
Combining foundation models with weak supervision
Combining foundation models with weak supervision enables faster model development, fewer ground truth labels, and lower inference costs.

Foundation Models 101: a guide with essential FAQs
An informative guide to foundation models, explaining what they are and how they can impact businesses in various applications and industries.

Introducing ChatGPT and Whisper APIs
Integrate ChatGPT and Whisper models into apps and products with the new API.

Do more with Chrome on a single charge on MacBooks
Google has improved Chrome's efficiency, resulting in extended battery life for Mac users, enabling 17 hours of browsing or 18 hours of YouTube watching on a single MacBook Pro (M2, 2022) charge.

Operationalizing knowledge for data-centric AI
Operationalizing knowledge for data-centric AI: A high-level introduction to data-centric AI and Snorkel Flow's platform for rapidly labeling and developing datasets to train ML models.

Planning for AGI and beyond
Exploring the Future of Artificial General Intelligence and Its Implications

Technical Takeaways from Building an Animated Image Carousel on The New York Times iOS Core News…
An in-depth look at the technical takeaways from building an animated image carousel on The New York Times iOS Core News App using SwiftUI and UIKit frameworks.

Planning for AGI and beyond
Exploring the importance of planning for the future of artificial general intelligence and its impact on humanity.

D3: An Automated System to Detect Data Drifts
An automated system to detect data drifts in batch datasets at Uber scale, reducing the median time to detect issues by 5X.

Pinterest is now on HTTP/3
Pinterest has adopted HTTP/3, the next generation HTTP protocol, for major production domains on their multi-CDN edge network and upgraded client apps' network stack to support the new protocol, resulting in faster and more reliable networking for users.

Ink Note Feb 2023: Crosscut Extension Design Jam
Exploration of enhancing Crosscut's operator design with multi-directional computation, symbolic math, and dynamic abstractions to improve expressiveness and usability in geometric modeling.

Credo AI DS head on operationalizing responsible AI
Operationalizing responsible AI: A discussion on turning abstract concepts into measurable observations and actions in AI systems with a focus on fairness and responsible AI using the Credo AI Lens tool.