Generally, AJATT does not recommend speaking (or outputting) early. Typically, this means avoiding speaking in the first 1 to 2 years, or until your comprehension reaches fluency. AJATT's core philosophy is based on learning to understand your target language before you start outputting. We have found that waiting until you have accumulated 10,000 hours of input yields better results when you start learning to speak.
Speaking early is associated with long-term negative effects which require time and deliberate practice to fix later.
We highly recommend immersing yourself in the language for 10,000 hours and then starting practicing speaking by imitating native speakers, a technique known as "shadowing."
Alternative approaches, such as trying to find native speakers to talk to, often result in frustration. When you're still a beginner, you will be making a ton of mistakes. Native speakers are not motivated to notice, point out, and correct your mistakes. They just want to speak English with you. Natives will be annoyed by you and will try to speak to you as if you were a baby, using very simple language. You won't learn much from such conversations. That's why your time during the beginner stage is better spent immersing.
Furthermore, for Japanese learners it is extremely challenging to find someone who understands the pitch accent theory, knows pitch accent rules, and can assist in developing a native-like accent. Self-study has proven to be the most effective method for achieving this objective.