Please enable JavaScript.
Coggle requires JavaScript to display documents.
Artificial Intelligence and Machine Learning - Coggle Diagram
Artificial Intelligence and Machine Learning
Cybersecurity risks
Scenario 1: AI deepfakes
Options: treating voice identification as a standard risk or mandate hardware security protocols
my choice: option 2. This is because human senses are not reliable indicators against AI synthesis
ethical theory: Ethics of care. This is because it focuses on preserving cooperative networks and is safer for vulnerable employees
Scenario 2: Medical AI models
Options: using fast AI outputs for medical advice or enforce strict filtering and manual human validation
my choice: option 2. This is because it uses dual verification and is more accurate in the long run. Medical decisions should not be quick and easy because sometimes they can be life or death.
Ethical theory: Universalism. This is because AI can make mistakes and be biased at times. Medical decisions is not something that should be done fast and pushed aside.
Privacy and security challenges
Scenario 1: Public surveillance
Options: total tracking for safety purposes or restricting facial recognition for emergency monitoring only.
My choice: restricting facial recognition capabilities. This will protect citizens and allow them to feel safer in other spaces. Along with this it protects public trust.
Ethical theory: Utilitarianism. This is because the harm of doing facial recognition all the time would outweigh the gain of doing it. It would lose a lot of peoples trust and feel like constant monitoring.
Scenario 2: Corporate data scraping
Options: use all public data as free access for the company (as well as use it for training) or implement opt-in structures to protect privacy
my choice: Implement a system to protect the employees privacy. This shows the expectation of privacy is important rather than using their creative data without consent.
Ethical theory: Kant’s ethics. This is because human creators must be treated as end in themselves. People and their ideas are not transactional.
Cybersecurity advantages
Scenario 1: Automated data patching
options: full autonomous live data patching or enforcing human in the loop approved data fixing
my choice: option 1 because cyber attacks happen during slow manual intervention. Live will be quicker and allow for less time for these attacks
Ethical theory: utilitarianism. This is because establishing an AI system that protects networks will allow for the best benefit.
Scenario 2: intelligent network decoys
Options: allowing AI to execute counter hacking or keeping AI inside and make it passive
my choice: option 2 because counter strikes can harm other systems you aren’t trying to harm
ethical theory: virtue ethics. This is because security institutions should use virtues of restraint over recklessness.
Ethical challenges
Scenario 1: AI hiring system
Options: keep AI for speed and apply manual tweaks or don’t use the tool and rebuild for equity
my choice: option 2. This is because there can be injustices in the AI hiring platforms that harm people. It also doesn’t allow for equal opportunity
ethical theory: Universalism. This is because AI holds biases that can very much harm people of different races, gender, and backgrounds.
Scenario 2: Weapon system using AI
Options: use full autonomy to eliminate casualties or mandate human in the loop lethal launch authorization policy.
my choices: option 2. This is because humans lives can be at risk and we shouldn’t have a system that makes mistakes do this work
ethical theory: rights based ethics. This is because every human has a right to live and an automated math equation doing this work isn’t always reliable and would harm others if it is done incorrectly