Created: 18 January 2023
“a futuristic image of an open artificial intelligence being” by Michael × DALL·E Human & AI
To try to understand what is going on with all this Open Artificial Intelligence (AI) stuff, you first need to understand what it means to be open, and what open educational resources can be. This should lead you to an idea and vision of how Open AI tools are shaking things up around the world in many sectors.
This article should give you an idea about Open AI, what products are available, what they can be used for, how they could positively or negatively impact the educational sector, and what educators can do with Open AI tools.
What is Open?
“a psychedelic fractal coming out from inside an open treasure chest on a beach” by Michael × DALL·E Human & AI
Open is a key term that refers to sources, material, etc that is free and open to use. Examples you may have heard of could be: Unsplash, Brave browser, Audacity, SecondLife, and that silly eCard generator your grandma still uses to send you an ebirthday card that has a .GIF of a rabbit popping out from behind a fence with some flowers.
The term ‘open content’ describes any copyrightable work (traditionally excluding software, which is described by other terms like ‘open source’) that is licensed in a manner that provides users with free and perpetual permission to engage in the 5R activities:
• Retain — the right to make, own, and control copies of the content (e.g., download, duplicate, store, and manage).
• Reuse — the right to use the content in a wide range of ways (e.g., in a class, in a study group, on a website, in a video).
• Revise — the right to adapt, adjust, modify, or alter the content itself (e.g., translate the content into another language).
• Remix — the right to combine the original or revised content with other open content to create something new (e.g., incorporate the content into a mashup).
• Redistribute — the right to share copies of the original content, your revisions, or your remixes with others (e.g., give a copy of the content to a friend).
(Wiley, 2015)
“an anime rabbit popping out from behind a fence with some flowers” by Michael X Maker.ai Human X AI
What are Open Educational Resources?
OERs are ‘educational materials made freely and legally available on the Internet for anyone to reuse, revise, remix and redistribute’ (The William and Flora Hewlett Foundation, 2013, p. 4) and they have been part of the educational landscape since at least 2001. The flexibility and power of OER is made possible by their open licence, granting the user various permissions in respect of the way they are used. (FutureLearn, 2021)
So now imagine AI being Open, let’s say, like a Google?
“a futuristic image of an open artificial intelligence being” by Michael × DALL·E Human & AI
What is OpenAI and ChatGPT?
OpenAI Playground and ChatGPT are two artificial intelligence (AI) tools developed by OpenAI, a research laboratory for artificial intelligence and machine learning.
The use of Artificial Intelligence (AI) technology has become increasingly popular in the field of education, especially with the emergence of OpenAI Playground and ChatGPT. OpenAI Playground is a web-based platform that allows users to explore machine learning algorithms by writing code and running experiments. ChatGPT is a natural language processing model that enables users to interact with a computer as if they were chatting with a human. This article will discuss the potential benefits and challenges that AI tools may bring to education around the world, and what educators should do about it. But what AI tools are there?
What Open AI tools are there?
“a cat opening a bedroom door that has magical light bursting through” by Michael × DALL·E Human & AI
Playground is a web-based tool that allows users to explore and edit machine learning models in an interactive and intuitive way. A user can understand the basics of machine learning. It provides a graphical interface that allows users to modify different parameters in order to create their own custom models, and to visualize the resulting behaviour and performance of their models. The application also allows users to save, share, and embed their models in other applications.
DALL-E 2 (text to image) is a tool that uses AI to generate images based on text input. Users can input a description of an image they want to create, and DALL-E will generate a corresponding image. DALL-E can be useful for students who are trying to visualize a concept or idea, as it can help them generate a visual representation of their thoughts.
Clip connects text and images. It efficiently learns visual concepts from natural language supervision. CLIP can be applied to any visual classification benchmark by simply providing the names of the visual categories to be recognized, similar to the “zero-shot” capabilities of GPT-2 and GPT-3
Jukebox is a deep learning-based music synthesis system. Jukebox allows users to generate music from scratch using its neural networks. The system has been trained on a large collection of music, including genres such as classical, jazz, and rock. Provided with a genre, artist and lyrics as the input, Jukebox outputs new music samples produced from scratch!
Whisper is a trained and open-sourced neural net that approaches human level robustness and accuracy on English speech recognition. Whisper is an automatic speech recognition (ASR) system trained on 680,000 hours of multilingual and multitask supervised data collected from the web. It uses a large and diverse dataset which leads to improved robustness to accents, background noise and technical language. Whisper can enable transcription in multiple languages, as well as translation from those languages into English.
