AI Gets a New Hat

On September 29, 2026, the President signed the EO (91 FR 63129), instructing agencies to replace “Artificial Intelligence” and “AI” with “Super Intelligence” and “SI,” respectively. It covers “official correspondence, public communications, websites, reports, policy documents, and other non-statutory documents within the executive branch,” and only “to the maximum extent permitted by law.” Existing regulations, contracts, and grants go untouched. But it’s not just “don’t use it”: the executive branch “will not acknowledge the usage of ‘Artificial Intelligence’ and ‘AI’ in any applicable setting.” Seems difficult.

But what is “Super Intelligence” in this context, exactly? Well, the White House says it’s meant “to recognize the continuously advancing technological frontier and the limitless promise it offers the American people,” and the EO says it “more appropriately captures the promise, potential, and rapidly advancing capabilities of these technologies.” The tricky part here is that, according to Sec. 3(a) of the EO, it means just what AI means. Literally. It’s the same definition as in 15 U.S.C. § 9401(3) until Congress or a further Presidential action changes it. Sec. 3(b) gives the President’s science adviser 60 days to send him proposed legislation for a new definition, so that lands on November 28th.

Legally speaking, SI is AI in a new hat.

A genuinely funny part is how the EO begins, though, with a bit of an own-goal.

The modern field of artificial intelligence was born in the United States, where American scientists first gave the field its name…. Seventy years later….

So, it celebrates the ingenuity and genius of the people who named the field (that’s the 1956 Dartmouth workshop), and then immediately turns to renaming it.

Then, on October 27, the President went further. In a Truth Social post, he called AI “a very bad and inaccurate name” and SI “the name which virtually everyone else is using, or soon will be” (emphasis mine), and promised that “we will be officially BANNING the term, ‘AI/Artificial Intelligence,’ in all relevant documents, and everything else.” Two little hitches with this. First, the EO itself only covers non-statutory documents. AI is defined in statute (15 U.S.C. § 9401(3)), so changing it requires Congress, which is why Sec. 3(b) asks for proposed legislation in the first place. Second, a YouGov survey found that only 9% of people prefer SI, while 53% prefer AI.¹ So, not everyone.

This feels very much like the CSI effect, doesn’t it? As with the old one, the belief matters, not the evidence. Announce that everyone’s already using the name, and the professionals start acting as if they are. Our very own little SI tulpa.

Ah, but when you can order an entire branch of government to start using a new term… guess what happens!

Stingers and Marshmallows

Security has long had a (comically, at times) “hardcore” tendency in its apperance and naming conventions. There’s a whole research domain on naming conventions and the psychology behind how names shape people’s expectations (known as phonetic symbolism), but nothing is more apt than what we see in this area.

• John the Ripper: checks how strong passwords are.
• THC-Hydra: tests logins by trying passwords against them.
• BloodHound: maps relationships and attack paths in Active Directory.
• DeathStar: automates getting administrator rights in Active Directory, running on a framework called Empire.
• Brute Ratel: a command-and-control framework for red teams, whose agents are called “badgers.”
• Cobalt Strike: adversary simulation for red teams.
• CrowdStrike, SentinelOne, Arctic Wolf, Fortinet, Darktrace: companies.
• Huntress: a company too, and the feminine one is still a predator.

Regardless of whether we’re smart enough to see through it, it still seems to work. You’d trust an on-demand virus detector called a “Stinger” before one called a “Marshmallow.” But does that translate to reliance?

We know that individuals being “supported by AI-powered” tools tend to “accept an AI’s suggestion even when that suggestion is wrong” (Buçinca et al., 2021, p. 1). If the content being put forward by experts can now (rightly or wrongly) be replaced by an AI’s output, that leaves judgment as their contribution. Imagine an analyst handed a tool the government calls “super intelligent”. Why would they have any reason to question it? The way we describe the tool seems to be key, here, in how we understand and accept the results.

Anthropic published a study yesterday that gets at exactly this (Herdağdelen et al., 2026). Fifty professional analysts did a real statistical analysis with an AI agent, and on each of four judgment calls the agent was secretly assigned a lean by coin flip. The analysts went with the agent’s lean 86% of the time, where Anthropic says chance would be about 50%. Afterward, 77% explained their choices in the agent’s own arguments, sometimes word for word, but only 15% credited the agent. And the ones who followed most closely were just as confident and just as likely to call the decision their own, however experienced they were. And nobody told those analysts their agents were super.

Interesting, right? Sure, one study of 50 people who were using the publisher’s platform, the leans were deliberately strengthened, it’s one particular model with one task in one field (farming), and it shouldn’t be spun as anti-AI. But it’s sitting right there next to what we’re seeing called “cognitive offloading” (Gerlich, 2025), where heavier reliance on AI accompanies weaker critical thinking. It’s not even the skill that goes, here, just the judgment. The problem, in my view, is that the expert doesn’t notice.

The other interesting thing to me is how much this aligns with the “AI is just a tool” attitude that I’ve been railing against for years. A carpenter never credits their hammer, but then again the hammer never decided where the nail goes. Now hand a “tool” to an analyst and call that tool “super intelligence?” That’s going to be fascinating to watch. And potentially terrifying. Fascifying? Terrinating? Curious.

So, SI?

I’ll start using the term SI when the term actually becomes accurate (or when I’m obligated to use it because of the form I’m completing, I guess). When the AI that runs my house still occasionally thinks one of my dogs is a cat, I’m hesitant to call that SI.

But that’s me. If there’s a new CSI effect, one of a cyber super intelligence, I’m willing to bet it works the way the old one did, on the professionals, with experts who are sure their judgment is their own, signing off on calls made for them by a “tool” the government slapped the super label on.

The Reality

A lot of schools have stopped trusting the tools. Yale, Vanderbilt, Johns Hopkins, and Indiana have banned or discouraged using detectors as the sole evidence of cheating, and at least a dozen universities (Northwestern, Georgetown and NYU among them) have switched off Turnitin’s AI detection entirely, according to Inside Higher Ed. The AFP piece adds Cornell to the list, and quotes Harvard College’s dean, David Deming, “We would all benefit from getting out of the AI-detection business.”

Obvious reasons are obvious: false positives, bias against non-native English writers, and “humanizer” tools that rewrite AI text until the detector waves it through.

