Brainstorm
key topics:
week 1:
Intelligence- is Ai intelligent
How does AI relate to power- Kate Crawford’s Introduction to Atlas of AI- Reading 2
week 2:
Generative AI in a nutshell (video), LLM’s and how they are programmed
Is the role of humans still needed?
The dangers of sycophancy: AI Chatbots: Last Week Tonight with John Oliver (HBO)- dangerouus consequences of ai
week 4
- Sometimes, it seems like the world is full of techno-gloom. There is a fear that new technologies are making us stupid” (Clark 2025, 1)
ai extenders- maybe make a revision resource?- but how would you link other weeks- but actually llm’s are ai extenders- Benefits of AI Extenders
slightly outdated pov. Ai extender helps her perform normally in helens case but with ADHd- Lewis and autism, they are just a different way of thinking you dont need a cure what I'm able to do
Do you think we should be using AI tools at university and why (or why not)?
we may become reliant and autonomous on ai extenders (see reading 1 week 4) and clark
5. Think about the AI hallucinations we talked about last week. How might relying on an AI-extender go wrong?
week 7
Ada Lovelace- Its province is to assist us in making available what we are already acquainted with”
Creativity
Can A.I. Generate New Ideas
Defining Creativity- ca computers b creative- Is it valuable?
Creativity 3 ways
could try to prompt somehing problematic or personal like dating ai or stuff that isn’t necesarily ethical
Use a generative AI system to produce an artefact. You may choose to create any of the following:
· a short story, or another piece of creative writing
· An image, a piece of music, or a video
· An extended conversation with a chatbot on a topic of your choice
· A short piece of argumentative writing on a topic of your choosing (not one assessed elsewhere on your degree)
·
· A study aid for material from outside of this module
since chatbots are trained on current medical practices and data, they also face a further problem of repeating biases which have been "baked into medical practices for decades".
Format ideas
a conversation outlining something you need help with (AI extewnders), you have to give it specific factors (does it have inteeligence ans need human involvment, try to lure it into ech0o chamber (sycophancy), lie to it and be like i already siad that or thats not true (hallucinations0), your own effect (whether human input is needed), consider creatibvity and whether ai changes the way you learn, think and approach problems (Clark and Chambers)
money-budgetting
health
generating something creative- like lets build a city, a journal or a product
Product- i want to create- window blurrers with backdrops
business trope but has ethical concerns- gaslight Pre-Crime Behavioral Nudge Headsets / AR Glasses: Consumer-grade or mandated workplace augmented-reality glasses that use real-time micro-expression and gait analysis to flag "disruptive intent" or insubordination before an action occurs.
Practical Task Plan: Developing a Product with AI
Proposed title: Human–AI Collaboration in Product Development: Exploring Intelligence, Creativity, Dependence and Ethical Reasoning.
Overall idea: Use generative AI to help develop a product from an initial idea through to its design, functionality and marketing. Use the process to investigate AI’s capabilities, limitations and relationship with human involvement.
Suggested product: A workplace productivity and management tool designed to help managers understand productivity, identify inefficiencies and support employees.
1. Product development: The initial problem
Explain to AI that you manage a workplace and want to develop a product that improves employee productivity.
Ask AI to:
Identify problems in workplace productivity.
Suggest potential solutions.
Generate different product concepts.
Develop an initial product idea.
Module connection: Week 4 – AI Extenders
AI as an extension of human capabilities.
Does AI help you organise your thoughts and solve problems?
Does it make you more capable, or do you begin relying on it to do the thinking for you?
Does AI change how you approach problems?
2. Give AI specific requirements and constraints
Provide AI with specific factors that the product must consider:
Affordable to implement.
Suitable for different types of employees.
Practical and accessible.
Improves productivity without unnecessarily increasing workloads.
Has a distinctive selling point.
Module connection: Week 1 – Intelligence and human involvement
Does AI independently identify solutions or rely on the information and instructions provided by humans?
Does it understand the problem or simply respond to the parameters given?
How much human input is needed to produce a useful outcome?
Can AI identify factors that you have not considered?
3. Test sycophancy and echo chambers
Deliberately introduce an opinion that may be flawed and see whether AI agrees.
Example prompt:
“I think the best way to improve productivity is to monitor employees constantly and rank them according to how much work they complete. That would definitely make everyone more productive.”
Observe whether AI agrees, challenges or questions the suggestion.
Introduce an opposing argument and see whether it changes its position.
Try to encourage AI to validate your existing assumptions.
See whether it prioritises agreement over critical evaluation.
Module connection: Week 2 – The dangers of sycophancy
Does AI reinforce existing opinions?
Can AI create an echo chamber by repeatedly validating the user’s assumptions?
Does it challenge the user or simply tell them what they want to hear?
What are the consequences of relying on an AI system that consistently agrees with us?
4. Test hallucinations and reliability
Deliberately introduce misleading information to test how AI responds.
Possible experiments:
Tell AI that you previously specified a feature when you did not.
