Saturday, 14 October 2017

Artificial Intelligence impact on HR- Thoughts from an novice mind!




The first thing I get reminded of when I think of Artificial Intelligence (AI) is ‘The White Queen’ of movie Resident Evil or ‘Jarvis’ from the movie Iron Man.

During the course of my 5 years of experience working in HR function and having a curiosity to know about technology impacting my own function, I took to reading and keeping myself abreast of the various new technologies. I have during this course also attended couple of conferences- heard speakers, panel discussions and vendor information guides on how AI impacts us and how it can be leveraged. To my large part of the curiosity of how technology, especially Artificial Intelligence impacts HR, I am not too convinced nor satisfied on how the impact could be from what I heard! However, like the popular saying ‘Experience is what you get, when you don’t get what you want’ holds true for me as well. With abundant literature at my disposal and having referred to a few, I do have a made a fair image of my own on how AI can influence HR and its key activities (in my own imagination).

Artificial Intelligence from an HR lens need to understood as a means quantification of data and more data and thereby creating models of correlation, regression, prediction, etc to form patterns and solutions to arrive at the best decision possible from the data sources made available to the intelligence platform/software/whateveryoumaywanttocallit.

I will now decode my understanding from the statement above with focus on the highlighted aspects. What we all know (or rather some still are trying to know) is that AI is based on data of anything (could be a person, practise, etc). What AI does in general is

a.       Quantification: AI at the core is technology and it is known to quantify data. We all know that anything fed into computer is stored in the form of data such as one’s personal data, social media data, etc. All the data available to AI leads to quantification of some kind which could correlate to give a strong sequence or correlation, etc.

b.      Data & More data: AI draws its connection from internet of the world and will have access to all possible information points about anything in particular to a fine specificity. Irrespective of what password you might have for your google account, AI could be so powerful to hypothesis a series of password combinations which you might have had and thereby crack it. Most probably your nick name and something, your family member’s name, etc which you could have spelt out in some corner internet connected part of the world. Any data online about you in person could thereby lead to an explosion of information thereby satisfying the algorithm of the AI resulting into more patterns. Starting from your employee ID to bank account info to post on social media and including the hidden apps which track your face without your permission (contrarily, you have given that crazy app your permission to do so), all of these technologies are tracking data about you or a thing. Did you ever think what google might be doing to the vast knowledge of information generated on Andriod, including the places you travelled on its map?? When we think of it, it’s scary, but true. So the future will see data and more data.


c.       Models:  the first thing when our boss or anyone gives us large volumes of data of any particular activity and asks us to make sense, what do we try to do?. We model it into existing theories, identify what causes what, gather some hints on causal effect triggers. That is nothing but small scale modelling of data to create stories & patterns. In summary, what you are doing is deciding on a future plan based on this data using your models, context and biased future plans. AI knows this well. Any behaviour or emotion can be modelled to give a decision based on patterns of data. Simple analytics is also a sort of modelling of data. The famous Watson Analytics website of IBM models data into patterns. AI can do much more to get to the depths of the modelling. Facebook shut down its AI program because AI developed its own code or language. It was modelling data transmission to the simplest and fastest mode possible than regular English or code written by human brains.

d.      Patterns & Solutions: Imagine a situation where a repeated activity online could result into something meaningful at workplace or personal life. What if we extrapolate this behaviour into a larger situation? This will soon be a reality. AI creatively finds patterns & solutions to throw the best decision available. An example could include a situation where a woman applicant for a job is 1 month pregnant and the prospective employer is already aware of it because AI could pull out the latest lab reports of the woman who had undergone a medical check or found out that she bought a medicine concerning that. While I know its illegal to ask for ‘are you pregnant’ in a formal interview and base your decision on this, but will this data point be completely ignored by the interviewer? Hell No! In all probability, this data influences a decision maker. Not just this but other patterns of purchase, medical visits, online activities can all lead to beautiful solutions putting the person ‘on the spot’.

