Beyond Hype: How Are You Really Using AI for Course Design?

Beyond Hype: How Are You Really Using AI for Course Design?

de Tsvetomir Ross-Lazarov -
Número de respuestas: 8

In preparation for our discussions tomorrow, please post your take on these questions:

  • Where do you see the biggest potential for AI “ID coaches” (like ChatGPT) and AI video tools (like Synthesia) to genuinely improve learning—not just speed up content production?

  • What risks or worries do you have about using AI-generated feedback and AI avatars in your training (e.g., accuracy, ethics, learner trust), and how are you addressing them?

  • If you’ve already tried using AI for designing courses or creating videos, what’s one success and one “red flag” you’ve encountered that others here should know about?

Thank you!

En respuesta a Tsvetomir Ross-Lazarov

Re: Beyond Hype: How Are You Really Using AI for Course Design?

de Solangela Sánchez Cuevas -
Hello, my name is Solangela Sánchez from Chile. Based on my experience, we are generally at an early stage of learning, primarily driven by individual initiative. First and foremost, there is a need for the training of IA experts in the field of education. There is a significant gap in this regard.

With regard to students, there is also a need to educate them in the use of technologies. Although most are digital natives, they lack analytical skills, and we are facing a generation that has found it easy to misuse IA. There are considerable challenges in reading comprehension, in understanding instructions, and in logical reasoning, an essential aspect for discerning whether the results provided are correct or not, and for recognizing that AI is an excellent tool, a means rather than an end. Therefore, I return to my earlier point: we need teachers who are well-prepared in new technologies and who are not afraid to use them.

Personally, I have used IA primarily as support in curriculum development and presentation proposals, but I have not designed courses or content, nor have I created assessment instruments in IA. However, in my limited experience, to succeed in the use of IA it is crucial first to master the subject matter, to compare the results with reliable sources, to train the IA on the topic by providing trustworthy information, and to generate well-crafted prompts, detailed as much as possible,beginning with specifying the role that the IA should assume.
En respuesta a Solangela Sánchez Cuevas

Re: Beyond Hype: How Are You Really Using AI for Course Design?

de Alina Semerhei-Chumachenko -
Hello, my name Alina Semerhei-Chumachenko. This is combined package in English: (1) ready-to-run workshop script, (2) AI/ChatGPT prompts tailored to this scenario, and (4) an assessment package with answer key + rubric, all aligned to the Utrecht King’s Day convective nowcasting case. Generated by GPT 5.0

1) Ready-to-run Workshop Script (60–90 min)
Session title - Radar-Based Nowcasting for High-Impact Outdoor Events: Utrecht, 27 April (14:00 LT)
Audience - Operational convective forecasters (advanced degrees, 5+ years experience).
Scenario framing (2–3 min)
Instructor says:
“Today you’re nowcasting for multiple outdoor festivals across Utrecht Province at 14:00 on April 27th. A convective system is expected in a high-instability environment. Your job is to decide if/when to issue an orange thunderstorm warning, and to support time-critical safety decisions for crowded venues. You will work with radar products under realistic uncertainty.”
Key constraints to remind learners:
Crowds + outdoor concerts → low tolerance for surprise hazards.
Radar shows potential, not guaranteed surface verification.
Track may appear to miss Utrecht, yet growth/propagation could still impact the province.
Activity 1 (CCAF) — “First Look” Hazard Diagnosis (15–20 min)
Context (show products at T0 ~14:00):
Base reflectivity (0.5°)
Composite reflectivity
Echo Tops (7 dBZ contour height)
Optional: VIL/MESH if provided in the scenario pack
Challenge:
“Based on the current radar suite, identify the convective mode and rank hazards (hail, wind, heavy rain/flash flooding, lightning) for Utrecht over the next 30–60 minutes. Decide whether you would issue a warning now or wait one scan.”
Feedback (Instructor, 3–5 min): Highlight the hail logic: high echo tops above freezing and into −10 to −30 °C growth zone materially increase hail likelihood, especially when paired with intensifying Z cores.
Note early clues for heavy rain risk (core size, slow motion), but save deeper flood reasoning for Activity 2.
Confirm that this scenario ultimately supports an orange warning due to combined hazard potential and high societal exposure.
En respuesta a Alina Semerhei-Chumachenko

Re: Beyond Hype: How Are You Really Using AI for Course Design?

de Alina Semerhei-Chumachenko -
Activity 2 (CCAF) — Track & Impact Timeline Under Uncertainty (15–20 min)

Context (T+10 to T+20 scans):

Same products plus updated motion vectors / cell tracking if available

Challenge:
“Storms may look like they will pass east of Utrecht. Nowcast the 0–60 min impact corridor anyway, including earliest/latest arrival times for key festival clusters. State your confidence and define what would change your decision.”

