Original article was published by Venkataswamy R on Artificial Intelligence on Medium
Artificial Intelligence has NO future! It will die soon…!
Well, let me write the following simple statements, Say True or False.
I am sure, your answer will be TRUE.
- Artificial intelligence is powerful, advanced and emerging technology.
- It creates lot of jobs in future.
My answer is NO.
You might have heard people quoting some benefits and significance of AI. but here let me mention the technological BAD FACE of AI.
- AI technique is simple and attractive but it has weak internal structure. People are going behind the fancy thing. They are not validating the sustainability of the technology.
- AI technique is static modeling of dynamic process. which completely makes the system as black box. A day will come when you say, “Probably it works, but I don’t know how”.
- AI technique can not understand nonlinear system. There must be some pattern. It always provides wrong result for uncertain input vector.
- Machine learning or Deep learning classifier model is just a linear polyhedral in vector space.
- We can’t rule out multiple equilibrium point or space in a system. AI technique is weak at these point or space.
- We had good old area called system identification which gives dynamic model of dynamic process. Expert require to work on the model. Technologist cornered or not using because it is difficult. AI is a subset of system identification.
- AI never gives the picture of internal component, internal signal, internal process or internal uncertainties of the system.
- AI technique adjusts its parameters without any physical connection and meaning.
- AI technique is not suitable for dynamic system. It can’t learn or consider or process instantly.
- The application associated with AI technique is easy to hack/crack. Decryption is very easy. So its completely insecure.
- Noise in the data set is never segregated in the training.
- Transformed domains are not there in AI.
- Creation of heuristic or measured data is impossible to create model for sophisticated unobservant system.
- Accuracy is defined data not for system. Accuracy is data dependent.
I really unable understand this,
There are 1000 students are there in a place. 10 diversified industries are located in the same place. Each company require 100 employees in next year. Now AI is emerging!!! Most of the students say 800 started learning AI and became expert. Do you think all of them will get job?
If you say Yes, then you probably should agree with the following points,
- 80–100% industry want to use AI technique.
- All those companies have a scope to incorporate AI. Some may be energy, construction, education and everywhere.
- If a company recruits 100 people all of them will work in AI.
- Sustainability of technology is 100%.
- You are assuming the job opening every year without any decline.
My suggestion would be,
- Do not go with mass in the direction of general propagation. Always move orthogonal.
- Stick to core domain with some degree of interdisciplinary capability.
- There are lot of people are there to see your failure. Don’t give them a chance by your negligence.
- Do not look for tomorrow. Look for day after tomorrow.
- Understand/debate on the technology. Do not inspire by others talk.
- AI has no future. It will die soon.