Thursday, October 26, 2017

Tech Giants Paying Huge Salaries for Scarce A.I. Talent : Why it matters

An interesting article in New York Times “Tech Giants Are Paying Huge Salaries for Scarce A.I. Talent (link)” is making rounds in social media and among the digirati.  Artificial intelligence has fascinated technologists and science fiction writers for decades, but the business world seems to be getting serious about its disruptive potential. Technologies including Deep Blue from IBM, DeepMind from Google or Microsoft’s Chatbots are going beyond press-mentions, and beginning to demonstrate value in solving real world problems.

A.I. also continues to be on “top 10 or “top 25” Digital Startup ideas. Promising AI and Machine learning focused startups are frequently being courted and acquired by the tech oligopoly — Apple, Amazon, Facebook, Google and Microsoft (link). In many cases, executives see it as an opportunity to on-board a pool of talent more than just acquiring a promising technology.

The author, Cade Metz, after discussion with “nine people who work for major tech companies or have entertained job offers from them,” explains how tech giants are paying “Huge Salaries” for AI Talent. A few key points from the article


  • In the entire world, fewer than 10,000 people have the skills necessary to tackle serious artificial intelligence research  
  • At the top end are executives with experience managing A.I. projects. Anthony Levandowski, a longtime employee who started with Google in 2007, took home over $120 million in incentives before joining Uber last year.
  • Typical A.I. specialists, including both Ph.D.s fresh out of school and people with less education and just a few years of experience, can be paid from $300,000 to $500,000 a year or more in salary and company stock
  • Costs at an A.I. lab called DeepMind, acquired by Google for a reported $650 million in 2014, when it employed about 50 people, illustrates the issue. The lab’s “staff costs” as it expanded to 400 employees totaled $138 million. That comes out to $345,000 an employee.

Some of these are broad generalizations and sound like “in the entire world, fewer than 1,000 researchers are working on the cure for XYZ cancer or ABC disease.” One can discount such hyperbole since the author got most of his inputs and figures from just “nine people.” Still, the premise of the article is still logical and rather straightforward:

“Tech’s biggest companies are placing huge bets on artificial intelligence, banking on things ranging from face-scanning smartphones and conversational coffee-table gadgets to computerized health care and autonomous vehicles. As they chase this future, they are doling out salaries that are startling even in an industry that has never been shy about lavishing a fortune on its top talent.”
Let us look at some of the implications of Why and to Whom this matters:


  • Technology Executives: In a classic case of “Airline Magazine Syndrome,” functional leaders and executives across businesses are beginning to lean on their IS Executives to demonstrate how they leverage Artificial Intelligence, big-data, visualization, robotics and other digitization techniques.  Technology vendors are sensing this opportunity and are cleverly rebranding their CRM, ERP and other products as “AI based,” sometimes by just adding cool-new chatbots to the existing platform. 
    • It is the responsibility of Enterprise Architects and technology leaders to see through the emperor’s clothes. 
    • Technology leaders can use this opportunity to engage and inform their stakeholders, and help them contextualize relevant user-stories and requirements where they will demonstrate value
  • Consulting firms: Consulting firms and System Integrators are jumping the AI bandwagon by adding offerings to ‘Digital Transformations.’  
    • It is necessary for consultants to stay abreast of emerging technologies. However, consultants must also take an objective view of their client’s requirements. While AI holds a lot of promise, in some cases it may just be a solution looking for a problem. 
  • Software Engineers: The article highlights how “companies like Google and Facebook are running classes that aim to teach “deep learning” and related techniques to existing employees.”
    • If you happen to be an engineer selected to learn “deep learning,” great. 
    • Otherwise, you can explore opportunities for self-paced learning on Fast.ai, Deeplearning.ai  etc. Keep in mind a certification or training alone may not suffice if your organization is not embarking on an AI based initiative
  • Startups and entrepreneurs: Many startups and entrepreneurs are looking to carve out a niche in this greenfield space 
    • If you aspire to be acquired by “the tech oligopoly,” you should focus on innovative application of AI and ML. However, this is much harder than it sounds. Such real world applications of practical value are not easy to visualize. 
  • Students of Computer Science: There are several ‘hot’ and emerging technologies competing for our mindshare, though Big Data, Robotics, Automation, AI and ML stand out. 
    • As a student of Computer Science, a specialization in AI and machine learning may help you stand out from the crowd. 
    • A specialization in these technologies will certainly help you land a better job, but don’t be under pipe-dreams of “$300,000 to $500,000” payouts. Those are going to be much harder to come by. 

I’m sure this is not the last word on this topic.



