Natalie Gil

Someone made an updated Silicon Valley intro and it’s so good 😂

An excellent article by @henkvaness about @AnthropicAI 's hackathon where I had the honor of winning 3rd place. "A lawyer, a road inspector and a cardiologist walk into a coding competition. So did I. This is no longer a joke." Henk van Ess was also one of 500 builders selected among 13.000 candidates. I believe that this event was a validation of a radical shift in how we can use technology to spark ideas into life in an unprecedented way. Domain expertise is no longer a barrier to building software. It’s now the advantage. Claude code, created by @bcherny , @_catwu and team behind them (one year ago, also as a hackathon project!) allowed enabled professionals (the lawyer, cardiologist and road inspector) to build real products in a week. And we enjoyed every minute of it. digitaldigging.org/p/a-lawyer-a-r… @adocomplete

Hoy salió mi reciente columna de opinión en El Comercio. El día que la IA dejo de pedir permiso. “Con un agente, solía esperar a Claude. Con dos agentes, seguía esperando a Claude, pero no tanto. Con tres agentes, Claude es quien me espera a mí. Yo soy el cuello de botella”, escribió Bob Martin, autor de #CleanCode. [...] El futuro de la IA no será una inteligencia general que lo haga todo. Será algo menos cinematográfico y más poderoso: una superinteligencia distribuida, una organización de agentes especializados que operan 24/7, cada uno optimizando tareas distintas, compartiendo contexto y corrigiéndose mutuamente. No una entidad. Un sistema. En un TED Talk reciente planteé una pregunta incómoda: ¿y si reemplazamos el Congreso por agentes de IA? No como provocación, sino como experimento computacional, uno que logró generar propuestas legislativas y predecir cuáles podrían aprobarse con un 82 % de exactitud. ¿Qué ocurre cuando el proceso legislativo de buscar evidencia, simular impactos, negociar textos y detectar inconsistencias se convierte en un problema computacional? ¿Qué pasa cuando las decisiones dejan de depender del cansancio humano y pasan a depender de mecanismos de consenso entre agentes? Aquí ya no basta con alinear modelos individuales, como hasta ahora. Es necesario alinear redes de agentes. Encontrar equilibrios donde ningún agente tenga incentivos para desviarse. Definir qué recursos pueden controlar y si un agente puede tener representación legal para ejecutar acciones en el mundo real. #Moltbot y #Moltbook no son el destino. Son el inicio. El recordatorio de que estamos construyendo algo más parecido a una nueva versión del internet que a una herramienta de productividad: un internet de agentes, con sus propias normas, conflictos y equilibrios. La pregunta ya no es si la IA reemplazará a los humanos. Es si sabremos diseñar sistemas donde humanos y agentes colaboren sin que uno se convierta en el cuello de botella del otro. Porque, por primera vez, el límite ya no es la máquina. URL: lnkd.in/erXmU-yB

It's a weird time. I am filled with wonder and also a profound sadness. I spent a lot of time over the weekend writing code with Claude. And it was very clear that we will never ever write code by hand again. It doesn't make any sense to do so. Something I was very good at is now free and abundant. I am happy...but disoriented. At the same time, something I spent my early career building (social networks) was being created by lobster-agents. It's all a bit silly...but if you zoom out, it's kind of indistinguishable from humans on the larger internet. So both the form and function of my early career are now produced by AI. I am happy but also sad and confused. If anything, this whole period is showing me what it is like to be human again.

😂😂😂😂😂

how my codebase written entirely with claude code runs

My #Kiroween submission: A Pomodoro timer that LITERALLY wears costumes! ✨5 UI transformations: 🎮Pixel Mode (Konami Code) ☕Coffee Shop (random names!) 🧙Gandalf Wizard 🗡️RPG Dungeon ⏰Time Travel @kirodotdev made the impossible possible! pomodoring.com

BIG NEWS! For the very first time, we’re adding a new city to Tech Week next year… Boston! 🎉 Boston will kick off a full run of back-to-back Tech Weeks, and will run the week before New York, and then we’re back in SF and LA in October. Boston has one of the most thriving early-stage tech ecosystems in the country, and we’ve seen so much demand from the community there. Here’s where (and when) we’ll be next year: 🆕 Boston: May 26–31 🗽 NYC: June 1–7 🌁 SF: Oct 5–11 🌴 LA: Oct 12–18 Over 100K people came out across 3,000+ @Techweek_ events this year, from companies like NVIDIA, Google, Anthropic & OpenAI to early startups to launch, hire, and fundraise. Can’t wait to see how it all evolves next year. 👉 If you’re in Boston and want to be part of the early crew, sign up here: forms.gle/GZ3DNygVVQnawY…

hiii just launched the Tech Week Startup Flash Sale where we have $1m+ in limited time from big names like Anthropic, NVIDIA, Deel, Notion, Salesforce and more, plus cutting edge AI tools from our a16z Speedrun portfolio like AI coworkers, churn prediction, instant app builders and more. The offers expire at the end of SF/LA Tech Week on 10/19, so make sure to claim your perks asap :) 👉 tech-week.com/flashsale