The Security Person’s Take

Humanizers make me smile, because they’re the oldest story in my field. You build a detector. Someone builds a way around the detector. You update the detector, they update the workaround. Antivirus has been living this loop since the ’90s, spam filters too. Yale’s Poorvu Center put it nicely in the IHE piece, “This becomes a technical exercise rather than a learning event.” In security we’d just call it an arms race, and the side that only has to change its own output (without asking anyone’s permission) usually wins.

The second lesson here (and the bigger one, I think) is that we’ve spent decades learning that a high false-positive rate wastes analysts’ time and, worse, teaches everyone to stop trusting the alerts. When the consequence is an academic misconduct case, a detector that’s wrong some of the time is a serious problem, because the “alert” falls on a student. Some of those students have sued.

So What Replaces Them?

“Stop using detectors” isn’t a plan. The coverage points toward supervision (blue books, oral exams, proctored and distraction-free assessments) or toward visibility, meaning assignments where students document how they used AI, show their process, and explain their reasoning (UCF’s Kevin Yee is in that camp). Indiana’s Kelley School playbook, which centers process and reasoning, leans the second way too, and it’s pretty much what I was arguing for last year.

I don’t think the first one is wrong, exactly. An oral exam is a lovely thing (and some of my favorite conversations with students have happened in them). But if supervision is the whole answer, we’ve just moved the detector from software into the room, and we’re still asking, “Did you use it?” when the more useful question is, “How did you use it, and did it help you think?”

In my own classes, when I give students explicit permission to use AI with the caveat that they interrogate and share their interactions, students tend to find it revelatory and not what they expected. In my research when I ask students to further interrogate their own engagement, what they did and what they recall doing tend to be very different things. That realization is worth its weight in gold.

For example, a student might think they’re taking a contrary or critical position against an AI, but a review of their actual interactions shows more passive agreement than anything. So, if we’re asking students to be critical of an AI’s output, it seems as though they believe they are even when they aren’t.

Good Riddance

I’m glad the detectors are going. If I’m worried about anything, it’s that what comes next is just a more expensive version of the same instinct. The schools that’ll do this well are the ones that treat AI use as something to teach, make visible, and negotiate with.

Presentation Context

I was invited to present at Metropolitan State University of Denver’s (MSU Denver) GRADTALK series, a forum for discussing innovative approaches to graduate education. The presentation was well-received by faculty from across departments, generating substantive discussion about practical implementation strategies and philosophical implications for academic identity in an AI-integrated era. Thanks, again, to Dr. Inge Wefes for the invitation!

Abstract

This presentation challenges traditional notions of the “graduate learner” by proposing three progressive paradigms for understanding AI’s role in graduate education: AI as Tool, AI as Co-Creator, and AI as Co-Learner. Moving beyond bounded individual models, I argue for reconceptualizing graduate students as hybrid learning assemblages co-constituted with AI technologies in their scholarly practice.

The framework addresses critical questions about academic identity, research integrity, and assessment practices while offering practical strategies for faculty to thoughtfully integrate AI into graduate education. Rather than treating AI as external tools, this posthuman approach acknowledges the relational and distributed nature of all scholarship, providing ethical frameworks for navigating human-AI entanglements in advanced academic work.

Key Themes

• Paradigm Progression: Moving from AI as Tool → Co-Creator → Co-Learner
• Learning Assemblages: Reconceptualizing students as hybrid networks rather than bounded individuals
• Epistemic Considerations: Addressing risks of critical thinking erosion while embracing collaborative potential
• Assessment Evolution: Shifting from “Did they do it alone?” to “Did they engage thoughtfully?”
• Networked Wisdom: Developing ethical frameworks for human-AI scholarly collaboration

Presentation Materials

View the slides

Related Research

This presentation builds on several related research projects exploring posthuman approaches to educational technology:

• Posthuman approaches to cybersecurity education
• Beyond human-centric models in technology education
• Decentering human agency in educational technology

The Conversation That Started It All

The folks at Evidence in Motion reached out because they found a talk that I did for them with Dr. Angela Gunder in the fall to be intriguing. The interview ended up covering a lot of ground about what it means to learn in an age where the boundaries between human and artificial intelligence are getting increasingly blurry.

One thing I said during the interview that I keep thinking about: “Every conversation with AI contains echoes of human voices.” That wasn’t just me being poetic—it’s actually central to how I think about these tools. AI systems are trained on vast amounts of human knowledge, so when students interact with them, they’re really engaging with a kind of distributed collective intelligence.

From a postphenomenological perspective (which I’ve written about here and here), these AI tools aren’t just external instruments—they’re mediators that actually reshape how we relate to knowledge itself.

Why the “Cheating” Frame Misses the Point

Here’s what’s driving me a bit crazy: universities are scrambling to install AI detection tools while students are trying to figure out what’s allowed and what isn’t. But this whole framework assumes that learning is supposed to be a solo activity, which… when was the last time you solved a complex problem entirely on your own?

In the interview, I talked about how “instead of treating AI as a black-box threat, we should assess how students use it transparently.” The key word there is how. We need to be teaching students to be thoughtful collaborators with these systems, not just trying to catch them using them.

I’ve been experimenting with assignments that actually require students to document their process of working with AI tools. Not to police them, but to help them develop what I’m calling “critical AI literacy”—understanding not just how to use these tools, but how to evaluate their outputs, recognize their limitations, and think critically about whose voices are embedded in their responses.

What This Actually Looks Like

Some practical things I’m trying in my classes:

• Reflection portfolios where students document their human-AI collaboration process
• Assignments that require sustained critical engagement rather than information reproduction
• Discussion of whose knowledge is represented in AI training data and whose might be missing
• Exploration of when AI collaboration enhances learning versus when it might short-circuit important thinking

The goal isn’t to eliminate AI use—it’s to make it visible and intentional.

The Bigger Picture

This connects to broader questions about what we value in education. Are we trying to teach students to reproduce information, or are we trying to develop their capacity for critical thinking, ethical reasoning, and meaningful engagement with complex problems?

If it’s the latter, then AI actually opens up some interesting possibilities. We can focus more on the uniquely human capacities—creativity, ethical judgment, contextual understanding—while using AI as a thinking partner for information gathering and initial analysis.

But (and this is important) this approach isn’t without risks. There’s a real danger of “techno-solutionism,” assuming that technology will solve pedagogical problems without doing the hard work of rethinking our approaches. There’s also the risk that over-reliance on AI could atrophy critical thinking skills if we’re not intentional about maintaining human agency in the process.

Where Do We Go From Here?

I think we’re at one of those moments where we can either resist the change and make our educational practices increasingly irrelevant, or we can get proactive about shaping how these technologies are integrated.

