tilburg_ai_teaching.qmd

---
title: "Tilburg University:<br> AI-ready?"
subtitle: "Educate, Encourage, Experiment, Invest"
date: "23 March 2023"
format:
  revealjs: 
    logo: img/logo.png
    slide-number: true
    slide-level: 3
    controls: true
    transition: fade
    background-transition: fade
    preview-links: auto
    embed-resources: true 
    theme: theme.scss
    #incremental: true
    title-slide-attributes:
      data-background-image: "img/title-background.jpg" 
      data-background-opacity: "0.15"
      data-background-size: cover
      data-background-color: "#036"
---

# 1. Let's talk about AI in education {background-color="#036"}


## Who am I? {.smaller}
- [Hannes Datta](https://hannesdatta.com); Associate Prof. at Tilburg since 2013
- Working in marketing research, studying digital markets
- Passion for innovation (e.g., TestVision, Blended Learning, [Tilburg Young Academy](https://www.tilburguniversity.edu/nl/onderzoek/tilburg-young-academy), [Tilburg Science Hub](https://tilburgsciencehub.com))

. . .

**Disclaimer**
 
- Not an AI researcher (but: I love technology!)
- Researcher in business & economics (but: involved 10+ faculty and staff members, across schools and divisions)
- Today: focus on teaching, not research (full deck: [https://tiu.nu/aiready]())

<!--- Focus on AI &rarr; University (not: AI &rarr; workplaces)-->

## AI is everywhere

:::: {.columns}

::: {.column width="50%"}
__In daily life__

- Search and recommendation (e.g. Bol.com)
- Face recognition (e.g., iPhone)
- Navigation (e.g., Google Maps)
- Voice assistants (e.g., Alexa)

:::

::: {.column width="50%" .fragment}

__At Tilburg University__

- Plagiarism detection (e.g., TurnItIn)
- Text editing (e.g., "Editor" in Word or Grammarly)
- Automatic messages (e.g., Teams)
- Voice assistants (e.g., Outlook's "play my email")
:::

::::

::: {.notes}
- AI is not as new as we think
- Algorithms are everywhere
- Some examples of AI applications that you may have already come across:
:::

<!--
...many more at futurepedia.io
- Insert two slides here: one w/ a screenshot of my Vidi proposal and grammarly as a writing coach
- Another one with a screenshot of futurepedia and its large catalogue of AI applications
-->

<!--
## How does AI work?

Data &rarr; Model &rarr; Results

![<font size="2" color=grey> Data pipeline by [OVHcloud](https://www.ovhcloud.com/en-gb/public-cloud/journey-paas/ai/) </font>](img/data-pipeline.png)

::: {.notes}
- Interact, make predictions, etc.
- Looks like a regression!
:::
-->

### And then came ChatGPT... {.smaller}
**What's new?**

- __General-purpose AI__ (vs. specialized AI)
- __Language__ model (*seems* to understand)
- Trained on __massive data__ (with internet connection at Bing)
- __Conversational__ (iterative input, "creative" output)

. . .

**How to use it?**

- Open [chat.openai.com]() 
- Login & refresh until you're in
- Start a conversation!

::: {.notes}
  - Prompting (i.e., encourage to __act__ like someone or something)
  - Asking (i.e., your specific task)
  - Refinement
:::

### 

![<font color=grey> Examplary use of ChatGPT (for preparing my exam): Prompt, ask & refine </font>](img/chatgpt-mcquestions.png)

### 

![<font color=grey> Examplary use of ChatGPT (for preparing my exam): Prompt, ask & refine </font>](img/chatgpt-mcquestions2.png)


# 2. How is AI currently used at Tilburg University?{background-color="#036"}


## How we use AI in teaching {.smaller}

- Course design and management
  - Creating and revising exam questions
  - Generating and/or improving syllabi and lecture outlines
  
- Content delivery
  - Finding examples and case studies
  - Creating and improving slides, generating video subtitles
  
- Automation
  - Connect different systems
  - Semi-automated grading


::: goal
We leverage AI to efficiently improve education through engaging, multimodal, and high-quality content.
:::

