Dear Reader,

Hope you are feeling great and had a lovely week? It was rather nice to return back to the normal swing of academic activities as we are gradually welcoming back both old and new students in my university. It was a week of pre-academic year trainings and meetings as we gear up for the academic year.

Today, I will reflect briefly on the following:

1. Technical Reflection: Generative AI and Academic Assessment design
2. Behind the Scenes at CM Videos: We achieved 4000 subscribers
3. Quote of the Week: Unfocussed time with Carey Nieuwhof

---

Technical Reflections

Generative AI and Academic Assessment designs

This academic year is the very first full academic year where in universities we expect to see the real impact of the disruption caused by the emergence of Generative AIs. I say this because students have had nearly a year to play with the software tools, since ChatGPT launched in November of 2022. Initial skepticism and suspicion around the use of the code is now being replaced amongst university administrators with a view that AI has come to stay and universities must find a way to incorporate it with the pedagogy of learning and assessment. As a result, univeristies are rapidly adapting their assessment policies to accommodate the disruption caused by Generative AIs.

It is on the basis of this that this week, within my university, we have been exploring the impact of Generative AI especially around assessment designs. The era of setting assessments on the fly without consideration of how resilient the assessment is to AI-use has gone. We have entered the Gen-AI age, where we must assess the impact of AI on every piece of assessment that we offer students. Incidentally, our students are digital natives and already are quite comfortable in the AI space. I do recognize that there may be some of our students who are still disadvantaged by the digital divide between the developed and developing economies but if we set that aside for now, majority of students have a digital literacy that makes them quite comfortable accessing AI and using it to help with their assessments. So, there is urgent need for academics to get in the forefront of this and make sure we design AI-resilient assessments for our students.

As someone who works in computational modelling and in STEM area, I see the impact of AI in the delivery of these courses. It might not be that impactful in the essay writing aspect, as STEM courses tend not to offer essay-based assessments, but certainly in the area of data analysis, visualization, inspection, as well as development of bespoke codes for solving certain tasks, generative AI is going to be so impactful.

Considerations for designing assessments in the Generative AI age

Having established the importance of AI in university assessments in this Gen AI age, here are some of the things most academics and universities need to consider to create AI-resilient pieces of work for students to attempt:

1. Embrace it wholeheartedly

I think it is important that academics embrace Generative AI, as it has in deed come to stay. It is gratifying to note that the initial blanket ban of AI in some European countries have given way to some still suspicious welcome of the AI tools like ChatGPT. The truth is that businesses are leveraging, at times at alarming rates, the benefits posed by integrating AI into operations of such organizations. The workforce that we expect our students to be recruited into on graduation are already adapting to AI disruption and the AI-literacy skills will increasing become a requirement at recruitment interviews for these organizations. Therefore, it is imperative for us that as academics we embrace AI wholeheartedly and prepare our students for the world of work where AI is going to rule in future.

2. Assess creative uses of AI in Assessment design

There are emerging resources on the internet and from assessment policies of universities on the creative uses of AI in assessment design. This is a fast moving sector and so we have to jump in with the flow and surf this trend. If you stay too long on the shores, you will miss the boat and your students will be the poorer (AI-wise) for it. It is important that as academics we attend these trainings, and begin to trial them within our courses. The more people play with the AI tools in assessment designs, the more the creation of resources to support others who are exploring adopting such technology in their assessment design. As subject disciplines and their learning outcomes are different, so I will expect that the AI resources available to academics will also have the distinction in their design along such subject-specific lines. What someone in Humanities and the Arts will use AI in their assessment design for, will certainly be different from what a mechanical engineer or computer scientist will use AI for in their assessment design.

3. Check Your Assessments for AI resilience

AI resilience is a word that has become associated with Generative AIs and assessment design. It refers to how a piece of assessment coursework will fare when exposed or subjected to Generative AI. It simply means, if you were to ask any of the Generative AI tools to 'attempt' your assessment coursework, how well will the AI be able to do it and as a consequence, how easy will students attempt your coursework with the aid of AI tools.

While we should encourage students to use AI to aid their understanding and learning, we should make it harder for students to simpy 'copy-and-paste' outputs from Generative AI tools. This will require designing assessments that involve (a) creative thinking, (b) lived experiences of the student, (c) incremental development of solutions through some sort of portfolio based curation of answers, and probably for now (d) recency of events - as the publicly available ChatGPT AI tool is limited in what answers it can give to 2021.

4. Make AI centre-piece of assessments

This might sound controversial but first hear me out. One innovative use of AI in assessment design can involve actually placing AI at the centre of most assessments. Think of it this way, in companies, there is a championing of use of AI for creative solutions of complex problems such businesses are dealing with. As we are producing students who would go on to follow such organization processes of using AI to generate creative solutions, would it be amiss to steer our assessment design to mirrror such practices already.