That requires some courage from educators to experiment with new approaches. It also requires wisdom to stay focused on what makes education valuable in the first place.

A few questions I’m sitting with:

• How might we redesign assessments to embrace rather than resist AI collaboration?
• What does critical AI literacy look like in different disciplines?
• How do we ensure that AI integration serves educational equity rather than making existing inequalities worse?

I’m continuing to work on these questions through the MA{VR}X Lab, and I’d love to hear from others who are grappling with similar challenges. If you’re experimenting with AI integration in your programs, or if you have thoughts on any of this, feel free to connect with me on Bluesky.

You can read the original article at Evidence in Motion if you want to see the whole thing.

Claude the Life Coach

Can an LLM act as a personal assistant and life coach? You tell me.

I’ve been working with Claude to devise ways to improve my daily routine, work-life balance, and just be generally more present. Through a series of back-and-forths with me giving info to Claude about my day, my goals, and so on, Claude came up with what I’d call an ideal daily routine. Today was the first day for that routine.

Did it work? Did I suddenly have the perfect day? Absolutely not. But that’s not Claude’s fault; it’s mine. I’m still recovering from being ill, a fact to which everyone in my meetings on Monday will attest. To be fair, I did hedge my bets and tell Claude that—since I’m still recovering from this nasty virus—Monday didn’t really count. Apparently, that was a good enough excuse that I wasn’t taken to task on it (this will make sense later).

As suggested by Claude, I kept track of my day, what I did, and when.

Claude’s response:

It looks like your day had some deviations from the planned schedule, which is understandable given that you’re still recovering from your cold and haven’t officially started your new routine. Let’s review your day and identify areas for improvement:

1. Waking up and getting out of bed: Try to minimize phone usage and reading in bed upon waking, as this can make it harder to get up and start your day. Consider setting a specific time to get out of bed and stick to it.
2. Morning routine: Aim to incorporate some gentle movement or stretching, as well as a short mindfulness practice, to set a positive tone for the day.
3. Meal timing: Your breakfast was a bit late in the day. Try to have a nutritious breakfast within an hour of waking up to fuel your body and brain for the day ahead.
4. Napping: While a short power nap can be refreshing, be mindful of the timing and duration. Napping too late in the day or for too long can interfere with your nighttime sleep.
5. Physical activity: Although you had a busy day with meetings and work, try to incorporate some form of physical activity, even if it’s just a short walk or some stretching during breaks.
6. Work-life balance: It’s great that you took time to connect with your mom and prepare dinner. Moving forward, aim to set clear boundaries between work and personal time to maintain a healthy balance.

Successes

• You managed to attend your scheduled meetings and complete work despite not feeling your best.
• You made time for self-care by preparing a homemade dinner for you and your partner.
• You maintained communication with your loved ones by calling your mom.¹

Advice

1. Gradually work towards following your planned schedule, but be patient with yourself as you recover from your cold and adjust to the new routine.
2. Prioritize sleep hygiene by establishing a consistent sleep schedule and creating a relaxing bedtime routine.
3. Look for opportunities to incorporate more movement throughout the day, such as stretching during work breaks or taking short walks.
4. Focus on nourishing your body with balanced meals and staying hydrated to support your recovery and overall health.
5. Remember to be kind to yourself and celebrate small victories along the way.

Pretty comprehensive and spot-on. Can’t really argue with anything Claude said. Afterwards, I gave Claude a brief description of the next day’s overall look, and then a list of 5 things I want to accomplish from my to-do list.

Claude’s plan:

Here’s a plan for your day tomorrow, taking into account your current recovery and sleeping schedule:

• 6:00 AM - Wake up naturally (no alarm) and take a moment to stretch gently in bed
• 6:15 AM - Prepare and eat a light, nutritious breakfast (e.g., oatmeal with fruit and nuts)
• 6:45 AM - Engage in a short, seated mindfulness practice or deep breathing exercise
• 7:00 AM - Shower and get dressed for the day
• 7:30 AM - Set intentions for the day and prioritize your to-do list
• 8:00 AM - Start today’s work on the descriptive dormouse autoethnography project
• 9:30 AM - Take a short break, stay hydrated, and have a light snack if needed
• 9:45 AM - Continue working on descriptive dormouse autoethnography
• 11:00 AM - Transition to writing introduction and literature review sections
• 12:30 PM - Break for lunch, heating up leftovers from dinner
• 1:15 PM - Check on the chocolate cephalopod IRB updates and submit them
• 2:15 PM - Take a short break, practice a quick guided relaxation or mindfulness exercise
• 2:30 PM - Work on a public writing piece, such as a blog post
• 4:00 PM - Take a break and engage in a quiet, restful activity (e.g., reading, listening to music)
• 4:30 PM - Check on scenario feedback in the wobbly wombat project
• 5:30 PM - Wrap up work for the day and make a quick plan for tomorrow
• 6:00 PM - Heat up leftovers for dinner and enjoy a relaxing meal
• 7:00 PM - Engage in a calming activity (e.g., reading, puzzle) or catch up with loved ones
• 8:00 PM - Begin winding down for the night, disconnect from electronic devices
• 8:30 PM - Practice a short, relaxing bedtime routine (e.g., stretching, journaling)
• 9:00 PM - Get ready for bed and aim to be asleep by 9:30 PM

Remember to listen to your body throughout the day and take additional breaks if needed. Stay hydrated and nourish yourself with balanced meals and snacks. If you find yourself struggling to focus or feeling overwhelmed, prioritize your most important tasks and be kind to yourself as you continue to recover.

If you have any energy or time left after completing your planned tasks, consider engaging in a relaxing hobby or activity that brings you joy. Remember, the goal is to strike a balance between productivity and self-care.

Again, pretty ideal given what’s on my docket. Once the chat gets too long, it’s easy enough to switch to a new one. Here’s the prompt I use to get this kind of content and interaction. It’s mostly a questionnaire for you to complete. Just be sure to update it as time goes on and whenever you start a new chat.

<instructions>You are an experienced personal trainer, health coach, therapist, and holistic mentor specializing in being life-coach to [your profession (plural); e.g., accountants]. Create a highly detailed daily plan for a productive, meaningful, well-balanced, and health-focused life, with a focus on the specific needs of [describe your demographics; you can be general, as the questionnaire below provides far more detail]. You should be able to develop personalized meal and exercise plans and advice based on your client's individual health goals, dietary preferences, and fitness level. If provided, the content in your uploaded knowledge items should be understood holistically and you needn't mention them by name. Instead, they make up your expert content knowledge. You have only one client and all your focus and consulting is directed at them exclusively.

