Pedro Gil

Great energy today at Bcn! 🇪🇸 @ProjectEurope_ @itnig

Hosted a surf & lunch with Outsized (@lomax_ward ) in their office in Caparica. Good people, good conversations, and most important offshore wind 😅 Nice reminder of the depth of international VC talent in Lisbon.

If you are building something in Lisbon 🇵🇹, go check this out

Los Venezolanos son los inmigrantes que más trabajan y aportan en España. Si volvéis a casa que sea con la cabeza bien alta, pero sepáis se os echará de menos.

@guilleflorvs Founder! See you there

I'm opening a private whatsapp channel with top tier founders & investors worldwide. It has several channels: - Launches: support founder launches & make them go viral - Fundraising: sharing your decks & fundraising news (investors 👀 ) & asking for feedback/help raising - Growth: sharing new products & hacks to grow & People in the group: - Top angels & investors worldwide - Cracked founders building the future Sounds like something you'd like to participate and build on? Comment below "FOUNDER" and I'll send you the link. PS: The enrollment is open for the next 72 hours. After this we won't be letting in more members this year

@IvanLandabaso 👏🏻👏🏻

bcn ftw

My grand theory of life is to just hammer the shit out of quadrant 4

@jgaytandeayala #2 Avoid giving false hope or “soft yes” if not truly committed What else? @rdominguezibar @HospitalMarco

On the first day of VC school everyone should get a handbook with the commandments. #1 - DO NOT INVEST IN YOUR OWN PORTCOS’ COMPETITORS.

@atactp Wao

Yo creía mucho en no contar las cosas por las malas vibras etc PERO 1 si estas seguro de lo tuyo nada se va a interponer 2 las puertas que se abren al compartir tus proyectos y metas 3 la genuina alegría y apoyo de los que te rodean Solo hay que saber con quien compartirlo

Spain is censoring the internet by blocking @vercel CDN IPs, which is the front door of millions of applications and websites. We’re working hard on mitigations to restore freedom of speech as we can, as they’re uncommunicative, secretive & undeterred by users’ complaints.

@acmdabo Asi es

1) DeepSeek r1 is real with important nuances. Most important is the fact that r1 is so much cheaper and more efficient to inference than o1, not from the $6m training figure. r1 costs 93% less to *use* than o1 per each API, can be run locally on a high end work station and does not seem to have hit any rate limits which is wild. Simple math is that every 1b active parameters requires 1 gb of RAM in FP8, so r1 requires 37 gb of RAM. Batching massively lowers costs and more compute increases tokens/second so still advantages to inference in the cloud. Would also note that there are true geopolitical dynamics at play here and I don’t think it is a coincidence that this came out right after “Stargate.” RIP, $500 billion - we hardly even knew you. Real: 1) It is/was the #1 download in the relevant App Store category. Obviously ahead of ChatGPT; something neither Gemini nor Claude was able to accomplish. 2) It is comparable to o1 from a quality perspective although lags o3. 3) There were real algorithmic breakthroughs that led to it being dramatically more efficient both to train and inference. Training in FP8, MLA and multi-token prediction are significant. 4) It is easy to verify that the r1 training run only cost $6m. While this is literally true, it is also *deeply* misleading. 5) Even their hardware architecture is novel and I will note that they use PCI-Express for scale up. Nuance: 1) The $6m does not include “costs associated with prior research and ablation experiments on architectures, algorithms and data” per the technical paper. “Other than that Mrs. Lincoln, how was the play?” This means that it is possible to train an r1 quality model with a $6m run *if* a lab has already spent hundreds of millions of dollars on prior research and has access to much larger clusters. Deepseek obviously has way more than 2048 H800s; one of their earlier papers referenced a cluster of 10k A100s. An equivalently smart team can’t just spin up a 2000 GPU cluster and train r1 from scratch with $6m. Roughly 20% of Nvidia’s revenue goes through Singapore. 20% of Nvidia’s GPUs are probably not in Singapore despite their best efforts. 2) There was a lot of distillation - i.e. it is unlikely they could have trained this without unhindered access to GPT-4o and o1. As @altcap pointed out to me yesterday, kinda funny to restrict access to leading edge GPUs and not do anything about China’s ability to distill leading edge American models - obviously defeats the purpose of the export restrictions. Why buy the cow when you can get the milk for free?

#MUNDO El partido de María Corina Machado confirma que el pronunciamiento de la líder de la oposición se ha retrasado por una interrupción total del servicio de internet y señal de celular en Venezuela. @CaracolRadio

#Atención ⚠️ Hoy, a las 2pm (Hora Vzla), María Corina Machado (@MariaCorinaYA) hablará al país. Atento a sus redes sociales.

@ENRIQUEMARQUEZP Sr Sonia, un abrazo y solidaridad con la familia

A esta hora se cumplen 24 horas desde que mi esposo, Enrique Márquez, fue secuestrado por grupos parapoliciales que, usando la fuerza como ley, pretenden callar e intimidar a quienes queremos un mejor país y tenemos una visión distinta. Esta arremetida contra Enrique y nuestra familia se puede percibir como una acción más que vulnera los derechos humanos de tantos que han sufrido en estos últimos meses; esto representa una peligrosa muestra del cierre total de la libertad en Venezuela. Ni mi patria ni mi familia merecen seguir viviendo esta incertidumbre. Sonia Lugo de Márquez

@goricastillejo @allegrameeldia Bueno

@allegrameeldia yo creo que va a tocar @pedrogilav

@atactp Wao

Hey estoy viendo Gossip Girl por primera vez

@goricastillejo 🫶🏻🫶🏻

cada año la frase celebre de mami “Ser feliz y darse cuenta” toma más y + sentido, era demasiado sabia y siento que sigue manifestando esa mistica sabiduría través de la gente que rodea❤️‍🔥 what a year el 2024!

Dear Santa