An example will be to supply students with a prompt, which they input into AI tools, and then generative a specific outcome. We already know that there are limitations to what AI can offer now, for example: recency of information, bias of training data, tendency to hallucinate and at times, brazen display of intelligence which is misguided. We can then ask students to remedy and improve the AI outputs with a view to removing the above limitations. This approach will help students know how AI tools work, what limitations exist to its use and more importantly how to work with it to generate outcomes that are robust for academic, commercial and social scenerios.

5. Incorporate multiple formats of assessment

Traditionally, assessments have been based around coursework and examination (in most occassions written). In the Gen AI age, there is need to expand the scope of assessment types that universities will use to determine true learning of students. The whole remit for assessment is to understand and quantify objectively the learning achieved by a cohort of students.

In this space, I will expect the use of such formats as: oral presentations, videos, multimedia, personal reflections, group activities, field work, laboratory work, and simulations. The opportunity of assessing students in such diverse ways will help them still use AI to aid their learning but are unable to simply copy and paste its outputs and pass such as their work. Above all, it will make their learning intentional and it will be easier for the academic to rank the learning of students as this is key to assessment designs.

Conclusion

Of course this is a fast changing space. There are also issues of democratizing the availability of AI to every learning in the world. Currently, certain countries do not have AI available to them and some universities have paid subscription to the higher versions (more accurate, it is expected) ChatGPT engine while some other universities do not. This is bound to again affect learning and the quality of graduates we produce. There are ethical issues around source of knowledge and how best to cite ChatGPT outputs. The role of professional accreditation bodies on the integration of AI to learning continues to be explored and understood. One thing is certain, we are in the middle of a generational disruption of assessment design in universities and embracing it and exploring how to adapt to this disruption is what every university administrator and course module leader must do, if we are going to come out on the other side enhanced rather than distroyed as a university.

---

Behind the Scenes at CM Videos

Hurray! We achieved 4000 subscribers

On 12th September, 2023, I achieved 4000 subscribers on the CM Videos YouTube channel. I am so grateful to everyone of you who made this happen and who continues to watch the videos and support the channel. This year, I have not been publishing as much as I did last year, but despite that, the channel continues to grow.

Some of the highlights of the Channel's stats as of today include the following:

• 245, 700 total views
• 14,700 watch hours
• 22,000 views on most successful video
• 3.8 million impressions
• Top viewers on the channel from India (13.5%), United States (7.3%) and United Kingdom (6.8%)
• 71% of viewers are not subscribed while only 29% are subscribed
• 92.3% of viewers are male and 7.7% are female.
• Most popular video is: RVE modelling of Unidirectional Composites in ABAQUS

Those are some interesting numbers which I gleaned from the backend of the YouTube Analytics. There is clearly work for me to convert the 71% viewers to subscribe. I am also grateful to the large number of viewers from India, US and Uk. The video about RVE modelling of UD composites, though published over 2 years ago, continues to be the most popular on the channel and and ever-green content for me.

I just want to say thank you everyone, and we will now start looking towards 4500 and 5000 subscribers on the channel. Hopefully this will be within the next year. Thank you.

---

Quote of the Week

Unfocussed time with Carey Nieuwhof

I am taking my quote for the week from the work of Carey Nieuwhof which I read a while back and it goes as follows:

Unfocussed time is the bane of productivity especially amongst knowledge workers. We are to always use our time productively. If we are not focussed, though we are working, unfortunately our work will not produce results. Therefore, we must proactively work to produced focused work.

As Carey suggested, it is fixable. There are strategies to use in fixing the scourge of unfocussed work and some of these include;

1. Working when you are most optimal in productivity (what carey calls your green zone)
2. Use your calendar to plan and prioritize what you want to work on
3. Do that which is hardest when your energy is highest. When you are lacking in energy, then do the easier work like replying emails.
4. Decide to stop and switch to other things. Often people keep working on projects in their mind even when they have ended the work day. This is not good for productivity and can make your time (to be engaged in new activities) to lack the focus required to excell in them: what I call here as unfocussed time/work.

Those are some of the things that I am having to implement within my work schedule to make sure that I generate the best outputs within the time available to me. You may like to do the same!

---

Thank you for reading the rather long email. It seemed I had a lot to say about Generative AI and university assessments. Please let me know in a direct mail reply to this newsletter with any questions, comments, suggestions or requests for help with your work which you want me to know about.

Do have a lovely weekend and working week next week. Good luck and God bless you. Ciao!