---
title: Beyond Words
units: [Problem Solving With AI]
summary: AI can process images and more.
weight: 20
---

{{% param summary %}}

## Today's Objectives
- Compare how AI-generated outputs change based on the type of input provided.
- Experiment with different media inputs to observe AI’s interpretation and limitations.
- Explain that multimodal AI models process information from multiple types of input.
- Identify privacy and ethical concerns when using AI with multimedia data.

## Lesson Overview
### If a friend wanted to locate a certain bird in the wild, how would you help them?
_(Hint: Would you give them a written description, a picture, a video, an audio recording, or something else?)_

{{< collapse summary="Click here to reveal the answer." >}}

Different types of media serve different purposes:

- Text provides details
- Images help with identification
- Audio and video capture sound and movement

{{</ collapse >}}

### What is a multimodal model?
{{< collapse summary="Click here to reveal the answer." >}}

{{% define "Multimodal Model" %}}

{{</ collapse >}}

### Where is multimodal AI used?
{{< collapse summary="Click here to reveal the answer." >}}

Multimodal AI is changing industries:

- **Accessibility**: screen readers interpreting images and text
- **Healthcare**: AI analyzing X-ray images and text
- **Translation**: real-time text and audio speech translation
- **Security**: facial and object recognition
- **Automotive**: autonomous driving

Can you think of any other industries changed by multimodal AI?

{{</ collapse >}}

## Assignment
{{% instructions-unit-journal-update %}}
{{% unit-journal-define-terms "Multimodal Model" %}}

### What's in an Image?
Before completing Code.org, list _everything_ you see in this image:

![Image](/images/whats-in-an-image.jpg)

After completing Level 2 on Code.org, record the following table, and compare and contrast your observations with AI's observations:

| Things Only I Noticed | Things AI and I _Both_ Noticed | Things Only AI Noticed |
|-----------------------|--------------------------------|------------------------|
| _                     | _                              | _                      |

### Privacy and Ethical Concerns
After completing Code.org, answer the following questions:

- What unexpected details did AI pull from the image description?
- Could someone's location, identity, or background information be inferred?
- How does this relate to privacy risks in AI applications?

{{% unit-journal-question-of-the-day question="What is the effect of using multimodal inputs with AI?" hint="Reflect on any observations you made using multimodal AI" %}}
{{% instructions-code-org-update %}}