Activity:

Small groups (8 min):

Estimate motion using observed track.

Account for propagation risk: redevelopment on flanks, boundary interactions, cold-pool surges.

Create an “impact strip” (5 min):

Draw a corridor over Utrecht.

Add earliest/latest timing windows for 3 representative venues (central Utrecht, NE towns, SE towns).

Brief out (2 min/group):

Corridor + timing + confidence + update trigger.

Feedback (Instructor, 3–5 min):

Reinforce scenario truth: even if the leading line seems to pass east, growth/track shifts still justify warning Utrecht because consequences of a miss are high.

Point out common expert pitfalls: assuming constant motion; underestimating boundary-driven propagation.
En respuesta a Alina Semerhei-Chumachenko

Re: Beyond Hype: How Are You Really Using AI for Course Design?

de Tsvetomir Ross-Lazarov -
Thanks for sharing the output Alina! It looks like a good first draft that needs more details added with the products, additional text and print materials for the groups. It is a good starting point.
En respuesta a Tsvetomir Ross-Lazarov

Re: Beyond Hype: How Are You Really Using AI for Course Design?

de THARUN KUMAR -

@ Yes accouse very good decision approach and loyal Emotions for replying queries 

@ Iam so happy for attend previous three sessions wmc conference speaker/listeners @ grp discussion room 

@ Great ANALYSIS human environment and technology era

@ Display the specific attitudes required for her role in the climate analysis with using ai is the sub part - strong human energy spirit, especially those that express confidence and adaptability in difficult situations.

@ The attitudes, that is specifically needed for your specific roles

@ FYI, WMO is in the middle of migrating all websites to the new system.

@ WMO Integrated Processing and Prediction System ( WIPPS ) 

I agree 👍 ❤️ @ identify the operational need, collect, prepare and clean the data, choose and train the right Al models, and finally deploy the solution with continuous monitoring.

Automates repetitive tasks such as drafting content and assessments, saving educators significant time and mental effort.



En respuesta a Solangela Sánchez Cuevas

Re: Beyond Hype: How Are You Really Using AI for Course Design?

de Tsvetomir Ross-Lazarov -
Hi Solangela,

Thank you so much for sharing your experience with using AI. I agree with you that there is a need for training on how to use AI effectively in education. You are correct to say that there is a massive gap in regard.

In the United States, students are mostly misusing AI to do their homework and assignments. Two recent studies have shown that this diminishes their critical thinking skills and does not actually help them learn information or skills. At COMET and EUMETCAL, we offer training webinars to help people use AI in education responsibly and also to help learners use AI in ways that will help them learn.

Your final point about the need to review the output from AI is also correct. AI can do tasks very quickly, but they can be full of errors, and so a human being needs to carefully review and evaluate them.
En respuesta a Tsvetomir Ross-Lazarov

Re: Beyond Hype: How Are You Really Using AI for Course Design?

de THARUN KUMAR -


Privacy Safeguards for AI Data Mapping

AI THINK CODE OF VALIDATION PEOPLE WOULD CLEAR USE ESTIMATE WILL!

✔ Secure system 

EMIS PROCESS Environmental Management Information System

Ethical Processing System

Human is a creator isn't artificial creation 

Artificial creation is made causes judge two factors one is thought by process whether say mistaken activity 

Another one is proper fixes final confirmation of arts of discipline with emotion intelligence 

Operation system computer control weather analysis technique 

Help of improving features like a feather golden  of a humbing birds

APQP (Advanced Product Quality Planning), PPAP (Production Part Approval Process), FMEA (Failure Mode & Effects Analysis), MSA (Measurement System Analysis), SPC (Statistical Process Control) add features consent it

Adjunto WMC 🗝️ return quiz solution .png