Gym is an interface simple, pythonic, and capable of representing general RL problems. It is all that ‘behind the curtains’ of websites, games, etc that many of us just have no clue about. It allows you to make or custom your own environment. Think like an Atari game for those of you who could be seen as a digital immigrant or digital native (Prensky, 2001).
(OpenAI, 2023; 2023a; 2023b; 2023c; 2023d; 2023e; 2023f; 2023g)
But wait, that is not all that is out there!
“a beach oasis with an open and delightful view” by Michael × DALL·E Human & AI
What other AI tools are there?
ChatBCG (AI for presentations) is a tool that uses AI to help users create professional-quality presentations. It can suggest design elements and layout options, as well as generate text for slides based on user input. ChatBCG (2023) can be helpful for students who are working on a presentation for a school project, as it can save time and effort in creating visual aids.
MakerAI (AI for documents and images) is a tool that uses AI to assist users in creating documents and art. It can suggest design elements and layout options, as well as generate text and images (i.e.: Linkedin articles, social media posts, blog posts, et al.) based on user input. MakerAI (2023) can be helpful for students who are working on a document or art project, as it can save time and effort in creating the final product.
VALL-E (text to audio) is Microsoft Research's tool that uses AI to generate audio files based on text input. Users can input text, and VALL-E will generate an audio file of a synthetic voice reading the text. VALL-E can be useful for students who are trying to learn a new language, as it can help them practice pronunciation and listen to the spoken language (Wang et al., 2023; Fowell, 2023).
Just to name a few…
What does all this mean?
“artificial intelligence robotic green bugs scurrying around on top of a motherboard with storms of electricity raining down” by Michael × DALL·E Human & AI
What this all means is that if you want to know the best American cookie dough brownie recipe, it will give it to you. If a child wants to describe the monster from their nightmare into an accurate computerised 2D/3D illustration, easy as. If you want to write an academic research proposal about the decline of Fish and Chip shops in the UK with the benefits of outdoor learning, it can do it. If you want a summary of a selection of books that focuses only on the topic of gender, that is no problem at all. Enter a prompt and boom! An essay, poem, coding, a World Cup grand final analysis, or any other kind of text can be generated automatically with complete originality. The options and possibilities are endless. Use your imagination as the input.
These tools have the potential to revolutionise how we teach about machine learning and AI. By allowing students to interact with an AI in a conversational way, they can gain a better understanding of the technology and its potential applications. These AI tools can be used to help students understand the fundamentals of machine learning and artificial intelligence, develop the necessary AI input-output skills in order to refine their work as accurately as the user desires, and to gain an appreciation of the potential and power of these technologies.
What are the Benefits?
“a 3d fantasy landscape with a wise wizard facing the abyss” by Michael × DALL·E Human & AI
Using AI tools mentioned in teaching and learning can have several benefits. For example, these tools can help students save time and effort in creating presentations, documents, art, and other projects. They can also help students visualize concepts and ideas, and practice pronunciation and listening skills. When considering the endless opportunities with Open Art, or Jukebox, this would require very explicit and particular language forms to be understood and known prior to producing any kind of work that would resemble the quality you expect from AI.
This kind of literacy synthesis can be a great help for students and educators, as it allows them to explore and experiment with subject knowledge and the metacognition of the AI itself, without having to spend a lot of time and resources. Additionally, the platform allows for collaboration between users, which can help foster creativity and innovation. OpenAI language models are only limited to the individual’s imagination.
Open AI technology presents many innovative options for educators to use in their classrooms. They offer the potential to help students create content in an interactive and engaging way, and to access information faster than ever before. However, these technologies also present some challenges that educators must consider when integrating them into their lessons.
What are the Challenges?
There are challenges to using AI tools in teaching and learning. One challenge is that these tools may not always produce accurate or high-quality output, as they are still being developed and refined. Another challenge is that students may become overly reliant on these tools, which could negatively impact their ability to think creatively and independently. It is important that students are taught to use these tools responsibly and as a supplement to their own learning and creativity, rather than relying on them solely.
One of the main challenges associated with these technologies is the need to ensure that students understand the ethical implications of working with AI applications. AI-based applications have the potential to be both powerful tools for learning and sources of potential bias and a reshaping of academic integrity.
For instance, some misunderstand Chat-GPT and see that it invents academic references and in most cases its so-called ‘database’ and seems to end at 2021. It is important to know that Chat-GPT is not a ‘knowledge’ base, it is a ‘language’ model (Sharples, 2023).