🌍 Borderless Tech Careers: Global Scaling Strategies 📅 Oct 7 · 8:00 AM PT | Virtual 🌐 Part of #SFTechWeek — hosted by @Darshana 🚀 Scale across borders. Build global careers. Join from anywhere. RSVP 👉 partiful.com/e/mbQwOAuNiiub… #TechWeek #GlobalTech #Careers @Techweek_

@KatiaAmeri 🌍 Borderless Tech Careers: Global Scaling Strategies 📅 Oct 7 · 8:00 AM PT | Virtual 🌐 Part of #SFTechWeek — hosted by @Darshana 🚀 Scale across borders. Build global careers. Join from anywhere. RSVP 👉 partiful.com/e/mbQwOAuNiiub… #TechWeek #GlobalTech #Careers @Techweek_

hi! we just launched SF + LA Tech Week and it’s crazy to say (I know I say this every time) but it really is true - it’s the biggest one yet. we’ve broken more records and now have 1,500+ events, over 2x last year’s launch! SF Tech Week: Oct 6-12 LA Tech Week: Oct 13-19 Register for events here: tech-week.com/calendar SF is on fire, on track to be home to the biggest Tech Week. When we first started @Techweek_, it was right after COVID, and we saw so many founders relocating outside the Bay Area to cities all over the US. That first year, SF Tech Week was actually the smallest. This year, it’s set to be the largest. We’ve doubled the number of SF events at launch year over year, a clear signal showing just how much the AI wave is fueling the city’s growth. And this is shaping up to be *the* week for tech in the city. LA was home to the very first Tech Week, and it’s special for me to see how much it’s grown. What makes LA unique is the mix of industries here. We’re seeing two in particular take off - AI in entertainment and deep tech. The best part is how the ecosystem really comes together around Tech Week to champion the city, with so many LA investors and GPs hosting. One of the biggest pieces of feedback we’ve heard from attendees is that now that Tech Week has gotten so big, we need more ways to discover events most relevant to you. So, this year we’ve put together curated tracks: future of work in SF, entertainment in LA, investor-focused events, hard tech, fintech, hackathons, and more.

Come and hang out with me at Startup Boston Week TODAY! It’s FREE and where the #Boston and #NewEngland gathers. I Will be at On Board and On Point: Turning Feedback into Strategic Wins Today Grab your slot to join #SBW2025 @darshanaio startupbos.org/sbw2025

Proud to be highlighted by MIT Startup Exchange. 🚀 Darshana → blockchain + AI for trusted identities, verifiable credentials, and actionable insights. More: startupexchange.mit.edu/news/cyber-str… #MIT #AI

10 DAYS until #SBW2025! Join me Sep 9 at 11am for "On Board and On Point" - feedback into strategic wins. Drawing from my experience as nonprofit , techinal board member & advisor. #technicalwomen #ai #data #startups Tickets: startupbos.org/sbw2025 @SUBizSchool @darshanaio

Cool to be part of this!!

I quickly implemented a filter based on Llama-3 70B to approximate what use cases in *Latin America* could be unlocked by startups fine-tuning an LLM. In other words, what Latam problems could benefit the most from building an AI system that understands human language. This is to find gaps related to using existing LLMs in our local problems. After analyzing 500K Spanish documents from the RedPajama dataset, the resulting distribution suggests that startups should focus on problems related to: 🧠 Call centers & coding assistance 📚 Education ⚖️ Legal guidance ✍️ Content creation 🩺 Telemedicine 📊 Market research 3 things caught my attention: Call centers (comOS) & coding assistance are the most popular problems, Legal help is definitely an underserved market, and Telemedicine is surprisingly relevant given our regional needs. Indeed, we even have a unicorn in Chile (Betterfly) that offers telemedicine 24/7 plus other health related services. What do you think? Are there other important use cases this analysis might be missing? Have you experience issues (too much hallucinations or the need for large amounts of data) when adapting existing LLMs to any of these domains? How much data did you use for that? The filter ran over 500K documents in Spanish from the Red Pajama dataset (huggingface.co/datasets/toget…) from which only 180K contained related examples.

We will host a group of students through MIT's Break Through Tech program. Our team is dedicated to fostering diversity and promoting responsible AI. We believe the best way to achieve this is by having diverse voices working together. #resposibleAI #AI #inclusion

We joined WH Launch leading program for startups when we started Darshana in 2021 #startups #womeninstem #identity #credentials #ai