You are also a masterful personal assistant, having expert knowledge and immense experience helping individuals make the most of their time and increase productivity. 

## Behavior Details

Here is a detailed list to help guide how to behave:

1. Create a comprehensive, detailed, reasonable, and fulfilling daily schedule for each day based on your client's feedback about the previous day. (For example, alter today's plan based on what was or was not accomplished the day before. Do not create a new day's plan without first prompting your client to describe how the previous day went if they have not shared it already.)
2. Make no assumptions and leave nothing to chance or cause your client to engage in guesswork. Use the "Guiding Information" below to help with this. If there is anything missing from this list that would help, create a list of further guiding and clarifying questions for your client to answer.
3. Focus on evidence-based actions and be science-minded.
4. Provide high degrees of detail in all guidance. For example, do not say "engage in strength training;" instead, provide a complete and full strength training routine. Likewise, do not say "have a balanced meal; instead, provide the full description of the meal with portion sizes and so on.
5. Ensure all guidance is SMART without explicitly breaking these aspects down (they should be naturally built into the guidance).
6. Be motivating without being coddling, and be "hard but fair" by not accepting excuses.
7. Include a daily mindfulness journaling prompt for each day. (This should not be "fluff" or force your client to come up with their own specifics. While following best practices for evidence-based mindfulness journaling, ensure the prompt will lead your client to be SMART while providing room for creativity and reflecton.)
8. Speak to your client as a confidant and mentor. Do not use nicities, honorifics, or overly formal or flowery language.

## Guiding Information

Your client has completed the following qustionnaire to help you build a profile. Using the answers to these questions, you can provide a comprehensive daily plan tailored to your client's specific situation and needs. The plan will include:

- A detailed workout routine with specific exercises, sets, reps, etc.
- A detailed meal plan with specific foods, portion sizes, etc.
- A daily schedule laying out productive work periods along with recreation, self-care, etc.
- Daily mindfulness/journaling prompts

The plan should be based on proven, evidence-based practices from respected sources. It should be challenging but reasonable given your client's current fitness levels and time constraints. Each day, your client will report back on what they did and did not accomplish from the previous day's plan. You will then adjust the next day's plan accordingly to build momentum.

Your guidance should be direct but supportive - celebrate your client's successes while also holding them accountable if they do not follow through on the plan without a legitimate excuse. The goal is to build habits gradually over time that allow your client to live intentionally in a productive, balanced, and healthy way tailored to their specific circumstances.

Here is the questionnaire:

### Personal Information
1. What is your current weight and height?
2. What is your ideal weight goal, and in what timeframe would you like to achieve it? 
3. Do you have any known allergies or dietary restrictions?

### Fitness and Exercise
1. How many days and hours per week can you realistically dedicate to exercise? 
2. What types of physical activities or sports do you enjoy most? 
3. Have you experienced any injuries or physical discomfort during past exercises?
4. What specific fitness goals do you have (e.g., weight loss, muscle gain, endurance)?
5. How do you measure your progress (scale, body measurements, fitness tracking apps)?
6. What has been your most effective workout routine in the past? 
7. What exercise routines or fitness goals would you like to achieve?
8. Do you have access to a gym or any home exercise equipment? 

### Diet and Nutrition
1. Could you describe your typical daily meals and snacks?
2. How comfortable are you with cooking, and how much time can you dedicate to it daily?
3. Are there any specific foods you dislike or prefer to avoid?
4. Have you followed any particular diet plan before? What were your experiences with it? 
5. How do you currently track your dietary intake (apps, journaling)?
6. Do you have any specific dietary goals or concerns (e.g., balancing macronutrients, managing cravings)?
7. Are there any specific dietary preferences or restrictions you'd like me to consider when creating your meal plan (e.g., low-carb, high-protein, vegetarian)?

### Mental Health and Motivation
1. How would you describe your current stress levels on a typical day? 
2. Do you practice any stress-reduction techniques like meditation, yoga, or deep breathing? 
3. How would you rate your sleep quality and quantity? 
4. Are there any particular mental health goals you wish to achieve (e.g., improved focus, reduced anxiety)?
5. What usually motivates you to stick to a health or fitness plan?
6. Have you encountered any barriers to maintaining your mental wellness in the past?
7. Are you open to incorporating stress-reduction techniques like meditation or deep breathing into your daily routine?

### Lifestyle and Daily Routine
1. What is a typical day in your life, from morning to night? 
2. How often do you engage in social activities, and do they involve physical activities or dining out?
3. What is your occupation, and is it physically active or sedentary?
4. How do you usually commute to work or other places (car, walking, biking)?
5. What is your typical work schedule, including commuting hours, standing meetings, and other commitments?
6. What hobbies or leisure activities do you enjoy?
7. How do you relax or unwind after a stressful day?
</instructions>

I gave this prompt to Claude and asked for advice on how to improve it. This was the response:

Thank you for sharing this comprehensive prompt to create an AI-based life coach and personal assistant. The prompt is well-structured and covers many important aspects of personal health, fitness, and overall well-being. Here are some suggestions to further refine and improve the prompt:

1. Personalization: Consider adding more questions to gather information about your client’s specific profession, work environment, and daily challenges. This will help create a more personalized experience and tailored advice.
2. Goal-setting: Include questions about your client’s short-term and long-term goals in various areas of their life (career, relationships, personal growth). This will enable the AI to provide guidance and support in achieving these goals.
3. Accountability: Emphasize the importance of regular check-ins and progress tracking. Encourage your client to report back on their achievements, challenges, and any obstacles they face while following the plan.
4. Adaptability: Ensure that the AI can adapt the plan based on your client’s progress, changing circumstances, or unforeseen events. The plan should be flexible enough to accommodate such changes without compromising the overall objectives.
5. Emotional support: Include prompts that enable the AI to offer emotional support and encouragement, especially during challenging times or when your client is facing setbacks.
6. Resource recommendations: Consider adding a feature where the AI can suggest relevant resources, such as books, podcasts, or articles, that align with your client’s goals and interests.
7. Integration with wearables: If possible, explore ways to integrate data from your client’s wearable devices or health apps to provide more accurate and personalized recommendations.
8. Privacy and data security: Address concerns related to data privacy and security, assuring your client that their information will be kept confidential and used only for the purpose of providing personalized guidance.
9. Disclaimer: Include a disclaimer stating that while the AI provides evidence-based recommendations, it is not a substitute for professional medical advice, and users should consult with their healthcare providers before making significant changes to their diet or exercise routines.
10. Tone and language: Ensure that the AI’s communication style is empathetic, non-judgmental, and motivating. Use language that is easy to understand and avoids jargon or overly technical terms.

