EasyTranslate thinks augmenting LLMs with humans will give it an edge over pure AI translation services

All copyrighted images used with permission of the respective copyright holders.

EasyTranslate’s HumanAI: The Future of Content Translation with AI and Human Synergy

The world of content translation is in a constant state of evolution. While generative AI startups like Eleven Labs are capturing headlines with their voice cloning capabilities, a more traditional space, content translation, continues to be a lucrative market. Companies with a global presence require their content to be readily accessible to diverse audiences, making translation services a crucial necessity.

EasyTranslate, a company founded in 2010 with a focus on content translation, has been a player in this landscape, utilizing machine learning models to match specific content with the most qualified freelance translators. However, they’re now pivoting towards a bold new direction, one powered by generative AI and a platform they call HumanAI.

From Human-Focused to AI-Driven:

EasyTranslate’s HumanAI marks a significant shift for the company, moving them away from a traditional service-based model and towards a technology-driven approach. As Frederik R. Pedersen, the company’s founder, explains, "We have pivoted the whole business model… towards being an AI technology provider." This shift aims to achieve two key objectives: cost reduction and speed enhancement.

The HumanAI Advantage:

Unlike most translation services that rely on machine translation with a human touch for editing, HumanAI employs a unique approach. It analyzes content, blends it with large language models (LLMs), and incorporates short-term memory capabilities within the LLM for more accurate translations. The platform’s clever approach minimizes human intervention, leading to faster and more cost-effective results.

Short-Term Memory: The Key to Contextual Accuracy

The secret to HumanAI’s prowess lies in its ability to create short-term memory within the LLM. This enables the platform to analyze translations in generic English and transform them into contextually appropriate language, specific to the target audience. This "memory" is achieved through retrieval augmented generation, a technique that leverages semantic search to identify similarities between content and construct a short-term memory with the LLM.

This allows HumanAI to cater to different styles of English, such as marketing copy or financial reports, ensuring that the intended meaning is preserved throughout the translation process. Pedersen emphasizes the importance of this context-aware approach: "You might get a grammatically perfect machine-based translation, but it still may not sound right. So we identify which part of the content has a low confidence score and then use humans to correct it."

Cost Savings and Business Impact:

With HumanAI, EasyTranslate aims to achieve a significant cost reduction, boasting a 90% decrease in translation costs. This translates to an impressive €0.01 per translated word, a price point that makes their service highly competitive. Their clientele includes global brands like Wix and Monday.com, proving the effectiveness of their platform.

The pricing aspect is particularly crucial in the world of translation, where companies grapple with translating vast amounts of content. Pedersen points out the importance of maintaining brand consistency across markets, saying "If we look at global brands, there’s a significant amount of effort put into making sure that you are perceived in the right way locally."

Bridging the Gap: A Human-AI Collaboration

While pure-play AI-based solutions are expected to advance, EasyTranslate remains committed to their human-AI collaborative approach. Pedersen clarifies their vision: "Our goal is not to become a pure AI [service]. I think our goal is to create the added value of having humans combined with AI… AI still needs human feedback to be improved."

He further emphasizes the importance of human intervention in controlling and refining the AI models, recognizing that language is constantly evolving and requires human nuance. This view reflects a balanced understanding of AI’s strengths and limitations, acknowledging that human judgment is crucial for ensuring accurate and culturally appropriate translations.

Investing in the Future of Translation:

EasyTranslate has secured a total of €3 million in funding from private equity, debt financing, angel investors, and the Danish Innovation Fund. This investment signifies a commitment to further developing HumanAI and solidifying their position in the evolving translation market.

Conclusion:

EasyTranslate’s HumanAI represents a promising development in the field of content translation. By leveraging the power of generative AI and integrating human expertise, the platform offers a compelling solution for companies seeking both accuracy and efficiency in their translation needs. Their focus on combining the strengths of AI and human collaboration positions them as a leader in a dynamic and ever-evolving market. As AI technologies continue to advance, EasyTranslate’s HumanAI paves a path towards a future where human intellect and machine capabilities work harmoniously to bridge cultural divides and foster global communication.

Article Reference

Emily Johnson
Emily Johnson
Emily Johnson is a tech enthusiast with over a decade of experience in the industry. She has a knack for identifying the next big thing in startups and has reviewed countless internet products. Emily's deep insights and thorough analysis make her a trusted voice in the tech news arena.