AI and Machine learning

AI and Machine learning

1. machine learning

   • artifical creation of human-like intelligence
   • the machine will work out the rules itself
   • structured data
     • data has to be organised
     • unsrtuctured behavior has not been organised
     • Machine learning is very powerful at finding patterns in unstructured learning

2. ai and jobs

   • robots have already replaced many jobs like
     • core manufacturing
     • electronics production
   • up to 40% of jobs could be replaced by automation in 2035
     • wealth differene
       • if a job is lost to AI then the company may get richer whilst others are getting poorer
       • however technology reductions such as electricity, it has created many new jobs
     • bias AI
       • consider bias when using AI
         • if the training data is bias, then it may exclude minority groups or views

3. Ethichs of AI

   • Automatic vehicles
     • AI is used in self driving cars
       • come across ethical dilemas
         • choices like either hitting a pedstian or a cycler
         • hard or ethical rules need to be used in the system
   • ethics of AI
     • it is important to consider the negative effects
       • how reliable is technology
       • how secure it the system

4. image recognition

   • problems with facial recognition
     • calculating the difference between points on the face
     • the face isn't looking at the camera directly
       • points on face will differ even if the head is rotated slightly
   • colour resolution
     • 2 bits of colour allow you to use four different colours (1 bit = 2 colours
       • e.g 2 to the power of 10 = 1024 colours.
   • Example: Identifying a Symbol (tick)
     • should be white around the tick
     • a long line left to top right
     • thicker in the bottom
     • smaller from bottom to top

5. Turing and Chatbots measuring intelligence

   • processing speed
     • 1Hz = 1 cycle/process
       • 1KHz = 1000 cycles
         • 1MHz = 1000KHz
           • 1GHz = 1000MHz
     • we can measure the ability to preform like chess,predicting facts etc
   • General behaviour (AI) hard
     • learning to peel and orange
       • confronting an unhappy friend
         • reviewing a film
   • the turning test
     • question: what is your primary goal in life?
     • Human responses are simple and not explained alot

6. do better

   • ethics of ai
   • facial recognition
     • describe what is a human and what is not
       • humans have 2 symetrical ovals on them
         • WHY AM I SO WEAK
           • crying is weak
         • those ovals have a small dot