Thanks for reading! Please click on Like, or Share, Tweet and Comment below to continue this conversation | Reposted from my Linkedin Pulse article

Sunday, October 22, 2017

Looking at the current scenario of too many immigrations to Canada, will it create lack of employment opportunities over there?

This was an interesting question that came via an online forum. The questioner adds "The trends which I am following make it very confusing to decide upon whether to opt for Canada or not."

My Response

There is an old tale of the famous explorer Vasco da Gama (link) landing in the coast of Kerala



On hearing of the landing of ‘foreigners’ on their shores, the king of Calicut sent an emissary with a glass brimming with milk.



On receiving the glass, Vasco da Gama quickly understood the message, and asked his assistant for spoon of sugar that he deftly mixed in the glass without spilling a drop.

The emissary went back to the king with the message, and the King smiled and welcomed Vasco da Gama and his entourage with fanfare.

The message back to the king of Calicut was simple: just like the spoon of sugar mixes and devolves in a glass brimming with milk without spilling it, we too will blend in without spoiling your equilibrium.

Immigrants to Canada, US and elsewhere have traditionally taken this approach of ‘sweetening the milk’ without spilling it or messing with the existing equilibrium.
This is likely to continue as long as newcomers come in with an open mind by enriching the host nations, and are willing to carve out opportunities without spilling the existing milk.

Monday, October 16, 2017

Indian Roads: the final frontier for Autonomous and Self-Driving Cars

I happen to be among a rare breed of NRI/PIOs who are as comfortable driving in Anytown USA as they are on busy and chaotic Bangalore roads. This said, it takes a few days of ‘acclimatization’ riding on Olas, Uber and Auto Rickshaws before I gain confidence to get behind the wheels. During my current stint in Bangalore, while driving a small Maruti car, I have been musing on former Uber CEO Travis Kalanick’s statement that “India will be the last place to get autonomous cars.”

When Kalanick made this proclamation a few months ago, it raised a few eyebrows among the digirati, and proponents of Autonomous vehicles and AI.

The technologies behind autonomous cars are advancing at a fast pace. Billions of dollars are being poured into it by Automakers and rideshare companies. Almost every day we see news of ‘yet another’ innovation in self-steering, LIDAR, GPS, Digital Maps and related technologies including AI and robotics. (link to a couple of recent deals) Some of these ideas and technologies sound futuristic and SCI-FIish, though many of these technologies are beginning to appear in high-end automobiles.

After driving on Bangalore roads for the past few months, I will have to concur with Kalanick’s proclamation. And here are 10 reasons why Bangalore roads will be the last place on earth to get autonomous cars:



#1. Right of way? When it comes to use of roads, everything and everyone has a right of way, Including pedestrians and cows. Drivers may swear at cyclists, pedestrians or even guys riding horses ziz-zagging through traffic, but shrug it off as par for the course. Apparently the guy riding a horse has as much right to be on the road as my Maruti.
Implication for designers of autonomous cars: Any self-driving technology will have to accommodate for erratic and unpredictable presence of vehicles and non-vehicles on roads that claim an equal right-of-way.



#2. Might is right – Paraphrasing George Orwell, when it comes to Indian roads, all vehicles are equal, but some vehicles are more equal than others. Bangalore traffic follows an informal pecking order with Public buses at the top of the food chain, followed by ‘G’ plated government and police vehicles, followed by larger SUVs, yellow-plated commercial vehicles and so on. Of course, Autorickshaws and bikes seem to have a license to ziz-zag as they please.
Implication: My description of the informal pecking order is merely an empirical observation. Good luck to analysts trying to decipher and codify the complex Orwell pecking order on Indian roads!

#3. Eye-contact – At busy intersections where traffic begins to crawl, pedestrians, bike-riders and others will try to make eye-contact, wave or make other gestures to indicate their intent
Implication: There is obviously a lot more than meets the eye. LIDAR, Digital sensors and cameras will have to be smart enough to speedily decipher such human cues from a distance.

#4. Horn-OK-Please – Trucks and busses on highways routinely have “horn please” painted behind to remind drivers that honking is not only expected but a normal mode of communication. Honking on Indian roads takes an art-form and is not a precise science. Honking can range from benign expression of impatience and a subtle warning to pedestrians and bikers to more serious expression of rage.
Implication: Making sense of a honk requires contextualized interpretation: Try deciphering a single horn, and what it’s trying to communicate from among a cacophony of honks in a busy street.


