AI tools ChatGPT and DALL·E 2 can write content and produce art, respectively, and have potential uses in education. However, concerns about the role of humans in the creative process must be considered. It is important to teach responsible use of AI.
“Game-changer” is a much used and abused word, especially in tech, but these last few months have seen staggering advances in artificial intelligence. All over social media there is talk and excitement about ChatGPT and DALL·E 2. Let’s take a short look at what they can do for users, and what all that means for educators and students. If, for example, AI can write and produce art, where does that leave us? How will we use AI to enhance the human experience and improve the world?
Language-rich artificial intelligence – ChatGPT
ChatGPT has got everyone talking. It’s early days, but so far, its achievements are impressive. Among other things, ChatGPT can code and write content. A student studying GCSE History could type “Write a 300-word essay on why the main reason why the League of Nations failed was because of how it was organised” and have their homework done in seconds!
Students and techies are abuzz on social media, with some describing it as “scarily good”. It’s worth noting, however, that OpenAI is keen to manage expectations. "ChatGPT sometimes writes plausible-sounding but incorrect or nonsensical answers,” it warns us.
Take a look at its answers to the history question above and others here, and judge for yourselves.
At the end of this piece, you may wish to type in “Write me a 1,000-word blog on ‘Is artificial intelligence the future of education?’" and see how we compare. To hazard a guess, we reckon you’ll find it useful, informative, and a decent starting point, but short on the specifics that illuminate the issues. Let’s not forget, ChatGPT is, in essence, researching and reassembling what has already been written on any given topic. By a human.
But credit where credit’s due. ChatGPT is clearly doing more than merely searching the internet. Like “game-changer”, “artificial intelligence” is an overused term, and some use it rather loosely. It’s a convenient catchall phrase, with the added bonus of the accompanying hype. But if AI is “the theory and development of computer systems able to perform tasks normally requiring human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages” (Oxford Reference definition), then this is very much what language-rich ChatGPT is doing. It is clearly showing the ability to follow instructions, interpret, reason, research, problem-solve, learn, and then create. And these capabilities open up a brave new world of possibilities.
Take content and course creation. University professors could instruct ChatGPT to write essays for students to critique. Some say the writing itself can sound rather workaday or stilted, but teachers could work with students on improving AI prose, injecting the human into the artificial.
Or, artificial intelligence could in the not-so-distant future take on the task of course design. With its ability to collate and rewrite, ChatGPT is a great automated brainstorming tool. It could quickly and painlessly generate ideas for professors to trim and tailor for their students.
Of course, the technology raises concerns about plagiarism, but we need to teach students to use AI, with all its capabilities and limitations, responsibly. This will remove the temptation of cheating and shift the focus to working with technology to solve real-life problems.
It’s also worth noting that OpenAI is developing a watermark tool to prevent plagiarism. Using a cryptographic pseudorandom function, the “watermark” will add an unnoticeable signal to the text.
AI art – DALL·E 2
Another boon to students and course creators alike, DALL·E 2 can create imagery to supplement the written word, producing original, realistic images and art from simple instructions. Simply type in concepts, attributes, and styles. It will do the rest. The examples it gives on its homepage are:
An astronaut | Teddy bears | A bowl of soup
riding a horse | lounging in a tropical resort in space | playing basketball with cats in space
in a photorealistic style | in the style of Andy Warhol | as a pencil drawing
University professors could use AI art generation for making diagrams or learning aids, further reducing time in course creation. Artists, photographers, and graphic designers are finding it invaluable for image manipulation, modification, and troubleshooting. Like ChatGPT, it is language rich, and so, with simple prompts, it can isolate a certain part of an image and modify it as instructed. Some say it has a stock photography feel, but this tech is, arguably, in its infancy. We are only beginning to fully appreciate its potential.
AI or digital assessment – Graide
Graide’s AI allows for the semi-automation of marking, lightening the burden which so hampers professors. Here AI doesn’t so much replace as replicate. Graide is guided by the teacher and repeats human decision-making, using a technology called “replay grading” ensuring that similar submissions are not marked twice, and grouping similar answers so they can be given a consistent grade at the same time. It’s also great at compiling and collating the results. Gone are the days of manually entering grades. Digital assessment means delegating the drudgery. Like autocorrect or suggestions when we write emails, this is something we now all take for granted.
Read more about how Graide works here.
AI is the great liberator. Better still, it is also a great investigator when it joins the dots and spots the gaps in students’ understanding. Teacher dashboards and knowledge graphs allow educators to identify and address, among other things, common misconceptions, in a timely fashion. Early intervention is crucial to STEM retention at university.
AI also enables adaptive or personalised learning: tech that delivers individualised content and tasks to suit each student’s strengths and weaknesses, providing just the right level of stretch, at scale. Often, educators “teach to the middle”: they don’t want to confuse the less able or bore the brightest. But adaptive tech circumvents this.
So, these are exciting times for AI, with developments driving change at a rate of knots. Is artificial intelligence the future of education? It will certainly play a key role, but that will be alongside humans, assisting and enhancing rather than replacing them. Perhaps the bigger questions are “What tasks suit tech, and which best suit teachers?”, and focusing our endeavours accordingly. Education is more than just delivering content and assessing. Teachers will remain the stars. They’re the ones who connect on a human level. The tech is there to free up time for that.