AI in Health Professions Education: July 2023

Wednesday 12 July 2023

Artificial Intelligence in Higher Education Institutions

The use and development of technologies have transformed the world we live in. We have transitioned from a world of Web 1.0, where everything happened through a computer with dial-up internet connection, static portals, and the need for experts to establish an online presence. Later, users were empowered, and content creation was no longer exclusive to programmers. Nowadays, we all have the possibility to create content in various formats such as videos, audios, or texts, to name a few. We have the convenience of conducting financial operations remotely. The world has been evolving, and new ways of offering goods and services have emerged. The ultimate dream has always been for technology to take a step further and assist us with activities exclusive to humans.

While there are various Artificial Intelligence technologies, this post specifically mentions ChatGPT. Ever since ChatGPT, announced as the first Artificial Intelligence portal, was released for public use (with access restrictions in some countries), different industries have been embracing its use to gain competitive advantages and excel in their productive sectors. Educators at various educational levels are wondering about its application in education. OpenAI developed ChatGPT, which was launched in June 2020.

Padilla (2019) attributes the following characteristics to an intelligent agent:

1. Possesses mental attitudes such as beliefs and intentions.

2. Has the ability to acquire knowledge.

3. Can solve problems, even by breaking down complex problems into simpler ones.

4. Capable of performing complex operations.

5. Understands and can make sense of ideas.

6. Plans, predicts consequences, and evaluates alternatives.

7. Knows the limits of its own abilities and knowledge.

8. Can distinguish despite the similarity of situations.

9. Can be original, creating new concepts or ideas.

10. Can generalize.

11. Can perceive and model the external world.

12. Can understand and use language and its symbols.

But what is Artificial Intelligence?

The simplest answer to convey, as stated by Padilla (2019), is "the automation of activities that we associate with human thought processes, such as decision-making, problem-solving, machines exhibiting the same capabilities as humans."

ChatGPT is the first Chat Generative Pre-trained Transformer and was designed to generate responses and maintain conversations in the context of a chat or user interaction. Some industries are using AI for automation, information analysis, and pattern-based simulations.

Applications in the educational field, according to Vera (2023), include:

- Intelligent conversational agents

- Online platforms with personalized learning

- Processing large amounts of data

The integration of Artificial Intelligence represents a challenge but should also be viewed as a supportive tool to enhance learning experiences. From content delivery to activity development and assessment processes, teachers have a new ally. The next installment will delve further into this topic.

Juan Pablo Nigenda Alvarez

Traducido al inglés por ChatGPT

Bibliografía

Padilla, R. D. M. (2019). La llegada de la inteligencia artificial a la educación. Revista de Investigación en Tecnologías de la Información: RITI, 7(14), 260-270.

Vera, F. (2023). Integración de la Inteligencia Artificial en la Educación superior: Desafíos y oportunidades. Transformar, 4(1), 17-34.

Sunday 9 July 2023

Augmenting Human Intelligence with AI

Augmenting Human Intelligence with AI

Poh-Sun Goh

10 July 2023, 0807am, Singapore Time

AI is another way of augmenting human intelligence, human capability. We are all familiar with well tried and tested methods to gain new perspectives, deepen insights, and broaden our outlook, by reading around a topic, speaking to experts, attending conferences and courses, and getting formal training. AI is another tool to add to our toolkit.

Further reading

Markauskaite, Lina & Marrone, Rebecca & Poquet, Oleksandra & Knight, Simon & Martinez-Maldonado, Roberto & Howard, Sarah & Tondeur, Jo & Laat, Maarten & Buckingham Shum, Simon & Gasevic, Dragan & Siemens, George. (2022). Rethinking the entwinement between artificial intelligence and human learning: What capabilities do learners need for a world with AI?. Computers and Education: Artificial Intelligence. 3. 100056. 10.1016/j.caeai.2022.100056.

https://www.sciencedirect.com/science/article/pii/S2666920X2200011X

Dimitriadou, E., Lanitis, A. A critical evaluation, challenges, and future perspectives of using artificial intelligence and emerging technologies in smart classrooms. Smart Learn. Environ. 10, 12 (2023). https://doi.org/10.1186/s40561-023-00231-3

https://www.ama-assn.org/practice-management/digital/augmented-intelligence-medicine

https://hbr.org/2021/03/ai-should-augment-human-intelligence-not-replace-it

https://www.capgemini.com/insights/expert-perspectives/ai-4-education/

Friday 7 July 2023

Questions that technology solutions are doing regarding AI

Mildred Lopez

This weekend I had the opportunity to attend Times Higher Education (THE) Latin America Universities Summit in partnership with Tecnologico de Monterrey. I participated in a networking session hosted by Turnitin https://www.turnitin.com/, and they posed three questions that I want to share with you:

Do you believe that the students using AI writing tools represent a challenge or an opportunity?

In my discussion table, we shared these arguments:

• The challenge for teachers as they need to rethink the assignments in something more than writing an essay, as it can be easily generated with Chatgpt.

• The challenge for teachers is that to change their ways, teachers need training, and they need time and money.

• The challenge for institutions is that many priorities are battling for the same resources, such as teacher training and buying new platforms.

• It is an opportunity for teachers to use their time wisely in activities that add value for learners, for example, in mentoring rather than grading.

• It is an opportunity for programs to migrate from memorization to the application of knowledge finally.

How do you integrate AI with teaching?

• In developing educational resources that are adaptable to the interests and needs of learners.

• In providing accurate and timely feedback for individual assignments, we can provide lots of input for classwork, and teachers can focus on the most complex projects where students create or propose ideas.

