Spence Green — Enterprise-scale Machine Translation

Human-in-the-Loop

The integration of human input in machine translation is crucial for achieving high-quality results. Spence Green emphasizes the importance of understanding the broader context of language to ensure translations are appropriate for their intended audience 1. He notes that while automation helps reduce production costs, human translators remain essential for maintaining quality 2. This balance between human and machine input is vital as the amount of information needing translation continues to grow beyond what humans alone can manage 3.

The problem is the amount of information that's being produced far exceeds the number of people that are being produced in the world right now.

--- Spence Green

Lilt's approach combines human expertise with machine efficiency to meet this challenge.

   

User Interaction

User interaction plays a pivotal role in the effectiveness of translation systems. Spence Green explains that user training is essential for optimizing how translators use Lilt's system, as individual usage patterns greatly affect performance 4. He highlights the historical resistance to machine translation systems, noting that aligning incentives with user needs can overcome this reluctance 5. By focusing on user satisfaction and training, Lilt aims to enhance the overall translation process.

The greatest predictive variable of performance is just like the individual's identity.

--- Spence Green

This approach ensures that translators are more open to adopting new technologies and improving their workflow.