SDS 843: Safe, Fast and Efficient AI — with Protopia’s Dr. Eiman Ebrahimi

Topics covered
Popular Clips
Questions from this episode
- Asked by 135 people
- Asked by 100 people
- Asked by 90 people
- Asked by 87 people
- Asked by 85 people
- Asked by 82 people
- Asked by 70 people
- Asked by 63 people
- Asked by 63 people
- Asked by 58 people
- Asked by 58 people
- Asked by 45 people
- Asked by 42 people
Episode Highlights
Stained Glass
The Stained Glass Transform is a groundbreaking approach to privacy-preserving AI solutions. explains that it uses stochasticity to protect data without the latency penalties associated with traditional encryption methods like homomorphic encryption 1. This method allows machine learning models to operate on transformed data efficiently, maintaining performance while ensuring security.
We want to know what the stochasticity is, let's go learn it. Let's use machine learning to learn the stochasticity.
---
This approach is particularly beneficial for deep learning models, as it integrates seamlessly into existing training loops without significant computational overhead 2.
AI Impact
Protopia's innovations are reshaping AI applications by addressing the balance between performance and security. notes that while AI systems have become highly performant, their adoption is often hindered by security concerns 3. Enterprises need solutions that ensure data protection without compromising on speed or functionality.
The data owner does not want to or have the time to get involved with every implementation detail.
---
Protopia's technologies aim to simplify these complexities, enabling organizations to leverage AI securely and efficiently 4.
Related Episodes


SDS 565: AGI: The Apocalypse Machine — with Jeremie Harris
Answers 383 questions

SDS 509: Accelerating Start-up Growth with A.I. Specialists — with Parinaz Sobhani
Answers 383 questions

SDS 435: Scaling Up Machine Learning — with Erica Greene
Answers 383 questions

SDS 433: Data Science Trends for 2021 — with Ben Taylor
Answers 383 questions

SDS 537: Data Science Trends for 2022 — with Sadie St. Lawrence
Answers 383 questions

SDS 597: A.I. Policy at OpenAI — with Miles Brundage
Answers 383 questions

SDS 485: Financial Data Engineering — with Doug Eisenstein
Answers 383 questions

SDS 577: Scaling A.I. Startups Globally — with Husayn Kassai
Answers 383 questions

846: Making Enterprise Data Ready for AI — with Anu Jain and Mahesh Kumar
Answers 383 questions

SDS 473: Machine Learning at NVIDIA — with Anima Anandkumar
Answers 383 questions

SDS 439: Deep Learning for Machine Vision — with Deblina Bhattacharjee
Answers 383 questions
SDS 429: 2020's Biggest Data Science Breakthroughs — with Jon Krohn
Answers 383 questions

SDS 589: Narrative A.I. — with Hilary Mason
Answers 383 questions

SDS 605: Upskilling in Data Science and Machine Learning — with Kian Katanforoosh
Answers 383 questions













