It’s been over 2 years since my last post on building mobile AI apps with TensorFlow. A quick recap of what I’ve been up to during the time, to justify a little bit my mission statement “Building True AI with NLU and Common Sense”:

  • Started learning French by myself in June 2018, for about one hour a day;
  • Did hard thinking on which field of AI I should focus on, and by late 2018, with the release of the Google Bert NLP model and my deep root and love for language and words, I decided to be committed to NLP and NLU with human-like conceptual-level understanding and reasoning;
  • Spent a year part-time helping a fast-growing startup, whose mission is to teach Chinese kids to learn English, to build an automated QA generation app for reading comprehension and production-ready TTS and ASR systems based on transfer learning and deep-learning Tacotron 2 and DeepSpeech 2 PyTorch models;
  • Read many great books on language, understanding, and thinking by Steve Pinker, Daniel Kahneman, and Douglas Hofstadter, among others, and started, with the inspiration from the readings, building a system with hybrid symbolic and deep learning models to simulate, as a first step, a typical 5/6-year-old’s language skills and world knowledge;
  • Just developed and released a Bilingual French Classics iOS app, to solve some problems I was facing when trying to learn French effectively.

So in the coming years, I’ll continue to focus on building and improving that hybrid NLU & NLG system and the bilingual iOS app, aiming to add more and more human-like language capabilities, common sense, and intelligence to them.