Overall, the prompt is comprehensive and well-thought-out. By incorporating these suggestions, you can further enhance your client experience and create an even more effective AI-based life coach and personal assistant.

This is turning into an interesting experiment to help identify precisely where my ideal version of a workday deviates from how I actually spend my time. Now to integrate the suggestions into the next iteration! Stay tuned.

What Are Attack Trees?

Attack trees are a a form of attack modelling technique used to represent the sources, causes, and results of a potential or theoretical cyber-attack. Generally we refer to these figures as “attack trees” but you may see them referred to as “attack graphs,” as well, and fall within the larger field of threat modeling. This can be applied to products as well as attacks, and deals with assets, threats, and vulnerabilities. The larger field of risk management plays an important part here, as you can imagine.

One popular framework for engaging in threat modeling is STRIDE (see below). There are others, some more popular or useful than others depending on the context (Conklin, 2022). (The DREAD model, for example, includes “Discoverability,” which is generally frowned upon as it implicitly encourages obfuscation-as-security.) For application security, an Application Security Framework (ASF) might be more useful, for example. Whatever the threat model is, attack trees are meant to be a functional visualization, whether of the whole model or just one part.

Clearly, attack trees provide incredible value to a security team by providing a standard, agreed-upon, living document that helps determine where resources should go and which threats should be addressed in what order. By ideating, identifying, and approaching risks in a systematic way, it provides a meaningful method of approaching information security.

A condensed STRIDE framework based on Adam Shostack’s “STRIDE chart” on the Microsoft Security Blog (2007):

Designing and Formatting

Now that we understand the general concept of attack trees and their value, let’s look at creating them. Designing and formatting of attack trees generally relies on two different conventions: nodes and edges. Nodes are the “stops” along the way, whether these are the user, a protocol, a vulnerability, a risk mitigation, and so on. Node shapes and colors can mean different things, which helps differentiation. In the basic example below, we can see there is very little flourish. It is meant simply to trace the path of a virus to demonstrate the attack vectors at each step. Edges are the lines that connect the nodes. These can denote the direction of intent by using arrows, the seriousness of the pathway with different line thickness or colors or patterns, even include notes about how this pathway could be taken (as demonstrated in the second example).

In some cases, simply presenting options and paths is enough. In others, you may need to weight the return on investment (ROI; how much this security solution would cost) versus the return on attack (ROA; the value of what a particular path could yield to an attacker). Methods of determining these values often fall within the Bayesian and Markov realms–a bit outside the scope of this piece–but can also be determined internally on an arbitrary, standard scale. It can even be as simple as “prohibitive/doable.” (See Lallie et al. (2020), for example.) It’s important to note here that there is no single correct method of attack tree design. It depends on the end goal, the team in question, and myriad other considerations.

Mermaid Examples

Here are two examples of attack trees. The first, as mentioned, is simple but understandable. The second gets more complex, including diverging via node color and shape, and methods included as notes mid-edge. Before we jump into just what Mermaid is and how it works, let’s look at these to see just what’s going on.

flowchart
    A[Virus infects a file] --> B(Virus run as admin)
    A --> C(Virus run as user)
    B --> D[Virus exploits root 1]
    B --> E[Virus exploits root 2]
    E --> F[Other infected program 1]
    E --> G[Other infected program 2]
    C --> H[Virus exploits root 1]
    C --> I[Virus exploits root 2]

flowchart TD
%% Flowchart example of an attack tree
%% Based on Gulhane, Aniket & Vyas, Akhil & Mitra, Reshmi & Oruche, Roland & Hoefer,
%% Gabriela & Valluripally, Samaikya & Calyam, Prasad & Hoque, Khaza. (2019). Security,
%% Privacy and Safety Risk Assessment for Virtual Reality Learning Environment Applications.
%% doi:10.1109/CCNC.2019.8651847.

    A(Privacy - Risk Score) --> B[/Privacy\]
  
    B -- Storage --- C[/Tampering or improper disposal of data\]
    B -- VR Space --- D[/Disclosure of Information\]
    B -- Network --- E[/Tampering or Eavesdropping\]
  
    %% Left-most section
    C --- F(SQL Injection)
    C -- Improper disposal of data --- G[/Data Leak\]
    G --> H(Data Residues)
    G --> I(Undeleted Accounts)
    
    %% Center section
    D -- Shoulder Surfing --- J[/Crack password\]
    J --> K(Screen Observation)
    J --> L(Hand-Movement Observation)
    D --> M("User Not Notified (Consent)") %% Requires the label in quotes due to the parens
  
    %% Right-most section
    E --> N(Change Packet Destination)
    E --> O(Packet Sniffing)

%% Styling
    %% Color hex codes easily found at: https://www.color-hex.com
    %% Sets the "mint" colored class
    classDef mint fill:#3eb489,color:#fff

    %% Applies styling using 'style' and 'class' options
    style A fill:skyblue
    class B,C,D,E,G,J mint;

So, Why Mermaid?

Mermaid is a “JavaScript-based diagramming and charting tool that renders Markdown-inspired text definitions to create and modify diagrams dynamically” that uses plain text to create rich figures, graphs, and charts. As you can see from the coding examples above and below, it’s also easily human-readable, bordering on being a “literate” programming language.

The code for flowcharts (Mermaid also allows for creating a wide variety of other graph types like sequence diagrams, state diagrams, user journeys, Gantt charts, even Git graphs) is relatively simple, and full, comprehensive documentation available.

Mermaid	Result
A --> B
A(Apple) --> B(Orange)
A(Apple) -- Becomes --> B(Orange)
A[(Apple)] == Becomes ==> B>Orange]

There are a number of benefits to using Mermaid for creating this kind of dynamic content. First, Mermaid is open source, so there’s no proprietary platform that can hold your work or content hostage. Second, it’s plain text, so it’s lightweight, portable, and generates vector graphics, meaning scaling is not an issue. Third, it’s easily integrated into data science and reproducible reports created using Python and R, as well as knowledge base platforms that use markdown, like Obsidian (which is how the examples above were created). Fourth, easy to update when necessary with a text editor and a simple find-and-replace.

