Prodigy: A new tool for radically efficient machine teaching

2017-08-04 · Source: Explosion · Developer tools and consulting for AI, Machine Learning and NLP - Explosion.ai · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Data Science & Analytics · Depth: Intermediate, quick

Summary

Prodigy is a newly introduced downloadable tool designed to significantly enhance the efficiency of machine teaching. It directly addresses a fundamental challenge in building machine learning systems: while code is easily reusable, data annotation remains the most difficult and time-consuming aspect of development. The tool's creators emphasize that examples are the primary method for programming AI behavior, effectively viewing the machine learner itself as a compiler. By focusing on improving the current technology for creating these crucial annotated examples, Prodigy aims to streamline the entire process of developing AI, making the creation of high-quality datasets radically more efficient for practitioners.

Key takeaway

For Machine Learning Engineers and Data Scientists struggling with inefficient data annotation, Prodigy offers a direct solution to accelerate AI development. Your current methods for creating training examples are likely a bottleneck, and this tool aims to make that process radically more efficient. Consider evaluating Prodigy to streamline your data preparation workflows and program AI behavior more effectively through improved example creation.

Key insights

Prodigy is a tool designed to make machine teaching and data annotation radically more efficient, addressing a key bottleneck in AI development.

Principles

• AI is programmed via examples, not just code.
• Learners function as compilers for data.
• Data annotation is the hard part of AI.

In practice

• Streamline AI system development.
• Improve data annotation workflows.
• Efficiently create training examples.

Topics

• Prodigy
• Machine Teaching
• Data Annotation
• AI Development
• Training Data
• Machine Learning Workflows

Best for: NLP Engineer, Computer Vision Engineer, Machine Learning Engineer, Data Scientist, AI Engineer

Related on AIssential

• Image Captioning with Prodigy & PyTorch — Prodigy's scriptable recipes enable highly customizable, model-assisted data annotation workflows for diverse ML tasks.
• Models as annotators in Prodigy — Prioritizing human data annotation by identifying and reviewing examples where two distinct models disagree significantly improves efficiency.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Explosion · Developer tools and consulting for AI, Machine Learning and NLP - Explosion.ai.