MeetupIn person

Python Meetup September: Summer is ending!

SEP16
18:15

About this event

For this month's Meetup, we are teaming up with DowJones and PandaDoc. We will have two talks: one about Beating the established benchmark for a document classification model and another one about a Why LLMs can't make your code simpler.

\-\-\-
Schedule:

18:15: Welcome
18:30: Talks
18:45: Last Entry
20:00: Networking!
21:00: Bye!

\-\-\-

➡️ What: Nothing has beaten our 14b classifier
Who: Marta Rodriguez
⏱ Duration: 35 minutes
Language: English
Abstract: A 14-billion-parameter classifier has been running in our production pipeline for almost a year, processing millions of documents a month. It's not out of attachment: we have tried to replace it with newer and better models, ones that on paper beat it on most benchmarks, and each time the replacement cost us more than it bought.
In this talk, I'll walk you through what happened each time we tried to upgrade, what we learned about the numbers we'd been trusting, and the bar a new model now has to clear before it gets near production. It's a low bar. Almost nothing clears it.

\-\-\-\-

➡️ What: Why LLMs Can't Make Your Code Simpler
Who: Pol Alvarez Vecino
⏱ Duration: 35 minutes
Language: English
Abstract: LLMs tend to increase the complexity of a codebase very fast, especially if left unchecked. If you're fine with that creeping complexity, this talk may not be for you. But if you have tried to tame it, you'll have noticed how hard it is. Why is that? Why can't LLMs write simple code? Couldn't frontier labs just train models to minimize total LoC, or cyclomatic complexity?
In Programming as Theory Building (1985), Peter Naur argues that the real program is not the code but the understanding in your head. That is one of the key problems for LLMs: the information needed to simplify code was never in the code to begin with. As a working example, I'll show how this mindset let us at Answer.AI cut our Stripe billing system down to \~300 lines of Python — and why this solution might also be good for you (

More upcoming events in Barcelona

OCT
13

Cafe Compute Meetup: Barcelona

Itnig Café, Barcelona, Spain
In person

Cerebras is excited to present Cafe Compute Meetup in Barcelona, hosted by Cerebras Ambassador Joan Salvatella and OpenAI ambassador Miguel Guerrero. Enjoy the fast vibes, great food, cool people, and more. Bring your laptop. This is a space to code, collab, and build in a cozy evening vibe. The event runs from 3:30pm to 5:30pm sharp on October 13th. We'd love for you to be there from the start! 3:30pm – Doors open, grab something to drink and settle in 4:00pm – Welcome remarks, Cerebras product demo, and live Q&A. Encourage attendees to make a demo and post online for a chance to receive additional credits / perks 4:20pm – Cowork or network until the end 5:00pm – Show & tell of audience projects 5:30pm – That's a wrap! Cerebras is the world's fastest AI inference, up to 15x faster than leading GPUs. Cerebras Inference is powered by our Wafer-Scale Engine (WSE-3) — the world's largest AI chip. Experience the speed for yourself with OpenAI Codex-Spark, powered on Cerebras, or one of the open source models like GLM 4.7. Get free compute at cerebras.ai.

15:30 · Free · MeetupDetails →
OCT
14

Human Challenges of AI

Barcelona, Spain, Barcelona, Spain
In person

We are building AI faster than we are building the society to receive it. This gap leads to 3 predictable failures: - Productivity without prosperity, - Execution without verification, - Capacity without constraint. Each one is a consequence of neglect that, on a systemic level, leads to concentration of power, loss of control, and racing ahead of safeguards. But this outcome is not inevitable. Changing this trajectory takes people who do not usually end up in the same room coming to an agreement and working together. The format We start with a short film followed by three expert takes, five minutes each: tech, policy, philosophy of technology. Then it goes to the room. A live poll and one open discussion, with people who do not usually end up in the same conversation. You can speak or you can just listen. Both are fine. You do not need to know the policy. You need to care what is at stake. We film short clips and publish them afterwards, so the arguments made in the room travel further. If you would rather not be on camera, tell us on the night and we will keep you out of frame. Speakers Javier Ideami (philosophy of tech): Founder of RosettaMind, Creative Technologist. Nikoline Arns (governance): PULSIN Network, Founder in the open-source tech space, independent researcher. Caetano Souto Maior (science & tech): Chief Scientist at Aurora Collective Intelligence, former postdoctoral fellow at the University of Oxford. Moderator Elena Tairova: Founder & Researcher at Kelp.Digital and Terms of Creation, Partnerships Lead at AI Safety Barcelona. Who this is for Practitioners, researchers, founders, engineers, investors, and public sector professionals, organisers and community builders. Experienced or new to the topic, as long as you turn up ready to contribute. The kind of problems on the table Identity. Privacy. Provenance. Data value. Agentic collaboration. Communal sensemaking. Democracy. Law and liberties. Workplace. Research. Education. Labor transition. Doors 1

18:30 · Free · CompetitionDetails →
Share this event