Posted Nov 16, 2014 by Martin Armstrong
QUESTION: Good day Martin
I have been spending some time reading up on IBM’s Watson.
As you probably know they are developing a thinking computer which thinks and learns as time goes on. They are creating an Echo system which will integrate healthcare, finance and the public sector with Watson. IBM is spending over $1 billion and using over 3000 people into Watson at this stage.
My question to you is if Socrates is to be built for man to live a better life and learn from its mistakes , wouldn’t it make sense to help integrate Socrates with Watson or work in the Watson’s Echo system?
You have a limited amount of money( $100 Million vs $1+billion) in implementing your goal.
Socrates could become even more powerful in other area using Watson’s cognitive abilities .
ANSWER: The approach of Watson may appear to be similar, but it is very different at the core. It is not a question of money. We have spent far more than $100 million over the past 20 years adjusted for inflation that well exceeded $1 billion today. It was more than $100 million just to produce that chart alone in current dollars. If you use an induction system from which a computer can study and learn like a child through experience, all the programmers in the world will not help. The extent of the knowledge base becomes the extent of the data set. Collecting that data set is priceless.
IBM clearly states that Watson needs you to learn. That is induction. Assuming they have the correct code to create the dynamic links necessary for fully cognitive pattern recognition and association, then machine learning can match that of the human mind.
Socrates is well beyond Watson in finance. Watson will still be prejudiced by its input. That is the mistake IBM is making in the field of economics. This is a huge problem in trying to really reach the state of a all-knowing mentor. Watson should perform well in medicine and physics where the data is not based upon opinion nor the prejudice that plagues economics. Just look at this battle between Marxism/Socialism and Capitalism. This is not based upon data. You have people like Piketty running around who still preach Marx and THINK that everyone should be equal. This is not science. That is like saying the world should be flat not round. You cannot change the economic system and how it functions. It really is the Invisible Hand that is the dynamic link which binds everyone everywhere together. We need each other to make the whole.
Then you face the task of teaching the computer HOW to analyze something. Do you employ the standard linear cause and effect, or do we embark of an entirely new level of analysis that embraces the full dynamic structure of the universe? This lies at the very core of the design question. Programmers cannot write code to accomplish something they do not understand. I am fortunate to have the design and engineering background augmented with programming that enabled me to draw on that field to write code based upon my observations driven by client demands worldwide. Trying to communicate between the two fields is a near impossible task. It requires training and experience in both. I just happened to be at the right place at the right time.
Piketty and others are still very much one-dimensional. This prevents machine learning in the field of social science. We cannot seem to shake free of these biases. This is the stark difference between Socrates and Watson in the field of social sciences. You cannot inject personal opinion or the analysis will fail.
Google is attempting to adopt the approach of a more dynamic thinking machine funding based on the work of Ray Kurzweil who wrote the book – How to Create a Mind. This is a realistic picture of how the mind functions. It is all really pattern recognition reduced to its basic seed.