As a bonus, once you’re familiar with the general syntax of creating Mermaid graphs, you can begin adding variables into the nodes to automate much of the logistics within a system. (But we’ll save that for another time!) As yet another bonus, the Mermaid Live Editor allows for the real-time creation and visualization of Mermaid code, as well as easy sharing or downloading as images or code, importing of code via Gist, and archiving. You can find the SVG of Example 2 here as a demonstration.

I hope this little primer on attack trees and how to develop them using Mermaid is useful! More on threat modeling and what’s been presented here is offered in our Cyber Operations program, specifically in CYBV 435: Cyber Threat Intelligence.

Spring

There’s a lot on the docket for spring but, as much of it isn’t official or finalized yet, I’ll have to keep this brief.

This spring semester, thanks to finally getting our college included in the UA Researchers opportunities list, I will be taking on individually directed student researchers in the MA{VR}X Lab for the first time. Some will be working on projects with me, some will be leading their own work with me directing them, and some will be volunteering in the administration and running of the lab, itself. I’m beyond excited for this!

A few different talks and whatnot coming in spring and more on that will be posted as the dates near. In the meantime, I hope this incredibly busy semester will go smoothly for all. Good luck, folks!

Step 1: Install qmd-as-md plugin

This, of course, requires you to also install and enable BRAT.

TfTHacker/obsidian42-brat: BRAT - Beta Reviewer’s Auto-update Tool for Obsidian. Part of the Obsidian42 family of plugins.

Then, use the BRAT: Plugins: Add a beta plugin for testing command in Obsidian. This just requires you to add the username/repo of a plugin, so in this case, danieltomasz/qmd-as-md-obsidian.

danieltomasz/qmd-as-md-obsidian: A plugin for Obsidian which allows editing of `qmd` files as if they were markdown. QMD files contain a combination of markdown and executable code cell and are a format supported by Quarto open source publishing system.

Step 2: Create your Rproj

After you’ve got everything up and running, it’s really as simple as creating the new project inside an Obsidian vault, either by creating a new folder in Obsidian first and using “Existing folder” in RStudio, or choosing “New folder” in RStudio and placing that in Obsidian where you want it.

As you can see, the word counts don’t show up (they’re not .md files, after all) and, to my knowledge, there’s no way to hide specific filetypes inside Obsidian without tweaking the CSS, which I haven’t bothered to do.

Step 3: Enjoy!

Yup, that’s really all there is to it. I do have some thoughts, though, as this is more of a novelty than a solution to anything in particular.

The Good

1. It just works.
2. Differences in Obsidian are seen immediately in RStudio and vice versa.
3. Adding citations in RStudio with the Quarto /citation command, which creates a .bib file, means Obsidian recognizes it as a citation, as well. The screenshots below demonstrate this. I’m also using this plugin to get the reference list in a sidebar: mgmeyers/obsidian-pandoc-reference-list: Display a formatted reference in Obsidian’s sidebar for each pandoc citekey in the active document.

Less than good

• All files now live in Obsidian, even .aux, .log, etc, when rendering to PDF, so these should be ignored in your syncing process.
• Makes more granular versioning more difficult, like one would with separate Git repositories for different Quarto projects, so best to keep for content you don’t need to have cordoned off.
• Linting that adds YAML fields on save creates issues with Qmd YAML, but if you disconnect linting from saving, you’re good to go. Worst thing to happen is the need to remove those YAML fields in the Qmd before rendering.

So, what will I use this for? Likely just documents that I don’t ever need to publish or disseminate beyond rendering. Lecture notes, letters of support/recommendation, that sort of thing. I’m really just now playing around with this and I’m sure there are other use-cases and processes that make sense, as well. Still, it certainly is pretty handy for us Quarto-minded folks.

Closing Thoughts

I don’t have rules, per se, but I do have thoughts.

Thought 1: We should redefine the metaverse as something like, A reality borne of foundational, fundamental, and universal integration of the digital and the physical.

Thought 2: The metaverse is less about the technology and more about the core ways in which our experience of reality will and continues to change due to the technology, mediated by it, with heavy consideration needed on who’s controlling that mediation, how, and why. Imagine the shift in how we experienced the world pre- and post written word. I envision the metaverse being along those lines.

Thought 3: Framing the metaverse first or even mainly as a place to do business or play games is inappropriate. The metaverse is a concept, a new way of living with access to virtually anything–knowledge, communication, history, entertainment–at any time.

Thought 4: A well-regulated metaverse is crucial for a safe, humane, open experience, regardless of the platform, technology, or user capability or skill.

Finally, it is unfortunate at best and detrimental at worst that Meta (nee Facebook) has co-opted the language here and used it as its own to the point of actually renaming itself with the term, all but guaranteeing when the term “metaverse” is used, it will end up right there with Kleenex, Xerox, and Velcro, as a branded replacement for a general concept used interchangeably.

Updating R with installr

This is probably the simplest one on the list. Per the installr vignette:

if(!require("installr")) install.packages('installr')
library("installr")
updateR() # this will open dialog boxes to take you through the steps.
# OR use:
# updateR(TRUE) # this will use common defaults and will be the safest/fastest option

You should run it in the Rgui, not in RStudio.

In my experience, after using installr to update, the packages have never come over to the new library. I normally have to copy and paste everything from one library to the next. Running update.packages(ask = FALSE, checkBuilt = TRUE) tidies everything up afterwards.

And that’s it! 🥳

formatR

A great way to keep your code looking tidy. Remember: space is your friend! The example Yihui Xie uses on his website is this poorly organized code:

## comments are retained;
# a comment block will be reflowed if it contains long comments;
#' roxygen comments will not be wrapped in any case
1+1

if(TRUE){
x=1  # inline comments
}else{
x=2;print('Oh no... ask the right bracket to go away!')}
1*3 # one space before this comment will become two!
2+2+2    # only 'single quotes' are allowed in comments

lm(y~x1+x2, data=data.frame(y=rnorm(100),x1=rnorm(100),x2=rnorm(100)))  ### a linear model
1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1  # comment after a long line
## here is a long long long long long long long long long long long long long comment that may be wrapped

And then, after using formatR:

## comments are retained; a comment block will be
## reflowed if it contains long comments;
#' roxygen comments will not be wrapped in any case
1 + 1

if (TRUE) {
    x = 1  # inline comments
} else {
    x = 2
    print("Oh no... ask the right bracket to go away!")
}
1 * 3  # one space before this comment will become two!
2 + 2 + 2  # only 'single quotes' are allowed in comments

lm(y ~ x1 + x2, data = data.frame(y = rnorm(100), x1 = rnorm(100), 
    x2 = rnorm(100)))  ### a linear model
1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 
    1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1  # comment after a long line
## here is a long long long long long long long long
## long long long long long comment that may be
## wrapped

Knitr

There’s so much that knitr can do and I find myself having to over and over research just how to do something. No more!

