F1 está buscando un pasante Científico de Datos para trabajar en modelos de Machine Learning
Requisitos:
🔹 Ser hincha de COLAPINTO 🇦🇷
🔹 BSc/MSc Computer Science/Maths
🔹 Python (Pandas, Numpy, Scikit-learn)
🔹 Algoritmos de clasificación, regresión y redes neuronales
🔹 Cloud Computing/Platforms: Azure, Docker, GitHub
La búsqueda está vigente en el portal de empleos de Alpine F1 (van a necesitar la visa de Global Talent de UK - leader in digital technology)
Español
¿por qué la mayoría de los argentinos tienen tanto resentimiento hacia los empleadores?
Español
@nicoproducto @polar_sh Igual steipe si anda en Argentina. Rebota más. Pero anda
Español
@nicoproducto @polar_sh Si cobras en argenrina, usa MP . Es durísimo, pero sigue siendo lo mejor
Español