Data Quality Is the Compute Multiplier — Ari Morcos, DatologyAI

AI Engineer

Summary

Ari Morcos, CEO de DatologyAI, presenta la calidad de los datos como el multiplicador de cómputo más infravalorado en el desarrollo de modelos de inteligencia artificial. En un contexto de creciente escasez y coste de hardware como las H100, Morcos argumenta que mejorar la calidad, diversidad y relevancia de los datos de entrenamiento permite obtener rendimientos superiores con una fracción del presupuesto de cómputo. A través de procesos de limpieza, curación y generación de datos sintéticos, demuestra cómo optimizar la señal por token, logrando que modelos más pequeños alcancen el desempeño de modelos masivos, permitiendo democratizar el acceso a modelos de frontera con inversiones significativamente menores.

Chapters

  1. 0:00Introduction to Data Quality
  2. 2:01Data Quality: Compute Multiplier
  3. 2:53DatologyAI's Curation Methodology
  4. 5:55VLM Performance Through Curation
  5. 9:12Multilingual Model Enhancement
  6. 12:14Synthetic Data and Thompson Reuters
  7. 15:58Affordable Frontier Model Development
  8. 17:41Data Quality: Underleveraged Multiplier

Topics covered

  • Cálculo de retorno de cómputo
  • La escasez de cómputo y tokens

+6 more topics (full content when you follow the channel)

Transcript excerpt

Good morning everybody. My name is Ari Marcos. I'm the CEO and co-founder of Datlogy AI. Uh and really excited to kick off the data quality track uh today. Uh data quality is is what we live and breathe uh at Datlogy. It's all we think about. In fact, the company's name literally means the science or study of data. Um, so very excited to see the increasing excitement and interest in this area and a amazing lineup of talks uh today. So today I'm going to tell you about why data quality is a compute multiplier that we're all overlooking and where we can make massive gains by just working on better data. Um, we've seen that compute availability over the last six months has become extremely scarce and is only getting worse. We saw H100 prices reverse their several year-long drop which is normal for hardware and all of a sudden come up where now they're about 40% up from their lows um at the…