AI is when machines :
- Exhibit Intelligence
- Perceive their environment
- Take actions/ make decision to maximize chance of success at a goal
Elon Musk was wrong The AI Singularity Won't Kill Us All.
Most people working in AI have a healthy skepticism for the idea of singularity. We know how hard it is to get even a little intelligence into a machine, let alone enough to achieve recursive self improvement.
Don't fear super intelligent AI remember we could unplug the machine.
Cognitive computers are:
- Made with algorithms
- Knowledgeable ONLY about what taught
- Control ONLY what we give them control of
- Aware of nuances and can continue to learn more
Cognitive computer (Algorithm) Can
- Do very boring work for you
- Often make better more consistent decisions than humans
- Be efficient won't get tired
Exhibit intelligence
- Transfer human concept and relationship
Dependent on experts
- Subject Matter Experts (SME) Availabilty
- Lawyers
- Machinishts
- Physicians
- Insurance Adjusters
Usually not EXPERIENCED in machine learning.
- Need close collaboration with those making algorithms
It is only as good as data and time spent improving it
Creating an AI requires
- Algorithms
- Documents
- Teaching
- Iteration
- Ground Truth
- Repeat
The goal is saving time, machine learning creates a more highly trained specialist not an ''All Knowing Being''
Examples of AI and Cognitive Computers
Consider for each example:
- What intelligence does the system need?
- What the AI perceiving in their environment?
- What actions are taken to maximize success of goals?
- Watson developed for quiz show jeopardy , won against champions in 2011 for 1million $
Humans teach what we feel is important , teach them to share our values.
Super knowings not super doing.
Can we trust machines as much as well trained human?
