Honest Data Science Talks

What in the Name of ML and AI!

#HDST

This article is Talk 01 of Honest Data Science Talks series. Each talk of around 500 words length, is personal thoughts and deliberations on the domain.

I met these people from a company that claimed to have been teaching AI (Artificial Intelligence) and ML (Machine Learning) to school kids and teachers. No, not even to high school but to lower and primary. Being curious to know what the program was all about, surprisingly, most of it was using a website with simple animation buttons to move objects up and down. I had seen such sites (exactly same ones) and tools earlier and they were not new inventions. What worried me most was that AI being labeled as basic programming. Okay, the much more significant concern was fooling out a community of people who really aren’t aware of AI and falsifying the conceptual background.

I pity the universities for once who taught Data Mining and Analysis, renamed the same syllabus to Business Intelligence and Analytics over the years, and then renamed the same to Machine Learning and now calling it as Artificial Intelligence (wow!). You ask me what? Surprisingly, the textbook/content for all four is almost the same. I know it’s a LOL moment (ROFL maybe). Oh, okay, no, I pity the students. I pity the new generations which have been forsaken with old things on a new name.

Using a statistical algorithm already known for years on different data sets and then publishing a paper with graphs and analysis (not really analysis) with specific data results cannot qualify for ML or AI research. Using an API, not knowing what it does internally and presenting the results is not AI. In one of the conferences I attended, swarm Intelligence results were compared with another API result not related to and presented that the latter did better. The presenter had no idea what swarm intelligence did as the inbuilt API did everything! To the hell, that it was never questioned and was accepted with no interventions. The sad state of affairs is that it’s going to be someone’s published research work under a degree program.

Doing it because everyone is doing is not intelligence. Maybe that is artificial intelligence heavily stressing on ‘artificial’ which is damn away from ‘natural’.

Restarting an old dumped project and branding under Smart Application is not really AI.  A smart city does not mean that it requires a ML or AI solution. A data scientist can never be certified with AI for there can be no one constant syllabus that certifies that, in the true meaning of its name. Using programming language tools to train and test and then predict with a probability is probably NOT AI at all.

It’s not a question of one’s work’s authenticity or credibility. It’s a question of branding and labeling. Are we doing the right thing is the question to ask. This pretty much reminds me of a phrase I had read somewhere which said – to verify if ‘m’ was silent in ‘masses’ that you followed.

I am sure that the AI is crying out somewhere to the ear-less community.

AI, as I understand, is NOT about machine intelligence. It’s about human intelligence. As long as we are not programming the machine to behave dumb and stupid, we are not programming humans. 

 

You are in Talk 01

Go to Talk 02: The Data Villains

14 thoughts on “What in the Name of ML and AI!

  1. Reblogged this on Its PH and commented:

    I have started a series post about Data Science (Much talked Buzz words) under the title – “Honest Data Science Talks”. My plan is to keep it least technical. Two posts are up already and first one can be reached here!

    Liked by 1 person

  2. Got a whole new perception about AI and how people are misrepresenting the concept of AI. The AI now what we are learning has ultimately boiled down in the terms of machine learning and deep learning which is not the case. Students need to be taught about the real thing of AI before touching into the specifics of it.

    Liked by 1 person

  3. Most of the people now a days mean AI and ML as just using some programming tools and API to predict data.But, it is actually not so. Its how one can analyse the data in the efficient way.Its all about human intelligence.

    Liked by 1 person

  4. This is what we did in our ML course. Used APIs, especially of boosting techniques and retained the one which resulted in higher accuracy with only abstract knowledge.
    To be honest, I am still not completely aware of the principle and implemetaion of most of the APIs.

    Liked by 1 person

  5. In my opinion, a small community is working on real artificial intelligence. I feel most of the theories from mathematics with proper modeling interpretation will give a breakthrough to artificial intelligence. All we are doing it is to use the current work on AI and showing some results. People are using it to give out some results and say that’s their contribution coming to live light while the real working community is hidden. They guess some support to such people might give a new dimension to AI.

    Liked by 1 person

Leave a Reply to Nikhil Cancel reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s