Therefore, it is important for educators to help students develop an understanding of the ethical implications of using these technologies. Is essay writing in various forms in higher education really dead (McMurtie, 2022)? Definitely not. It is more about certain pedagogical and assessment practices are now dead.
Another challenge associated with Open AI technology is the potential for distraction. Educators must ensure that students are using the technology for educational purposes and not for recreational activities. These tools also bring with them new challenges regarding originality and learning. For example, students may be more likely to take shortcuts and rely on pre-made responses instead of forming their own original ideas. Additionally, the conversational nature of the technology can make it difficult for teachers to effectively evaluate student progress and understanding (Giles, 2020).
AI tools can be useful resources for students and educators, but they should be used with caution and in conjunction with other learning methods. However, various pedagogical forms will definitely struggle or be made redundant with the inclusion of AI tools in education. What is most important is how assessment and pedagogy/andragogy evolves with the technology surrounding it.
“An assortment of little green artificial intelligence bugs scurrying around on top of a motherboard with an Earth-like landscape to it with storms of electricity raining down” by Michael × DALL·E Human & AI
What pedagogy and assessment can deal with AI tools already?
I’ve always preached through learning and experience that the humanist philosophy of an educational framework of the International Baccalaureate (IB) programme(s) have always been one step ahead of the rest of the world, when it comes to progressive and relevant learning and teaching in the 21st-century. It goes without saying that such a framework replicated in other ways and means can definitely manage and incorporate AI tools into their learning environment without having any of the current dilemmas, concerns and issues other educational institutions are currently facing. Some of the proposals like banning Chat-GPT or going back to handwritten essays is a clear indication either of: (a). political affiliation to political discourse in education; (b). unpreparedness by educational institutions regarding educational technology, pedagogy/andragogy, and assessment practices relevant to transdisciplinary themes, reflection and agency: ‘how the world works’, ‘where we are in place and time’, ‘how we organise ourselves’, ‘how we express ourselves’, ‘who we are’ and ‘sharing the planet’ (International Baccalaureate Organisation, 2023).
An approach to learning and teaching that utilises intricate parts of an IB framework would require educators to move away from formative, standardisation, direct instruction and industrial-aged practices, such as those seen in some national curriculum around the world.
What are the Possibilities for Educators?
“a teacher with technological skills of the future” by Michael × DALL·E Human & AI
There are possible ways to navigate through the challenges for educators with these AI tools available. Here is a set that I have built on from Educator Anna Mills on Twitter (2022) on 40 ways to prevent learning loss, deter AI generator abuse, and discover the benefits for educators:
Analysis of images and videos (AI is still developing in this area)
Analysis that draws on class discussion (Note: AI could easily transcribe a recorded discussion)
Analysis of a long text
Writing about recent events
Writing research proposals
Updating academic integrity policies
Education focused on thinking
Learning about the limitations and faults of the generators produced work
Assignments that are meaningful and intrinsically motivating
Utilising an un-grading approach (developmental?)
Awareness of typical AI-features of generated texts (i.e.: bland style of writing, invented sources, inconsistent reasoning, et al.)
Assignments requiring a process of production
Interpretation and perspective critique
Include video and audio submissions
Metacognitive reflection
Handwritten assignments
In-class writing
Pedagogical and assessment changes to teaching, learning and module/curriculum design
Surveillance software
Software that tests students’ knowledge on their own written work (AI generated questions?)
Language development in order to refine AI generated products
Ensuring that teachers have the skills required for 21st-century
Synthesis of language appropriate for various tasks
Multilingualism development; the ability to communicate and learn languages other than mother tongue
Heightening the human ability to critique
Planning units of work as per required inputs
Assessing text according to inputs
Assistance in editing and referencing tasks, skills and processes
Strengthening the knowledge and ability of referencing and accuracy of sources
Creativity for transdisciplinary purposes
Multiliteracy development for transdisciplinary learning and teaching
Developing new ICT and 21st-century skills by using AI tools
Understanding the challenges, benefits and limitations of AI
Personalised tutoring
Automated assessment grading/marking
Automated research assistance
Automated test taking
Engaging young learners in technology that will be part of their daily life in the future
Creating original music according to the input and intentions of the creator
Require hyperlinked reference lists and in-text citations in written work from students
Educators must embrace AI tools immediately!
“a teacher with technological skills of the future” by Michael × DALL·E Human & AI
AI in education is nothing new. This has been happening for a few years now, but only recently has become more mainstream. This is the scariest thing about it. Educators and institutions need to act now or run the risk of being left behind.
