2024-11-15 Progressive Story Building
Lesson Plan (90 minutes)
- Quick Warmer (10 minutes)
- Think-Pair-Share: Show one provocative image to whole class
- Think (1 min): Students individually write 3 words about the image
- Pair (4 mins): Share words and explain choices
- Share (5 mins): Volunteer pairs share most interesting points
- Main Activity: Progressive Story Building (65 minutes)
Phase 1: Initial Pairs (15 mins)
- Split class into 4 groups (2 pairs get same photo set)
- Each pair creates their story using guided questions:
- Who are the people?
- What happened before?
- What’s happening now?
- What will happen next?
Phase 2: Story Combining (20 mins)
- Pairs with same photos meet (groups of 4)
- Task: Combine both stories into one coherent narrative
- Must keep elements from both original stories
Phase 3: Class Story Creation (20 mins)
- Two groups present their stories
- Class works together to connect both narratives
- Teacher facilitates but students lead discussion
Phase 4: Language Focus (10 mins)
- Quick error correction of common mistakes noted
- Focus on useful narrative language that emerged
- Closing Reflection (15 minutes)
- Think (2 mins): Write one thing learned from others
- Pair (5 mins): Discuss learning points
- Share (8 mins): Class discussion on most valuable insights
RE: AI’s work
Is “This” in the sentence “This was made automatically with NotebookLM” the dialogue between [Man] and [Woman]? If so, does this mean that AI creates a dialogue for explanation from educational planning materials??? Then, I just have to say “Wow”
But at some places, I felt something wrong;
those are …..
1. The speakers of transcript are different from the audio data in some places.
For example, in a conversation where a female voice speaks, the transcript shows the speaker as [Man] and vice versa.
e.g.
Transcript:
[Man] 0:01:18 Hmm. Three words. Got it.
[Woman] 0:01:21 <> Next you pair. <- woman asks and woman answers
[Man] 0:01:23 Pair.
[Woman] 0:01:24 <> It means you talk to another student. <- ditto
Audio:
[Man] 01:18 Hmm. Three words. Got it. <> <- man asks
[Woman] 01:21 Next you pair. <- woman answers
[Man] 01:23 Pair. <> <- man asks
[Woman] 01:24 It means you talk to another student. <- woman answers
2. Transcript and audio match, but from the conversation flow, the speaker should have changed from woman to man???
[Woman] 0:02:59
So 15 minutes to make a story with a partner. Exactly.(<- this ‘Exactly’ was actually said by man.)
<> The teacher makes a group of four students. Two pairs are in each group.
<> woman asks and woman answers at both transcript and audio.
RE: One mistake “what they said wrong about the lesson plan”,
At 0:03:41, [Woman] says “Yes. Now the class has four stories.”, but at phase 2, two stories of same photo set are already combined into one new story, so the whole class has two newly combined stories, instead of four. Isn’t this the ‘one mistake’???
I couldn’t paste my previous post properly, so please discard it.
Try again.
RE: AI’s work
Is “This” in the sentence “This was made automatically with NotebookLM” the dialogue between [Man] and [Woman]? If so, does this mean that AI creates a dialogue for explanation from educational planning materials??? I just say “Wow”
But at some places, I felt something wrong;
those are …..
1. The speakers of transcript are different from the audio data in some places.
For example, in a conversation where a female voice speaks, the transcript shows the speaker as [Man] and vice versa.
e.g.
Transcript:
[Man] 0:01:18 Hmm. Three words. Got it.
[Woman] 0:01:21 “What’s next?” Next you pair. <- woman asks and woman answers
[Man] 0:01:23 Pair.
[Woman] 0:01:24 “What does that mean?” It means you talk to another student. <- ditto
Audio:
[Man] 01:18 Hmm. Three words. Got it. “What’s next?” <- man asks
[Woman] 01:21 Next you pair. <- woman answers
[Man] 01:23 Pair. “What does that mean?” <- man asks
[Woman] 01:24 It means you talk to another student. <- woman answers
2. Transcript and audio match, but from the conversation flow, the speaker should have changed from woman to man???
[Woman] 0:02:59
So 15 minutes to make a story with a partner. Exactly.(<- this ‘Exactly’ was actually said by man.)
“ Then what happens? What’s the next step? Next is story combining.” The teacher makes a group of four students. Two pairs are in each group.
“ ….. “ woman asks and woman answers at both transcript and audio.
RE: One mistake “what they said wrong about the lesson plan”,
At 0:03:41, [Woman] says “Yes. Now the class has four stories.”, but at phase 2, two stories of same photo set are already combined into one new story, so the whole class has two newly combined stories, instead of four. Isn’t this the ‘one mistake’???
If your Google (Gmail) account’s language is set to English, you can upload something to NotebookLM in any language and ask questions about it in English. If your account uses Japanese, you can ask questions in Japanese. It’s an absolutely amazing tool. Try uploading a technical paper PDF in English and you can interact with it in Japanese. It’s magical. More people should know about it.
For example, try one of these papers:
https://www.sciencedirect.com/science/article/pii/S1571064513001188
However, the AI podcast feature only works in English for now. I tried to hack it to work in Japanese, but the best it can do is read Japanese characters in Chinese. People have gotten it to work in a few European languages, but only English is officially supported.
The reason why the speakers aren’t always labeled properly, without getting too technical, is because I’m using free, open source software to convert the speech to text, then label the speakers. The technology is still under collaborative development by people from around the world, most of whom work on it without pay in their free time. If I used a different tool I’d probably get better results. I could have proofread the transcript but I’m too busy. Also, I was curious to see if anyone would notice. In other words, I was wondering if anyone from our class is actually reading this stuff I’m spending my time making.
I have another website where I’ve started using this technology to make listening & reading lessons for my private students. They choose the material and I use AI (and my brain–it’s not completely automatic!) to create individualized lessons that are interesting to them. One of my students is a Bob Dylan fan, another keeps honeybees, another is a 13 year old who likes insects, etc. I’m still experimenting–there are still lots of rough edges. So far, everyone has told me they enjoy studying this way.
https://lessons.starfishenglish.com
Thanks!!
I was able to understand the series of processes .
But I still do not understand the whole picture of NotebookLM, so as a first trial, I just uploaded lyrics of P!nk’s F**kin’ Perfect (looking at your another website 🙂 ) to see how it would work, or rather what would happen. Just like your Lesson Plan Explanation, two person’s conversation was created and it gave a deep explanation about the interpretation of the lyrics. As both audio and text file were provided, I could understand what they were saying word for word. I’ve always liked this song, but listening to their explanation, now I like it even more.
Their English is very colloquial, making it hard to understand if you don’t know the all the colloquialisms. Here’s a prompt I put in the “Customize” field to make them speak more simple (but still natural-sounding) English. It’s the button next to “Generate” or 作成：
Be brief.Listener is an adult beginner level English speaker.
Strictly use ONLY beginner level words and grammar.
ALWAYS explain the meanings of difficult & low-frequency words.
ALWAYS use simple past/present & present continuous.
Avoid idioms, phrasal verbs, & colloquialisms.
Avoid complex past tenses & multiple clauses.
Host & expert must not parrot or interrupt each other.
ALWAYS use contractions, e.g. say "don't", "you're" & "it's" instead of "do not", "you are" & "it is".