#5. Streets with potholes, dug up and after monsoons, waterlogged. After the recent rains, unfortunate bikers and scooters in Bangalore trying to avoid new potholes have skid, and been fatally hit by oncoming traffic.
Implication: The challenge is not just about potholes but the unpredictability it induces, requiring split-second reflexes among motorists. LIDAR, DGPS and Digital maps may not be able to predict the next roadblock, pothole or dug-up road.

#6. Dealing with fender-benders – Driving on congested roads with bumper-to-bumper traffic will inevitably lead to fender-benders, or worse. Crowds of gawkers instantly gather around traffic accidents while some mobs are known to vent their 'fury' on helpless drivers.
Implication: One can only guess what happens when a self-driven car meets with an accident. Is it going to call on its AI driven digital assistant, and wait for the police while the other motorist invokes the mob to act?


#7. Dealing with traffic-cops –These men (and women) in uniform try hard to bring a bit of order in chaotic, overcrowded roads. Motorists flagged down by traffic-cops are generally expected to pay a ‘spot fine’ and try to ‘negotiate’ down the fine or plead their case.
Implication: What happens when a self-driven car is flagged down for a traffic violation? Does it to call its digital assistant to help with the ‘negotiation’ or just pay the fine and take the ticket? The social implications here are unclear.

#8. There are rules and then there are un-codified mores. Transport authorities have defined basic rules like ‘driving on the left side of the road,’ ‘overtaking from the right’ etc which motorists are expected to learn before getting a driver’s license. In addition, there are un-codified mores like the art of slowing at intersections and inching forward without stopping and yielding.
Implication: Obvious challenge of designing a system to decipher and work with ever changing, un-codified mores

#9. Driving towards landmarks: Directions from point-A to point-B in major localities are easy enough. It gets trickier when the destination happens to be in a small, narrow by-lanes. Even though I have my coordinates geo-tagged in Ola and Uber, the drivers invariably call to ask for a nearby ‘landmark’ that includes a dental clinic, liquor shop behind the house or the hardware-shop on the 80-feet road.
Implication: GPS and Digital Maps will need a tremendous amount of intelligence (AI ?) to identify and recognize small and large landmarks as humans understand them.

#10. Human Behavior – People in India have been learning to adapt to an increasing number of cars and automobiles on existing narrow roads and lanes. Even in a span of the past 10 years, one can see a remarkable reduction in the number of bicycles on roads and narrow roads morphing into parking spots with hardly any space left for motors to navigate.
Implication: Even the best designed system will have to continually keep up with changing human behavior, and surging population trying to adapt to the constraints.

Some of these observations may sound tongue and cheek, and can certainly be addressed with the right technologies and system design. Even in Bangalore, one might begin seeing autonomous vehicles in ‘controlled traffic’ environments in University, Defense, Governmental and Corporate campuses as the Infosys’ proof-of-concept in their campus highlights.


Perhaps the futurists eyeing the Indian market should step back from self-driving car hype, and focus on an unmet need: a relatively inexpensive electric car that can ‘idle’ smog-free in congested traffic! Wonder what will get Elon Musk to rethink his decision of bringing Teslas to India?
Thanks for reading! Please click on Like, Share, Tweet and Comment below to continue this conversation | Reposted from Linkedin Pulse | [Images in the writeup are googled stock pictures]

Sunday, October 8, 2017

Q&A on EA: What is the future of enterprise architecture? Is it good to become an enterprise architect now?

Here are a couple of questions that came to me via an online forum. My responses follow

Do you believe that demand for Enterprise Architects will keep growing over the next years?

It is well understood that the practice of EA is much broader than IS Architecture. Taking a simplistic definition from Wikipedia
“Enterprise architecture (EA) is "a well-defined practice for conducting enterprise analysis, design, planning, and implementation, using a comprehensive approach at all times, for the successful development and execution of strategy….”
What does this mean? People who can bridge the gap between strategy definition and execution will continue to be in demand. In many large organizations, these folks fill the dedicated role of “Enterprise Architects.” In smaller organization, it may not be a dedicated role but rather a senior executive or manager also taking on the role of Enterprise Architect. Some organizations may supplement the role by engaging external consultants.
Regardless of how organizations approach the role, the need for Enterprise Architects will continue to grow.

----------------

In which field should I do masters if my long term goal is to become an enterprise architect?

A very interesting question. Enterprise Architects that I have worked with come from a variety of technical, business and functional backgrounds. A few had masters degrees and some even had PhDs: An EA I worked with had a PhD in Physics and had worked an CERN before he came to the corporate world.
So, to answer you question: a strong educational background will certainly help you get your foot in the door to gain technical or functional expertise. Such experiences gained in the corporate world are going to be more valued as you try to become an EA.