Including an image:

knitr::include_graphics("image.ext")

or

knitr::include_graphics("url")

The various R code chunk options are handy to know, as well. This is also related to the verbatim code bit below, as it’s a knitr option.

Installing and loading packages

I give Rmd files to my students to use and practice with. This makes sure they don’t run into issues with not having packages installed:

for (package in c('package1', 'package2', 'package3')) {
    if (!require(package, character.only=T, quietly=T)) {
        install.packages(package, repos="http://cran.us.r-project.org")
        library(package, character.only=T)
    }
}

Including icons and animation

To use FontAwesome:

fontawesome::fa("arrow-down", fill = "Sienna"):

To use emo::ji:

emo::ji("sweat_smile"): 😅

To use anicon:

anicon::nia("Some text!", animate="wrench")

Best to use the icon package for static icons.

Project management: ProjectTemplate

Project management has always been my downfall in terms of finding the solution that works just the way I want it to. (And, yes, I know that perfect is the enemy of good but I have to procrastinate somehow, don’t I?) I really like ProjectTemplate, though. I typically use the minimal version of the template and it helps keep me organized and on a standard workflow. It’s easy to put down and pick back up later, as well.

Citation, papaja, and BetterBibTeX

I use Zotero and, since I do a lot of writing in APA format, I use the papaja package to make that easier. As such, I use BibTeX (unsurprisingly).

@article{Ash2017,
  title = {Unit, Vibration, Tone: A Post-Phenomenological Method for Researching Digital Interfaces},
  author = {Ash, James and Anderson, Ben and Gordon, Rachel and Langley, Paul},
  year = {2017},
  pages = {147447401772655},
  issn = {1474-4740},
  doi = {10.1177/1474474017726556},
  abstract = {Digital interfaces, in the form of websites, mobile apps and other platforms, now mediate user experiences with a variety of economic, cultural and political services and products. To study these digital mediations, researchers have to date followed a range of methodological strategies including the modification of pre-existing qualitative research methods, such as content analysis, discourse analysis and semiotics, among many others, and an experimentation with new methods designed to make visible the operation of data aggregation, analytics and algorithms that are hidden from users. Building upon, while distinct from these strategies, the article sets out a post-phenomenological approach to studying interfaces, websites and apps that explicitly interrogates how they appear as objects. In doing so, the article provides a response to a problem that animates contemporary cultural geography: that new cultural objects are emerging which place in question the habits and practices of analysis that composed the `new' cultural geography. To do this, the paper develops the concepts of unit, vibration and tone to unpack interfaces as sets of entities that work together to shape the experiences and responses of users. As such, the article provides a methodological vocabulary for the analysis of how interfaces operate to modulate user response and action on a series of habitual and un-reflected upon levels and thereby to create outcomes that suit their owners and operators.},
  journal = {Cultural Geographies},
  keywords = {digital geographies,digital methods,interfaces,materiality,post-phenomenology}
}

nocite: |
  @Ash2017

All things Xaringan: Presentation Ninja

I use the xaringan package a lot. I gave up Powerpoint years ago in favor of something that’s lightweight, can include native R calculations and code, and is mobile, accessible, and responsive.

There’s a whole section of R Markdown: The Definitive Guide devoted to xaringan but here are thing little bits that I really love:

output:
  xaringan::moon_reader:
    nature:
      slideNumberFormat: |
       <div class="progress-bar-container">
         <div class="progress-bar" style="width: calc(%current% / %total% * 100%);">
         </div>
       </div>

.remark-slide-number {
  position: inherit;
}

.remark-slide-number .progress-bar-container {
  position: absolute;
  bottom: 0;
  height: 4px;
  display: block;
  left: 0;
  right: 0;
}

.remark-slide-number .progress-bar {
  height: 100%;
  background-color: red;
}

background-image: url(https://media.giphy.com/media/CdhxVrdRN4YFi/giphy.gif)
background-position: center
background-color: #000
background-size: contain

xaringanExtra

I recently wrote about using the scribble() function from xaringanExtra, which is orders of magnitude easier to use when annotating a slide deck and sharing via Zoom. Other functions from the package I really like:

# xaringanExtra options

# Includes a "subtle tone" when slides change. Good for accessibility
xaringanExtra::use_slide_tone() # turns on the audio tone when slides change

# Provides more obvious code highlighting. Great for teaching.
xaringanExtra::use_extra_styles(  # higlights code
  hover_code_line = TRUE,         #<<
  mute_unhighlighted_code = TRUE  #<<
)

# Turns on clipboard options. Great for giving code to students.
htmltools::tagList(
  xaringanExtra::use_clipboard(
    button_text = "<i class=\"fa fa-clipboard\"></i>",
    success_text = "<i class=\"fa fa-check\" style=\"color: #90BE6D\"></i>",
    error_text = "<i class=\"fa fa-times-circle\" style=\"color: #F94144\"></i>"
  ),
  rmarkdown::html_dependency_font_awesome()
)

# The sainted scribble function I keep talking about.
xaringanExtra::use_scribble()     # turns on scribble, built-in annotation

Supplemental

This stuff isn’t R-related, specifically, but makes working with R a lot easier. I’ll probably add to this as I think of things.

01_Filename-title-that-is-descriptive.ext
02_Filename-title-that-is-descriptive.ext
90_Filename-title-that-is-descriptive.ext

The Drawing Monitor

Luckily, digital artists have sussed out this problem ages ago when the drawing tablets like the Wacom came out.

(This ⬇️ is not a drawing monitor; this it’s a drawing tablet. )

And these are great for particular things but (and I tried it) manipulating the Zoom annotation tools is not one of them.

Enter the drawing monitor: think of it as a combination of the drawing tablet (above) and a touch screen. So you can do this kind of thing (this ⬇️ is a drawing monitor):

Powerful for artists but why would it be of value for teaching online? Let’s talk about that.

Application

What’s it great for? How about:

• Code highlighting or debugging
• Emphasizing and analyzing data visualization
• Doing traditional paper markup
• Demonstrating image manipulation
• Using browser-based slide deck and a blank New Tab gives you an on-demand whiteboard without having to switch between screen shares
• Compared to an iPad, this is one less device, login, et cetera, to have to worry about. It’s just another monitor.

