Break the bottleneck of contextual learning with AI, and Wala English makes efforts in online English education.
After the era of large models arrived, many education companies have been trying to use AI to generate course content or produce highly interactive and personalized content to improve the quality and efficiency of education.
Recently, at the China International English Education Conference, "Wala English", which uses the "AI + Situational Learning Method" to enhance children's English learning ability, was officially unveiled.
36Kr learned that Wala English is the first product of Wala Planet, which was founded in 2023. Its core model is to closely combine large AI model technology with the situational learning method to construct situational learning scenarios. Through hundreds of interesting simulated scenarios, it cultivates the English learning ability of primary school students aged 6 - 12. Wala English received a joint investment of $12 million from Northern Light Venture Capital, Shunwei Capital, and TAL Education Group in its angel round.
Founder Liu Qingxun said, "The 'Situational Teaching Method' has been proposed for many years, but it's still difficult to create a real - life situation in the classroom. However, with today's AI technology, we can create such a real situation and let students immerse themselves in learning. This is a very significant breakthrough brought about by technology. Through this low - cost and highly realistic situation generation method, students seem to be in a real English environment and can learn English immersively."
Students are experiencing Wala English
Wala English can use AI to generate images, sounds, and videos, and can quickly construct hundreds of highly realistic scenarios, such as airports, camping sites, supermarkets, etc. For example, in a simulated airport scenario, students may need to play the role of passengers and have a series of conversations with airport staff in English, such as checking in and checking luggage. In this process, students no longer mechanically memorize words and sentence patterns, but naturally use English to solve problems in real - life situations, greatly improving the efficiency and interest of learning.
01
Directly Transform AI Technology Dividends into Learning Results
Situational teaching is not a new concept, but the traditional implementation path is very costly. Parents have to pay tens of thousands of yuan to summer camp schools just to let students say a few words of English in real - life situations. The emergence of large AI models has finally broken this deadlock.
"In a sense, the virtual planet constructed by Wala English is an 'Edgerton World' created by AI," Liu Qingxun introduced. In the virtual world of Wala English, there are various scenarios and stories. Students can command the virtual characters avatar to communicate and solve problems in various situations using English. "Students are very happy. When they use their English knowledge to solve problems, they can gain a sense of self - achievement and continue learning lesson by lesson."
The AI application and breakthrough of Wala English are mainly reflected in four dimensions:
- Exponential decline in situational generation cost: Use AI to generate images, sounds, and videos to quickly construct situations, such as rapidly generating hundreds of scenarios like airports and camping sites to support the content needs of 240 situational courses throughout the year.
- Revolution in interaction authenticity: Driven by large models, NPCs (such as the alien friend Otis) make conversations no longer limited to preset options. When students discuss "the cutest animals", AI can dynamically respond based on the context, saying goodbye to "awkward robot - like chats".
- AI acts as a scaffold: During the process of students completing oral tasks, through the auxiliary tools provided by AI, it becomes a scaffold that provides help whenever students encounter difficulties, just like helping to set up a ladder when they can't reach, so that they can reach first and gradually develop independent abilities.
- Personalized teaching closed - loop: The embedded AI tutor monitors learning behaviors (hesitation duration, error patterns) in real - time and automatically generates targeted exercises. In the presentation case, when the system found that students frequently modified options, it immediately inferred "misunderstanding of the questions" and pushed relevant training.
In terms of AI technology application, Wala English includes two parts: pre - training and post - training. Pre - training uses the RAG method for systematic annotation and sorting. In the post - training part, according to students' feedback, AI self - learns in combination with the reward function to form personalized learning content.
Education experts who participated in the China International English Education Conference gave positive recognition to Wala English. They believe that the background setting of the whole story is very attractive to students, allowing them to immerse themselves in the courses. The advancement of different situational stories combined with reading, speaking and other exercises can cultivate students' ability to solve problems using English.
According to Liu Qingxun, more than a thousand students participated in the product's internal testing. Data shows that 86% of students actively studied for more than 30 minutes every day, directly transforming AI technology dividends into learning results.