Open AI tools are pioneering advancements in the field of natural language processing and artificial intelligence, and their implications for the future of education are far-reaching. They have the potential to revolutionise the way students learn and interact with information, providing a more efficient and engaging way to access and understand educational material. With their potential to provide personalised and interactive learning experiences, Engaging and working with all kinds of AI tools is the first step towards a more innovative, inclusive and effective educational model. It would be good to suggest a study where AI generated university assignments are created and submitted as a virtual student to further understand student achievement and rigour of higher education assessment and design. However, it is word on the street that Higher Education institutions in Australia are considering going back to the pen and paper exams, moving from ignore to ban (Sharples, 2023)
With proper implementation and oversight, AI tools can be an invaluable resource for education around the world. The scary thing is, these are not the only AI tools available today. You might have noticed that all the images in this article were generated by the command described below each image, some were created more than once. Please note that this has been a collaboration with AI and it only took me about an hour. Do you feel threatened? What next?
“a psychedelic fractal coming out from inside an open treasure chest on a beach” by Michael × DALL·E Human & AI
References
ChatBCG (2023). ChatBCG. ChatBA [online] Available at: https://www.chatba.com/
Fowell, A. (2023) Microsoft Reveals Latest Text-To-Speech AI Research, VALL-E, TechTimes [online] Available at: https://www.techtimes.com/articles/286057/20230110/microsoft-reveals-latest-text-speech-ai-research-vall-e.htm
FutureLearn (2021) H880: Technology and Learning: Opening Up Education: Week 2, The Open University, Milton Keynes, UK
Giles, A. (2020). Open AI Playground and ChatGPT: New Challenges for Educators. Educational Technology Trends. https://edtechreview.in/trends-insights/trends/4358-open-ai-playground-and-chatgpt-new-challenges-for-educators
International Baccalaureate Organisation (2023) About the IB, [online] Available at: https://www.ibo.org/about-the-ib/
MakerAI (2023). MakerAI. Content Fly Inc. [online] Available at: https://maker.ai/
McMurtie, B. (2022) ‘AI and the Future of Undergraduate Writing’, The Chronicle For Higher Education, [online] Available at: https://www.chronicle.com/article/ai-and-the-future-of-undergraduate-writing
Mills, A. (2022) ‘How do we prevent learning loss due to AI generators?’ 14 December 2022 Mike Sharples @sharplm, Post on Twitter [online] Available at: https://twitter.com/sharplm/status/1602953413704429570
OpenAI. (2023). Playground. [online] Available at: https://beta.openai.com/playground
OpenAI. (2023a). ChatGPT. [online] Available at: https://openai.com/blog/chatgpt/
OpenAI (2023c). OpenAI Jukebox [online] Available at: https://openai.com/blog/jukebox/
OpenAI (2023d). OpenAI Clip [online] Available at: https://openai.com/blog/clip/
OpenAI (2023e). OpenAI Gym [online] Available at: https://gym.openai.com/
Open AI (2023f) Whisper Available at: https://openai.com/blog/whisper/
OpenAI (2023g). DALL-E 2. Available at: https://openai.com/dall-e-2/
Prensky, M. (2001) Digital natives, Digital immigrants, MCG University Press, Vol. 9, No. 5, [online] Available at: https://www.marcprensky.com/writing/Prensky%20-%20Digital%20Natives,%20Digital%20Immigrants%20-%20Part1.pdf
Sharples, M. (2023) 18 January 2023 @sharplm Post via Twitter [online] Available at: https://twitter.com/sharplm/status/1615607290320519169?cxt=HHwWgoCwwamR5essAAAA
The William and Flora Hewlett Foundation (2013) White Paper: Open Educational Resources: Breaking the Lockbox on Education. Available at: http://www.hewlett.org/sites/default/files/OER%20White%20Paper%20Nov%2022%202013%20Final.pdf Accessed 20 March 2019
Wang, C., Chen, S., Wu, Y., Zhang, Z., Zhou, L., Liu, S., Chen, Z., Liu, Y., Wang, H., Li, J., He, L., Zhao, S, and Wei, F. (2023) Neural Codec Language Models are Zero-Shot Text to Speech Synthesizers, Cornell University [online] Available at: https://arxiv.org/abs/2301.02111
Wiley, D. (2015) ‘Defining the “Open” in Open Content and Open Educational Resources’ [Online]. Available at http://opencontent.org/definition/ Accessed 4 March 2019
Comments