• As a teaching assistant, to curate content.

What is the impact that writing with AI has on assessment?

It is making evident the problems we already hold regarding ethics and integrity. Although these are not new, it is harder to detect plagiarism or understand the percentage of the content generated by humans or AI.

What do you think about these questions?

Disclosure:

I do not hold any commercial relationship with the platform.

Wednesday 5 July 2023

Reflecting on current Learning and Work processes, and current limitations of AI

Reflecting on current Learning and Work processes, and current limitations of AI

Poh-Sun Goh

6 July 2023, 0337am, Singapore Time

Currently when making notes and preliminary reading and research around a topic I start by outlining what I currently know, and then sequentially and iteratively do key word online searches, reviewing topic and review articles, and further background and source material review. Then refine this material to the task at hand, and with a target audience. Current limitations of AI are the datasets it has been trained on, and the absence of dynamic, ongoing access and feedback (from human trainers) of the latest, topical and trending, as well as integrative work. My feel is that AI has a role as part of our toolkit, as one of several sources of information and assistance we might use, but with a clearly define role, and awareness of its limitations when, and if, we use AI, to assist, and augment our work.

Saturday 1 July 2023

Beyond Chatgpt: AI resources for everyone

Mildred López

Recently, AI has been the buzzword everywhere. Even while grocery shopping, I overheard a discussion about AI where two clients in the fruit section were talking about this “Chagpt thing” supposedly taking over the world as it has a life of its own. AI is advancing rapidly; society needs to be involved in its development and the definition of management policies; however, AI is not only Chatgpt. There are many proposals that we can leverage to enhance our daily life and our medical education practice. I believe it is essential for us to explore the different uses to truly grasp what this is all about.

1. PaLM - Google

The company uses PaLM 2 as the large generation model to power features such as Bard and the PaLM API. The proposal is to build AI with a responsible approach.

https://ai.google/

You can learn more about it here:

https://www.youtube.com/watch?v=880TBXMuzmk

2. Watson - IBM

It is a suite of technologies developed by IBM. It uses as well natural language format to interact with humans.

https://www.ibm.com/mx-es/watson

It became famous in 2011 when it defeated humans on the TV’s Jeopardy! Family show. You can learn more about it here:

https://www.youtube.com/watch?v=P18EdAKuC1U

3. Chatgpt - OpenAI

It is a chatbot built on OpenAI’s platform to “get answers, find creative inspiration and learn something new.”

https://openai.com/chatgpt

Meet OpenAI CEO in this interview:

https://www.youtube.com/watch?v=540vzMlf-54

4. Azure AI – Microsoft

AI services for developers with access to vision, language, and speech analysis through learning models.

https://azure.microsoft.com/

These developments support many tools that can be incorporated for writing, generating images, etc. I’ll share a list with a few that I have collected, but feel free to share more in the comment section.

Writing

1. Grammarly: it is a writing assistance platform that enhances the grammar, spelling, tone, and pace of English writing. It can be used in documents, emails, and most websites.

https://app.grammarly.com/

2. Hypotenuse AI: content writer specialized in selling products, including the advantage of plagiarism check.

https://es.hypotenuse.ai/

3. Sider AI: it has many tools within: summarizes articles, finds similar pages, explains any complex concepts, and compares answers between AI and humans

https://sider.ai/

4. Wordtune: “Say exactly what you mean through clear, compelling, and authentic writing.”

https://www.wordtune.com

Image and video generation

1. Synthesia: turns text into speech through more than 140 avatars and creates a video using AI.

https://www.synthesia.io/

2. Gamma AI: creates documents, presentations, and webpages through prompts. The content that was generated can be edited through the platform or offline.

https://gamma.app/

Chatgpt plugins. https://openai.com/blog/chatgpt-plugins

1. AskYourPDF: share a document in a pdf format, and then Chatgpt can generate questions about that document. It can be powerful to help students review large quantities of data.

https://askyourpdf.com/

2. Prompt perfect: a plugin that aids in crafting the best instructions for Chatgpt (or other. Language processors).

https://promptperfect.jina.ai/

What other developments, tools, and ideas have you tried?

Learning and Training in the age of AI - going Micro, Macro and Meta

Learning and Training in the age of AI - going Micro, Macro and Meta

Poh-Sun Goh

Sunday, 2 July 2023, 0242am, Singapore Time

Having access to the world's knowledge (i.e. with access to the Internet), or an intelligent algorithm (i.e. AI or Artificial Intelligence), does not mean we have become more learned, become trained, more capable or skilled. Being able to look up an answer to how to do something online, or have an AI tool 'tell you' is only the starting point.

Going Micro and Macro requires human assessment of both the answer(s), as well as its relevance to the question or task at hand, and its broader context. Going Meta includes human evaluation of underlying logic, and overarching principles and theory that might be relevant. Human input is particularly useful to do beyond surface answers to deeper considerations - from What and How, to Why and If.

There is value in figuring out the answer or how to do something compared to being given the answer or shown how. Discovery and exploratory learning activities opens the door to creative and new approaches to doing things. Not to mention deeper understanding and insight into why certain ideas and solutions were chosen, and implemented in a specific context.

Imagine the task of finding and fitting all the pieces to complete a picture puzzle. Whilst software and AI can accelerate the process of completing this task, there is value in examining and assessing each piece of the puzzle, fitting these together by trial and error, and progressively assembling the whole construct. When we are given an answer, and shown a solution, we lose the opportunity to explore and examine alternative and other avenues and ideas, and taking full advantage of questioning to learn. Fast and quick does not necessarily mean better or best - answer, idea, solution or path.