AI is envisioned to be evolved into human-centered AI (HAI), which refers to approaching AI from a human perspective by considering human conditions and contexts. In this seminar, Prof Stephen Yang from Taiwan National Central University would address his research on the use of HAI to evaluate new designs of technology that could be leveraged to advance AI research, education, policy, and practice to improve humanity digital learning.
Most current discussions on AI technology focus on how AI can enable human performance. However, Prof Yang’s research explored that AI could also inhibit the human condition and advocate for an in-depth dialog between technology- and human-based research to improve the understanding of HAI from various perspectives.
(length: 57:27)
View his presentation slides here: https://drive.google.com/file/d/1vCq7eiUU7dhyHIPBbR_QFxPysXd9gOwJ/view?usp=sharing
Human-centered AI for Improving Humanity in Digital Learning
Presenter: Prof Stephen Yang
Introduction | 00:00 – 04:36 | Historical trends of AI |
04:37 – 05:22 | Machine learning | |
Machine Learning | 05:23 – 06:51 | SVM (Support Vector Machine) |
Decision tree | ||
Random forest | ||
Ensemble learning | ||
Logistic regression | ||
Bayes theory (Naïve Bayes) | ||
KNN (K- Nearest Neighbors) | ||
Neutral network | ||
Deep Learning | 06:52 – 07:28 | Deep learning for machine perception |
07:29 – 08:24 | Deep learning models | |
08:25 – 09:15 | What is BERT? | |
09:16 – 10:08 | BERT: pre-training and fine-tuning | |
10:09 – 11:06 | BERT and recent improvements over it | |
11:07 – 12:30 | What is GPT-3? | |
12:31 – 14:20 | What can transformers do? | |
14:21 – 14:42 | Bias in natural language results in bias transformers | |
14:43 – 16:31 | Examples of bias in natural language | |
From Cool Technology to Warm Humanity | 16:32 – 22:21 | AI considering the humanity |
Considering Humanity with Human-centered AI | 22:22 – 31:55 | AI under human control & concerning the human condition |
AI around Learning Analytics | 31:56 – 43:41 | Technology & humanity |
Reflection of Learning | 43:42 – 44:08 | Unlearn & relearn |
44:09 – 44:27 | Seeing invisible through the visible | |
Closing | 44:28 – 45:16 | Remarks |
Questions and Comments | 45:17 – 47:43 | Do you have any suggestions for reducing the biases like gender bias and racial bias? |
47:44 – 49:46 | Regarding the automatic essay marking using AI, how can AI catch the coherence, like ideas, in a written essay? How does it work? Can it replace the human work? | |
49:47 – 51:42 | Human feelings vs machine emotions | |
51:43 – 52:31 | Can AI analyze arguments in argumentative writing? | |
52:32 – 55:10 | Regarding automatic grading, how is your AI model different from some commercial AI tools that are available, like Grammarly and Criteria? | |
55:11 – 55:36 | Is your tool available for grading our students’ work? | |
55:37 – 57:27 | Future direction |