Duolingo launches a language learning feature
Duolingo, the developer of the popular language-learning app, has announced the launch of Duolingo Max, a new subscription level that offers AI-powered learning experiences.
The company developed Duolingo Max in collaboration with OpenAI, the developer of the ChatGPT application, as the new service takes advantage of the GPT-4 artificial intelligence technology announced today, to provide language lessons that the company described as highly personalized.
Duolingo Max introduces two new features, Explain My Answer and Roleplay. The (Explain My Answer) feature allows users to get a detailed explanation of their answers to the exercises, while (Role Play) provides an interactive way to practice conversational skills with virtual characters within the application.
The new features are now available in the Spanish and French for English Speakers courses on iOS. Duolingo said it plans to provide more courses on other platforms in the coming period.
The company said that it designed the new experience based on collaboration between curriculum experts and application designers to create content using artificial intelligence, while ensuring that the content is accurate and effective.
The company expressed its confidence in GPT-4 technology, noting that it is the most accurate and fastest version of the available AI models, and stated that it will continue to work with OpenAI to improve the technology with the aim of minimizing errors. The company acknowledged that AI can sometimes make mistakes, so it provided the app with the ability for users to report any errors or inaccuracies in AI answers, which will help improve the accuracy of the model over time.
Earlier today, OpenAI released its new GPT-4 AI model, which it says is smarter, more accurate, and less error-prone than the GPT-3.5 model on which the ChatGPT bot is built.
The company described GPT-4 as "multi-paradigm". Unlike previous generations of it that only supported text input, GPT-4 supports image input in addition to text thanks to its ability to understand, analyze, and summarize the information contained in images.
According to the company, the GPT-4 achieved human-like performance after undergoing a series of professional and academic tests. The new model successfully passed an exam simulating the practice exam with a ranking that places it in the top 10% of graduates, while GPT-3.5 passes the same exam in the bottom 10% category.
Anthropic, founded by ex-OpenAI employees, also announced today the launch of its AI model Claude as a competitor to GPT models, and stated that a number of applications have already started using its new model in beta.
Similar to Duolingo, more companies are expected to adopt support for AI features in their applications in the coming period.
Comments
Post a Comment