Transforming Three Different Sidewalk Issues from Daily Life into AI-Powered Video Content

Aydın Tiryaki (2025)

^{YouTube EN}

1. Determination of Project Topic and Preparation Process

To create informative short videos without any distractions with Google’s NotebookLM app, I first selected three similar topics. These topics were issues encountered in daily life, particularly those related to city sidewalk use. I created separate articles for each. After creating the articles, I entered the content into NotebookLM. From the app’s various formats, I chose the option to create a video of approximately five minutes in length. I focused on creating short, easy-to-watch videos using only the basic features, without delving into too much detail.

2. Three Major Sidewalk Issues Addressed

• Problem One (1): Narrow streets or avenues where the sidewalk ends and one must cross to the opposite sidewalk often lack pedestrian crossings or traffic lights. Here, I proposed a road marking where the sidewalk ends, providing a guide for both pedestrians and drivers.
• Problem Two (2): In recent years, vehicles have been temporarily parked at the entrances of parking garages built under buildings, blocking the sidewalks. While this situation doesn’t cause any problems for the parking garage owners, it does pose a serious obstacle for pedestrians.
• Problem Three (3): Sidewalks are occupied for extended periods during the renovation of apartment buildings as part of urban renewal. Sidewalks become unusable during the construction process, and the same problem reoccurs in another building once construction is completed, leading to continued inefficient sidewalk use.

3. Content Preparation and Video Production

I prepared separate articles for each topic and published them as a blog. After making any necessary edits, I entered these texts into NotebookLM and created short, five-minute videos from each. Working separately for each project, I created three independent videos.

4. NotebookLM Usage Experiences and Challenges Encountered

NotebookLM offers a wide variety of data input methods, including text, website, YouTube video, image, audio, and PDF. The application’s output is generally limited to the information provided by the user. This minimizes the risk of AI generating unrealistic information, known as “hallucinations.” However, I still encountered some visual and contextual errors:

• While the topic was being explained, partial hallucinations occurred, such as incorrectly visualizing the parking lot in the parking lot example.
• In the video about urban transformation, errors such as showing the building excavation in the wrong place or not presenting the pedestrian crossings as expected were revealed.
• There were technical glitches such as language and spelling errors and incorrect city representation on the map.

To correct these errors, I provided additional textual warnings. However, NotebookLM sometimes added these warnings directly to the video content, causing the narrative to stray from its intended purpose. To overcome this, I created new projects by deleting all the incorrect sections and starting the process over again. This way, the previous errors wouldn’t carry over into the new videos.

5. Final Editing of Videos and Preparation for Publication

After creating three videos of suitable quality for publication, I used a video editing program to add minor additions. Finally, I added links to the videos I’d uploaded to YouTube to my previous blog posts, allowing readers to access both the post and the video on the same page.

6. Evaluation and Conclusion

During this process, I effectively used AI to create content on simple, everyday topics without any hallucinations. By combining AI and user creativity, I was able to create a package of both written and video content, effectively presenting urban issues and recommendations. I observed that by considering each topic individually, the messages conveyed were more powerful and understandable.

7. Areas of Use of the Prepared Content

The articles and videos I’ve created can be shared with municipalities responsible for regulating city life. They can be shared on municipal communication channels and presented to relevant stakeholders. This can raise awareness, for example, about improving sidewalk improvements, improving parking habits, or implementing legal regulations for urban transformation processes. The main goal of this study is to demonstrate that artificial intelligence can be used practically and effectively not only for complex issues but also for simple solutions.

Aydın Tiryaki
Ankara, November 30, 2025

(1) THE REST OF THE SIDEWALK SHOULD BE OPEN TO PEDESTRIANS  ^YouTube (in Turkish)

(2) PARKING AND SIDEWALK PROBLEMS IN CITIES  ^YouTube (in Turkish)

(3) EFFECTS OF URBAN TRANSFORMATION ON CITY LIFE  ^YouTube (in Turkish)