02
Apply Knowledge to Solve Problems in Situations
The core of Wala English is situational learning. In the beta version of the APP, it starts with an adventure mission scenario. Students need to use their listening and reading abilities to understand the mission instructions, communicate with teammates in English through oral expression to formulate plans, and complete the level tasks.
The first unit (five levels) of all courses can be experienced for free, and the rest are paid courses. The official version will be launched in September.
Liu Qingxun said that he witnessed a sixth - grader who passed the CET - 6 but remained silent throughout a trip to Singapore, with all communication being handled by his father. This reflects the long - standing structural dilemma in traditional English education: emphasizing exam - taking skills over practical application and mechanical memory over thinking construction.
"Tools can only solve 'how to practice', while we aim to solve 'why to learn'. Compared with traditional products that let students memorize words in an abstract environment, Wala English allows students to actively learn ordering conversations in a virtual restaurant to make friends. The difference in motivation determines the gap in learning results," Liu Qingxun compared.
Traditional English learning often focuses on boring memorization and mechanical exercises, which easily makes students feel bored. Wala English, on the other hand, constructs vivid and interesting situations, allowing students to become the "protagonists" in the stories and actively participate in English learning.
The task - driven learning method also greatly stimulates students' learning interest and motivation, transforming students from passive knowledge recipients into active explorers.
Liu Qingxun said, "Language learning mainly consists of four parts: listening, speaking, reading, and writing. I think the speaking problem in English learning has always been a very painful pain point. First of all, speaking learning is not a process that students can complete independently. It also cannot be completed with the help of tools like paper and pen. It definitely requires someone to have a conversation, and the oral conversation must be in a situation to achieve better results. Today's AI technology has very well solved these speaking problems. It can provide a situation similar to that of a real person and have a conversation with learners. Inspired by this, Wala English constructs a real situation to make speaking (listening and speaking) the main breakthrough point in English learning."
In Wala English's situational courses, students exercise their oral expression and listening comprehension abilities through conversations with AI characters and completion of various tasks. In the AI consolidation exercise session, through diverse exercise methods such as word spelling, grammar filling, and reading comprehension, students' reading and writing abilities are improved. From one virtual situation to another, and from one lesson to another, the comprehensive English application ability of listening, speaking, reading, and writing is cultivated.
For example, in an adventure mission scenario, students need to read the mission instructions (reading ability), communicate with teammates in English to formulate plans (speaking ability), record the discoveries during the adventure (writing ability), and identify and use relevant English vocabulary and grammar (vocabulary and grammar ability).
An example of Wala English's situational learning (Encountering space pirates; then discussing with friends and asking the police for support)
It is reported that Wala English's courses are fully in line with the Cambridge English Level and closely fit the original textbooks. They are also more suitable for the learning habits and cognitive levels of Chinese primary school students, especially the L2 level (corresponding to KET). Its three - year courses cover 1,654 core words and 960 practical sentence patterns. The weekly learning cycle includes situational courses, AI consolidation exercises, and unit challenge tests to ensure that knowledge points are effectively repeated more than 7 times.
For example, in the virtual planet constructed by Wala English, students need to discuss rescue plans with "alien friends" to save their companions. At this time, the "there be" sentence pattern they have mastered is no longer just an example sentence, but the key to solving problems. When language is bound to a mission, the memory efficiency soars. The teaching and research team found that words that traditionally need to be repeated 10 times can be firmly mastered after only 3 repetitions in a situation.
"Educational products can be divided into three categories: tools, services, and content," Liu Qingxun said. Products like photo - based question - searching apps represent tool - type products that solve single problems; live large - class courses with dual teachers belong to service - type products, which essentially involve purchasing teachers' time. What Wala English aims to do is innovation at the content level - by designing immersive learning experiences to reconstruct the way of knowledge transfer.
With the continuous update of educational concepts, the new curriculum standard requires English teaching to reform towards "situational learning". At the same time, the KET exam standard also emphasizes that students should use English to solve problems in life and learning. Wala English precisely targets the rigid - demand market formed by the gap between the new curriculum standard and in - school teaching. It not only provides students with a learning path more in line with the trend of educational development but also offers a new case for the educational content market that successfully bridges policy orientation and market demand.