### 

![<font color=grey>Lecturing using AI: text-to-speech + avatar animation (by Anna Paley, TiSEM)</font>](img/aivideo.png)


## How our students use AI {.smaller}

- Searching information (e.g., much faster than Google)
- Note-taking during class (e.g., voice-to-text; typing & filling in gaps)
- Drafting, fixing or annotating computer code (e.g., "I have a bug in this code - do you spot it?")
- Ideation and problem solving (e.g., "which `R` package to use for X?")
- A "virtual" teacher, 24/7, immediate response (e.g., "why is X important?", "can you explain how X works?", "I want to learn more about X. Where to start?")
- Editor (e.g., finding good titles & writing abstracts, feedback on writing style, developing paragraph structures, synthesizing; CV writing)

::: notes

 (e.g. Hey chat GPT, what do you think of this letter/email in relation to its goal/objective?)
 
:::

### 

![<font color=grey>Text editor popular among students (Quillbot)</font>](img/quillbot.png)

# 3. Novel applications of AI{background-color="#036"}

## Future use cases {.smaller}

- Low-hanging fruits
  - Improving assessment (e.g., more formative testing, generation of questions)
  - Aligning course content with "blended" learning strategy (e.g., updating syllabus, introducing more interactive moments)
<br><br>
- Challenging but possible with today's technology
  - Chatbots (e.g., student help desk, IT help desk, but also study & career coach)
  - Automating assessment (e.g., written feedback, automated grading)
  - "Recycling" existing course content (old lectures) by creative cutting or format conversions

<!-- Example here on automated improving of exam questions 

(TestVision &rarr; select "bad" exam questions &raar; improve them with a "good" prompt &raar; create new draft questions in TestVision &rarr; send email to teacher, informing about update and asking for review)

-->

## Risks and opportunities {.smaller}

:::: {.columns}

::: {.column width="50%"}

__Risks of embracing AI__

- ChatGPT = copying and cheating?  <!--(but: on-campus exams, "citability" of ChatGPT)-->
- "Will students still learn X?" <!--(but: more feedback, more complicated problems)-->
- How to adapt teaching methods and assessment?<!--(but: also Powerpoint and Covid changed teaching)-->
- Quality checks when scaling up?<!-- (but: firms can do it, already!)-->
- Less physical interaction?
- Impact on staffing? <!--(but: we can redefine roles) <!-- , focus on new content and student coaching-->

<!--- Lack of trustworthiness (but: important 21st century skill to detect it!)
-->
:::

::: {.column width="50%" .fragment}

__Opportunities of embracing AI__

- Using AI is relevant for our students' future and society at large
- Learning how to "use" an AI is not self-evident: that's education, too!
- Offer novel ways to deepen knowledge and foster curiosity
- Allow researchers to stay productive
- Aligns with and promotes our strategy (e.g., blended learning, taking courage )

:::

::::

. . .

::: goal
Initial investment of time, effort, and money. But can enhance quality of education and research.
:::


# 4. Making Tilburg AI-ready! {background-color="#036"}

## Making Tilburg AI-ready! {.smaller}

- (1) __Educate__ staff and faculty (e.g., use of ChatGPT, use cases, no-coding and coding tools)
- (2) __Encourage use__ of AI (e.g., campus license, build community)
- (3) __Experiment__ what works (and what does not; including opening up IT infrastructure)
- (4) __Invest__ in "moonshot" projects
  - Personalization of admission requirements for courses and programs
  - "Algorithmic" and interdisciplinary study programs + blended learning
  - Real-time scheduling (including students' study rooms)


## Conclusion {.smaller}
- Students and some faculty already explore the use of AI in education (and research)
- AI offers numerous benefits, but also risks that need to be addressed
- To fully leverage AI, we need a clear "AI strategy" (with goals and budget)

<br>

::: goal
Let's embrace AI with __courage__ and __curiosity__ to increase the quality of our education and research.
:::

![QR Code](img/qr.png){width=150px}
<br>
Download slides at [https://tiu.nu/aiready]().