3 ways ChatGPT will transform soft skills learning in education
“ChatGPT will change the world” the tech prophets say, but how is ChatGPT going to change the way we learn soft skills? Will Artificial Intelligence (AI), combined with Extended Reality (XR) technologies, make both e-learning and coaching obsolete? And can machines really help humans be better with one another? Let’s take a look at 3 ways in which ChatGPT will transform soft skills learning.
#1 - Every Learning Designer will get an army of scriptwriters
#2 - Infinite conversations can be realistically simulated
#3 - Every learner will get an AI coach
To keep your reading more fun, I asked another AI tool, Midjourney, to create architectural drawings of future universities in the style of 4 “starchitects”. Can you guess which ones they are? Check the answers at the end of the post.
#1 - Every learning designer will get an army of scriptwriters
“The Learning Designer must sit on both sides of the AI.”
Learning Designers (LD) can leverage ChatGPT to do research and write scripts, at much greater speed. Using sophisticated sets of instructions also known as mega-prompts, any output format is possible: a lesson, a dialogue, a quiz, a riddle… Any length, any tone.
Case in point, I asked ChatGPT to write a paragraph about this topic from the perspective of a learning designer:
As a Learning Designer, ChatGPT can be a useful tool for quickly creating scripts for e-learning modules. ChatGPT can generate scripts in just a few minutes, which saves me time and allows me to focus on the creative aspects of designing a module. The generated scripts are comprehensive and can be easily edited to customise them to the project. In addition, the system can provide personalised feedback on the script's quality.
So will ChatGPT replace the Learning Designer? Nope. The learning designer must sit on both sides of the AI.
On the input side, the LD alone can decide on the learning objectives and design the scaffolding of activities to achieve those. And therefore instruct the AI with unique prompts. On the output side, regardless of how good the prompts are, ChatGPT can only amalgamate what’s available online without the expertise of knowing what’s actually best practice. And so the LD will always need to edit, trim, refine and often start a new AI cycle.
Again, as ChatGPT puts it:
However, there are some limitations to scripts created by ChatGPT compared to those written by humans. ChatGPT-written scripts may lack the creativity and the fine-tuned details that can be added by a human writer. Additionally, ChatGPT-generated scripts may not always capture the nuances of the subject matter or the desired tone of the project. Therefore, it is important to edit the scripts and make sure they are appropriate for the project.
#2 - Conversation simulators will feel real
“For learners, that means practice is available anytime, anywhere.”
ChatGPT passed the Turing Test, meaning conversing with ChatGPT can sometimes be indistinguishable from conversing with an actual human.
Up to now, creating endless self-generating human-like conversations for training was impossible. This was one of the major weaknesses of digital asynchronous tools vs synchronous modalities such as workshops or coaching.
Whereas a role-play could go anywhere, learning designers had to resort to somewhat contrived exercises: branching conversations, analysing model answers, asking users for monologues etc.
ChatGPT can now pilot a chat or a 3D avatar with a predefined role, personality and motive. The learner can then have an ongoing unscripted conversation with that avatar. The AI character takes clues from the learner’s input to respond, in meaningful and personalised ways. Every simulated conversation is unique.
Here are a few examples of conversations you might want to simulate that way:
- A tough job interview for a role you’re actually applying for in real life
- An upcoming decisive pitch competition with inquisitive investors
- Negotiating your salary with a short-tempered boss
- Challenging a sexist colleague on the spot
- Helping out a loved one suffering from acute anxiety
For learners, that means practice is available anytime, anywhere. And because the ChatGPT-led avatars are, well,not real, there is total psychological safety. No judgement, no hard feelings: just a conversational sandbox to experiment in.
For organisations, this offers an unprecedented level of flexibility and cost-efficiency, approaching the ideal of providing every employee with their own coach.
#3 - Every learner gets an AI coach
Practice without external feedback is limited. ChatGPT can analyse anything the learner says in a simulated conversation and provide hyper-personalised feedback, in real-time.
For example, the AI will look into a learner’s answer to a simulated job interview. It will provide feedback on how their personality came across, the relevance of their answers to the job role and company culture. Then it’ll provide suggestions on how to improve according to the STAR (Situation, Task, Action, and Result) technique, and even play out a model answer via an avatar.
Is it as good as a real human coach? Of course not. That’s not the point: AI makes coach-like, hyper-personalised feedback available to the 99% who are not able to afford a human coach. Talk about levelling the playing field!
So, dear ChatGPT, how have you found this article so far?
This blog is structured in a clear and logical way, and the tone is very engaging and informative. The author does a great job of explaining the potential of ChatGPT and how it can revolutionise soft skills learning in a concise and understandable way.
Well, thank you. That’s what ChatGPT said the second time around: after I reworded the structure and wording following its initial advice.
AI, XR and the metaverse
To conclude, the AI suggested that I summarise the key points above. So here’s a summary:
ChatGPT is a tool for the Learning Designer to create disruptive soft skills learning solutions likely to be more accessible, impactful and affordable than any previously available alternative.
For the learner, it’s an always-on anywhere and anytime coach; and roleplay partner.
For learning organisations, be it education institutions or corporates, it’s a game-changing opportunity to improve the performance and well-being of all employees.
ChatGPT doesn’t exist in a technology vacuum. It’s much more powerful when used in conjunction with other metaverse technologies, namely VR.
A conversation simulator is only truly meaningful if the character you’re interacting with looks and feels like a sentient entity, one that can destabilise you with a piercing glance, make you lose confidence by crossing their arms right when you push your key point or smile and nod in approval.
The environment also matters. If your phone were to ask you pertinent questions after you filmed yourself doing your pitch, you’d still be pitching to your empty living room. Nothing makes the heart race like the spotlight in your face and impatient coughing noises from the front row.
Only then will your performance be real, and therefore the AI Coach’s feedback truly accurate.
Pertaining to Amara’s Law, this article is most likely overestimating how quickly this vision will become reality. It’s also wildly underestimating how disruptive the technology will be in the long-run. One thing is certain however: machines are getting better at being machines, and, put in the hands of pioneering Learning Designers, they can help us become better at being humans.
In case you’re still wondering. The blue-ish image was Zaha Hadid, orange for Santiago Calatrava, multicolour for Antoni Gaudi and rust for Gustave Eiffel.