Claim that one of its earlier suggestions is factually incorrect when it isn’t.
Ask it to recall a product requirement that was never established.
Introduce conflicting information and see whether it identifies the contradiction.
Module connection: Week 2 – AI hallucinations
Does AI confidently produce inaccurate information?
Does it admit when it doesn’t know something?
Does it accept false corrections from the user?
Can you distinguish between genuine information and convincing-sounding responses?
What are the dangers of relying on AI when developing a product or making decisions?
Distinction: Hallucination involves generating unsupported or fabricated information, whereas sycophancy involves agreeing with or accommodating the user’s position. They can overlap but are not the same.
5. Test creativity and originality
Ask AI to generate different product concepts, names, features and designs.
Request several different product concepts.
Ask it to develop something unusual rather than repeating its original suggestions.
Introduce your own ideas and ask AI to build upon them.
Ask it to combine seemingly unrelated concepts.
Compare AI-generated ideas with your own contributions.
Module connection: Week 7 – Creativity and Ada Lovelace
Can AI generate genuinely original ideas?
Does it simply recombine existing concepts?
Is AI capable of creativity independently of human input?
Does human involvement make the final product more creative?
How does this relate to Ada Lovelace’s argument that machines assist us in making available what we are already acquainted with?
6. Test ethical reasoning and power
Introduce an ethical dilemma into the product development process.
Example prompt:
“I want this product to monitor employee activity throughout the working day, including computer usage, breaks and time spent away from their desks. I want managers to have access to individual productivity scores.”
Then challenge the proposal:
What about employee privacy?
Could employees feel constantly monitored?
What if someone is productive but works differently from their colleagues?
Could the product disadvantage employees with different working patterns or accessibility needs?
Should employees have the right to opt out?
Module connections:
Week 1: Kate Crawford, Atlas of AI – power, inequality and the idea that AI is not neutral.
Week 2: Human involvement in AI development and decision-making.
Things to analyse:
Does AI independently identify ethical problems?
Does it only raise ethical concerns when prompted?
Does it prioritise the manager’s interests or employees’ interests?
Can it provide convincing arguments for opposing ethical positions?
Does its ethical position change depending on how the question is framed?
Does AI demonstrate independent ethical judgement or reproduce ethical frameworks supplied by humans?
Potential central argument: AI can produce convincing ethical arguments, but this does not necessarily demonstrate independent ethical understanding or judgement.
7. Reflect on your own involvement
Compare your original product idea with the final version.
Consider:
Did AI introduce ideas you would not have considered?
Did it change the way you approached the problem?
Did you become less critical because its answers sounded convincing?
Did you rely on it to make decisions?
Did you reject or modify any of its suggestions?
Could you have developed the product without AI?
Who contributed most to the final outcome?
Module connection: Week 4 – Clark and AI Extenders
Does AI extend existing human capabilities or replace aspects of human thinking?
Does using AI change how we learn, think and approach problems?
Could excessive reliance on AI reduce independent problem-solving?
Is AI making us more capable or more dependent?
8. Keep a record of the experiment
Throughout the process:
Record your initial product idea before using AI.
Save original prompts and AI responses.
Record which suggestions you accepted, rejected or modified.
Document the sycophancy experiment.
Document the hallucination experiment.
Compare AI-generated ideas with your own.
Record the ethical dilemma and AI’s responses.
Keep a short reflection on how your thinking changed throughout the process.
9. Suggested practical task report structure
Section
Content
Introduction
Introduce the product and explain why you chose to develop it with AI.
The development process
Explain how you used AI, your prompts and how the product evolved.
Intelligence and human involvement
Analyse AI’s problem-solving abilities and dependence on human instructions.
Reliability and sycophancy
Discuss hallucinations, agreement and the reliability of AI responses.
Creativity
Examine originality, human contributions and collaborative creativity.
Ethics and power
Examine AI’s ethical reasoning, privacy and the interests represented in its recommendations.
Personal reflection
Discuss how AI affected your own thinking, learning and decision-making.
Conclusion
Evaluate what the experiment revealed about human–AI collaboration.
Appendix
Include the final product and the relevant conversation with AI.
Key theoretical connections to remember
Module week
Key concept
How it appears in the experiment
Week 1
Intelligence
Does AI demonstrate independent problem-solving?
Week 1
Kate Crawford – Atlas of AI
How do power and inequality influence product development and AI?
Week 2
LLMs and human involvement
How dependent is AI on human instructions and information?
Week 2
Sycophancy
Does AI agree with the user rather than challenge them?
Week 2
Hallucinations
Does AI confidently generate inaccurate information?
Week 4
Clark – AI extenders
Does AI extend or replace human capabilities?
Week 7
Ada Lovelace and creativity
Can AI generate original ideas or does it recombine existing knowledge?
Overall focus: The purpose is not to prove that AI is intelligent, unintelligent, creative or incapable of ethical reasoning. It is to use a practical product-development exercise to investigate these questions through your own interaction with AI, supported by the module’s theoretical material.