e.      Best Decision Possible: What is the best decision possible? Or what qualifies for a best decision? Does making decision based on data considered the best decision? Well!! Nowadays it is; data based decision making is considered good off late. But the question in AI world will be how much data based decision making will be good. Decision making is often called an art because there is some intuition, gut, context, consequence, etc., which all have to be thought through. Will AI be able to make the best decision? Imagine a situation where even the softer aspects such as empathy, sympathy, care, concern are coded which actually influence a decision. In the AI world, there will be no intuition. Most often, organizations tend to be supportive to an employee during crisis/critical times. Will AI consider these critical times or play on pure data. If a person meets with an accident and gets partially disabled, will AI be generous to offer consider the human factors or take a hard stand based on the disabled employees future perceived contribution towards the organization and decide? Well, it’s a food for thought.


f.        Data Sources: We are nowadays a mobile-phone beings. With atleast 99% of our data being captured through mobiles in various applications. How many of us in haste click on ‘Allow permission’ to the app which is asking for permission to track more than required data. At least a 99% of us do.
"Our privacy is being attacked on multiple fronts," Cook said in a speech that he delivered remotely, according to EPIC. "I'm speaking to you from Silicon Valley, where some of the most prominent and successful companies have built their businesses by lulling their customers into complacency about their personal information. They're gobbling up everything they can learn about you and trying to monetize it. We think that's wrong.” These are the words of Apple CEO Tim Cook  in 2015.
                We are constantly sharing GB’s of data into the internet system. Imagine a situation where every data gets integrated from whatsapp to gmail to android servers to facebook. That’s whole lot of data pool about you! With modernisation of technology, a whole lot of privacy laws are being broken and we are the subjects. AI in the future can have access to all this data. Another example- Can my whatsapp chat, facebook chat, twitter page all when integrated read a better story about me? I guess yes! These are all the pools of information which are currently not integrated by soon will be.

This is my understanding of AI infused world. I’m not portraying a scary pic (my apologies if I did), but all I am trying to say is AI will surely disrupt the industry and some jobs with the kind of efficiency it could bring to most systems, processes and decision making. The intuitive part of an decision be replaced by the rational part in that decision. The key question I do try to answer at time is ‘Do we really need these large amounts of data to make decision? Or are we just getting into a world where data becomes only filth to such an extent that all privacy barriers are broken?’

Wednesday, 15 February 2017

Uber and Ola Workers Strike! A discussion strengthened…!

The recent strike situation by the App based travel gig workers (Uber & Ola) in Delhi/NCR could throw much light into the on-going discussions relating to the employer-employee relationship status between the company (Uber, Ola) gig worker (workers on strike). While most of such discussions have taken place only in western countries, seldom was this topic discussed in India. The reason largely in my view being ‘there is huge unorganized sector workforce to whom the ‘terms of employment’ do not apply, and so the gig workers been classified in this large unorganized sector’.
On one hand the strike has proved problematic for the company, customers and the workers themselves to an extent, whereas on the other hand it has been a blessing for a few other stakeholders like the government authorities who can now breathe freely (literally) due to lesser traffic on roads and lower pollution levels estimated.
While to some extent the nature of relationship between gig worker and these companies could be established w.r.t payments in the form of wages, guidelines on bonus, guidelines on hours of work, etc. but the core relationship through a formal contract was never established between the company and the workmen till date thereby denying this segment of workforce ‘the gig-workers’ their share of social security benefits such as health insurances, provident funds, etc. The recent strikes could help elevate these discussions to a next level in strengthening the employer-employee bond. The current strike situation could also prove as a base reference points for the courts to create a sharper case to clearly call out a clear employee-employer bond basis the principles of collective bargaining which is formal machinery provided under law to negotiate (which currently is the situation happening)
One big blow in the face of these app based travel companies could come when the workers formally register a union and fight a case. It might be devastating to the business model, but in the larger interest of sounds employment practices, it’s probably the next step ahead. However as a Customer I do want this situation to settle soon, as a disguised Marxist I stand by the next generation workforce in establishing a formal employer-employee bond which bestows greater workers benefits to them, but as a HR professional, I would advise that the ‘gig worker’ model of business is not permanent state in the midst of developing workforce regulations in India and alternative models should soon emerge so as to support a profitable business. 
Further discussion to this view point can be read in the given link http://www.epi.org/publication/uber-business-model-does-not-justify-a-new-independent-worker-category/