In Lesson 20 we saw that the problem was GIGO. Now we look at the solution. In the next three lessons we will see three practise items on the GBB. The first one will be: Free Fridays on Public Transport. These will be full 10-point GBBs and therefore assisted by AI. Human binary thinkers usually can do around 3-5 points.
Lesson 23 Daily Feedback Question:
Do a GBB about how you currently feel about AI.
• Post below a GOOD point you like about AI.
• Post below a BAD point you don’t like about.
• Post below a BETTER point you think could be better about AI.
Next Lesson: GBB Practise #1
GOOD: speed
BAD: dangerous limitations
GOOD: clarity on how best to engineer effective prompts to protects us from those limitations
GOOD: Control: Humans can exert control over the AI, directing it towards useful and relevant outputs.
BAD: Contextual Limitations: AI may not fully grasp the nuanced or emotional context behind human prompts.
Encourage Balanced Use: Educate users to balance AI use with independent thought and problem-solving.
Good: condenses information into a very readable, concise and understandable response
Bad: privacy concerns is a real limitation
Better: ensure you don’t just rely on Chat GPT but also think about the answers, that is a sort of hybrid approach
G: Fast response times. Awesome programming skills
B: Acts as if the final word is correct. And it is not always!
B2: Provides concepts and connections that I may not have considered.
Good: Accessibility – AI and ChatGPT provide access to a vast amount of knowledge and information, allowing people to learn and educate themselves conveniently.
Bad: Emotional intelligence – AI lacks emotional intelligence, which plays a crucial role in effective teaching and learning, limiting the holistic development of individuals.
Better: Ethical AI education – Educating individuals about the ethical considerations, biases, and limitations of AI systems can empower them to make informed decisions and navigate the AI landscape responsibly.
G: Fast response
B: Limited depth of response
B2: Utilises the nuances and subtleties of language (NLP), culture, science and context to give a more insightful and emergent responses
ChatGPT frames answers in a natural language, in an understandable context.
ChatGPT answers appear to be accurate, but many times they are not.
ChatGPT could self-assess the accuracy of its responses, to assist users.
I like the speed of response and for the most part some interesting and valuable responses.
It’s not personalised and doesn’t pick up on nuances well.
It would be better if it knew your preferences better – though that is a bit scary too!
G: Very rapid research and summary reponses on wide ranging and complex questions including creative output.
B1: Apparent authority not necessarily justified due to process / source integrity issues.
B2: Open library of Chat GPT responses, subject to open source review process to academic standard leading to validated reponses and response databases.
CharGPT can give good results, but only if well.guided.
It can give totally invented results.
It would be better if it had live data available to it.