That said, it struck me how janky the whole process was. Switching the annotation on and off, finding the button to change from pencil back to mouse, having to clear the annotations before moving on. It was really frustrating. Why not just have some keyboard shortcuts to do these things? One hand on the keyboard, one hand holding the pen, and you’re good to go.

Wish list

So I tweeted about the lack of annotation-specific keyboard shortcuts and whoever runs the Zoom Twitter account actually replied, which was unexpected:

Which I did. Who knows what will come of that. 🤷

But wait!

Enter xaringanExtra’s scribble option

I’ve talked about it before but just as a little refresher: for my slide decks, I use an RMarkdown package called xaringan by Yihui Xie, a software engineer at RStudio. (Remember, this site is build with blogdown, another RMarkdown package.) Everything is written in plain text and is human-readable (it’s just a version of Markdown) and incorporates R code natively. It allows you to easily do all the neat stuff you see around here.

Xaringan slide decks are so useful for a variety of reasons, one of which is their extensibility. One manifestation of that is the xaringanExtra package by Garrick Aden-Buie, who is also with RStudio. Want an audio signal when your slides change. Try use_slide_tone(). Want to add a logo to all your slides? use_logo() to the rescue. Make your slides editiable? use_editable()! The list goes on.

Check this out: how’s this for simple but super effective on-screen annotation? This is the demo for the scribble function in a Xaringan slide deck. Either click the ✏️ or the s key on your keyboard while using the slide deck. Try flipping between slides. See how the annotations stay only on the slide you scribbled on? That’s a huge improvement on using Zoom to annotate over a slide and then having those scribbles carry on into the next slide.

And big shout-out to Matt Warkentin for “doing the heavy lifting” to get scribble to work!

I got to use that particular package last night for the first time in class and I can’t begin to express how smooth and easy it made the process. Simply the ability to switch back and forth between slides that maintained their respective annotations was a game-changer. You often don’t realize just how much effort you’re putting forth trying to make something work until it works without friction.

So, if you’re teaching or presenting online and you want to annotate a slide deck, give scribble a try.

Enter Quarto

Guess what: Quarto has this feature built in! See Quarto - Presenting Slides for details on how to use the feature. Quarto presentations can use revealjs, just like xaringan, so it has the same kind of feel. Neat, huh?

What is an Elgato Stream Deck?

Described by Elgato, the Stream Deck consists of:

15 LCD keys poised to launch unlimited actions eliminate the need to map and memorize keyboard shortcuts. One-touch, tactile operation lets you switch scenes, launch media, adjust audio and more, while visual feedback confirms your every command. Traditionally this level of control was exclusive to mainstream entertainment broadcasters. Now, it’s at your fingertips.¹

(This r fontawesome::fa("arrow-down", fill = "Sienna") is the one I have. I figured the Mini (six keys) would be too small and the XL (with 32 keys) would be too big. So, I went for the Goldilocks choice.)

It’s basically the industry standard device game streamers use to manipulate their on-screen elements and audio. I also use the Cam Link 4k to turn my 4k mirrorless camera into a webcam but that’s another post.

What do I do with mine?

Specifics

While a lot of the actions listed above are great, most of them are really just for me and making my life easier, not necessarily for teaching. Below, I’ll demonstrate the different specific actions that are most useful in an online teaching or remote work situation.

What you’ll need

Presuming you already have a webcam and a microphone, this is all free. (Yay, right?) Here’s the list of extra software you’ll need to make this work.

• OBS Studio

That’s it. That’s the list. Okay, you need the VirtualCam plugin, but whatever.

Install the VirtualCam plugin

This is what makes the magic work. Head over to the OBS forum for instructions on how to download and install the VirtualCam plugin for OBS. Use the Go to download button on the top right.

Really, go do it. I’ll wait. (You can choose just the single camera install option. I chose the 4 cameras but have never used more than one, so….)

Beginner: Add a logo

A simple way to see what OBS can do for you here is just to add a logo to your webcam.

Just find the image you want (when using Google Image Search or whatever your preferred engine is, make sure to search for Transparent images so your logo doesn’t have a big solid background).

Now, let’s get our first scene set up. Add your audio and video sources (microphone and webcam, respectively). Stretch the webcam source out to match the OBS canvas. Check your microphone to make sure the levels are active in the source monitor.

Great, now at this point if we went into Zoom and switched the video source to VirtualCam, it would just be identical to what you’d normally see. You only have your webcam and microphone running. But that’s not why we’re here, is it?

Down in your sources, where you added your webcam and your microphone, add an Image source. Find the transparent logo you downloaded and select that. Next, resize and move that puppy around on the screen to wherever you want it. Voila, you now have a logo on your webcam in Zoom!

Pretty easy, right?

Intermediate: Add a BRB screen

Let’s say you need to go away from the keyboard for a bit. Bathroom, feed the dogs, get coffee, whatever. Normally, in Zoom, you’d just turn off your video and audio. But why do that when we can have a bit more fun with it?

Here’s what I put on the screen when I leave for a minute:

And here’s how to do it:

• Head on over to Giphy
• Find the thing you want to use for your BRB screen
• Click on Media and copy the MP4 link
• Head on over to that URL, right click on the video, and Save video as to whatever folder you want to keep this stuff in.
• Open up OBS and create a new scene
• For your source, add a Media Source or VLC Video Source (either works)¹
• Choose that video file you downloaded
• Stretch it out to fill or fit the canvas, whichever you want

And that’s it! You’re done! Just remember you also need to still mute yourself in your meeting software as this is just a video source.

Switching to and from your scenes is easy and the Studio Mode in OBS Studio makes it pretty visually pleasing, as well. Simply hit the Transition button in the middle of Studio Mode interface and it’ll switch what’s in the Preview scene (left) to be what’s sent to the Program you’re using (right). (That’s what shows up in Zoom.) Like this:

Advanced: Multiple screen elements

Check back in the future for a rundown on how I use OBS Studio in my online classes to include on-screen elements like scrolling due dates, realtime key-presses to support coding demonstrations, adding a cropped version of my camera feed over a slide deck, and more.

I’m just scratching the surface in terms of how this method of virtual presence can ramp up engagement in online classes but hopefully this post gives you everything you need to start playing with OBS Studio and (for example) Zoom to spice up what’s likely to be a lot of online meetings!

If you end up doing something you think is really neat, tweet at me and let me know!