” Artificial Intelligence – 101 _ India Strides Ahead ” May 21, 2018

 

 

 

 

 

 

 

 

 

” Whoever leads in AI will rule the World “,  said Vladimir Putin.

It is significant to note that Russia and China are investing more in AI than the world’s most innovative nation, the U.S.  China has created a $ 2.1 billion research infrastructure dedicated to AI in West Beijing.

India is not behind. Indian government and Indian companies hold AI as a disruptive force in bringing about dramatic changes in specially manufacturing (Make in India) to be able to compete globally with leading enterprises. Start-ups, Venture Capital, New Product Development, and Academic Research are being encouraged.

Of significance is the fact that Finance Minister, Arun Jaitley, in his budget speech on February 1, 2018,  emphasized NITI Aayog spearheading a ‘National Program on AI’ including research and development. The budget allocation to promote AI,  Machine Language,  3D Printing and other tech has been doubled to RS 3073 crores this year. Technologies on par with Silicon Valley Counterpart is the chief focus by many companies here.

Major countries are taking steps to empower their Defense Systems by AI. A department of Defence Production in India concurs with the world view that AI will influence our common lives for the better. As with major powers taking steps to empower defence systems by AI,  India also thrusts itself in the same league with Sri Chandrasekhar, Tata Sons, designated as Chairman, Task Force, to achieve this goal. What are the things that the country has to prepare itself for?  What framework in terms of leveraging strength and capabilities of India to develop ? – are what the experts will try to formulate in this regard first.

PwC India report/survey details: 58.74%  say that AI will aid Economic growth;  Cybersecurity;  Global Health;  and Improve efficiencies in imparting education. What is easy and convenient today (use of devices and features in devices) will take on greater responsibilities in terms of education, training, and advising. This is true of AI tools. The survey says.

The decision that business managers and CEOs of enterprises would be challenged with in future in terms of AI use will be to either be  ‘AI advisor confident’  or  be an ‘advisor confident of a judicious AI – Human  mix’  _   experts opine.

There is nevertheless firm consensus  on the fact that Machine Learning solutions (specially replacing repetitive manual jobs) will more likely be a disruptor in sectors as  IT/ITES; Health;  Education;  Manufacturing;  and  Banking/Financials.

            AI  has just begun in India and in Education it is making its presence felt in courses and classes even while facing lack of qualified faculty. But the term “AI” that Professor John McCarthy (Dartmouth College) coined in 1956, when there was  no sign or talk of an efficient algorithm has advanced slowly to receive a sudden, optimistic, and positive momentum now in solving the real world problems. Data explosion is one of the chief factors here.

            Here is what we can possibly take from Australia whose priorities are in (a) Education; and (b) Public Safety.

Artificial Intelligence and Digitisation help transform Australia in specially sectors as important to people/society as  Education and Public Safety. AI market which was worth A$ 333 million in 2016 is expected to grow to A$ 2 billion by 2025.  The country has identified increasing student demand;  competition among institutions;  and lower acquisition costs as factors spurring growth in the education sector. They hope that AI and tech that stimulate AI applications will help bring about technology/innovations even while enabling effective collaboration and engagement among the participants. Benefits to students, training to staff, and meeting the costs adequately for fulfillment have been the chief focus in investment.

Total Education tech market is expected to reach A$ 1.73 billion by 2022. In Public Safety sector covering Law Enforcement; Ambulance, Fire, and other emergency services; the focus has been on “Critical Communications” in specially situations characterised as high risk and unforseeable occasions – a market that is expected to grow from A$ 296 million in 2017 to A$ 373 million by 2022 according to Andre Clarke of Frost and Sullivan research team. Critical events like shootings in offices, schools, public places;  terrorist attacks; call for swift communication. There are providers who unify communications such as voice, video, and messaging across multiple end points which enterprises and other administrators can take help of to keep employees and people safe, for which they are obligated. AdHoc and Everbridge are market leaders in providing tools for such critical communications to formulate a crisis communication system in place.

AI today is dominated globally by Microsoft, Adobe, and Accenture.

Microsoft has done a study in Telengana, India, on school drop outs (specially 10th class) assigning several characteristics and their priorities. They hold it is a lead AI study.

Accenture India lists 5 conditions necessary for the growth of AI in India. (a) Forge a national AI plan and multistakeholder partnerships in key sectors;  (b) Strengthen India’s AI R&D ecosystem including through better international cooperation;  (c)  Create a workforce of the future fostering inclusive workforce models;  (d)  Enable and broaden access to data;  (e)  Embrace smart regulation to safeguard Responsible AI. 

The team also believes strongly that the era of AI is now.

Emma Kendrew,  AI lead for  Accenture Tech., holds a slightly different view. She agrees that hype is reaching a peak, driven by busy corporate marketing machines hoping to take advantage of the possibilities opened up by Big Data & the Cloud as well as burgeoning customer interest.

Andrew Ng, Professor at Stanford University, places AI as the next Electricity. He avers that AI will now play the role in development just like Electricity played in the early 20th Century.

In sum, it may be said that AI is a tech weapon in the armoury of any nation with ambitions to be the ruling force in future.

” Dieu  avec nous “

Monday,   May  21,  2018  –  8.29 p.m. (IST).

Tidbit : ” AI is more important to humans than fire or electricity. It is also important to understand that AI is not a One Shop solution to all problems.”  –  Sundar Pichai,  CEO, Google.

 

 

Standard

” Artificial Intelligence – 101 ” _ Impact Through a Case . May 21, 2018.

 

 

 

 

 

 

 

 

 

 

              Many start-ups claim to do AI when in reality what they actually do is rule based automation systems;

              Many suppliers claim that they have AI systems when in reality what they have is just narrowly defined ML ( a simple software that does image recognition …);

               Many claim that they have completed an AI study when in reality what they did was just a “pilot” study with an aim to proving an element of the concept.

               AI in reality is waging an Automation Battle with promising chances of success. It is the ‘next wave’ of Automation. For success to be real, more time and effort (much more than anybody can anticipate …) are needed. And we can confidently say in good measure that they are being followed in the right track now. This in itself is an excellent beginning.

               Path-breaking announcements in innovation are often accompanied by some associated evils – (pros and cons as we refer to in simple research terms) and AI is not an exception to this. It happened in the case of computers (1960s);  in the case of internet (1990s);  in the case of mobiles (2000s);  but they were not just strengthened with their  roots firmly entrenched, but brought a sea change in the way business and industry accelerated and enhanced their profits while making people richer to enjoy their lives. It will not be surprising if we witness a similar scenario in the case of AI too in about 5 to10 years from now. A major concern that, however, dominates the discussions in professional meetings while seriously occupying the minds of individuals is the possibility of job displacements.

               Forrester Research claims that there will be a 17% loss of jobs;  but there will also be a 10% creation of new jobs; in effect causing a net decline in jobs to the extent of 7%. In the long term, it is expected that this gap will disappear.

               Striking, in particular, is the Study from Stockholm School of Economics (SSE) titled : “The substitution of Labour”  –  a part of 3 year research project on, “The internet and it’s direct and indirect effects on innovation and the Swedish Economy”  funded by Sweden’s Internet Foundation.

               This is just presenting a host of the findings :  The study investigated the potential long term impacts of AI, ML,  and Robotics in the general area of labour replacement.

1.  Innovation of Automation technologies will be gradual;

2.  Long term potential is job creation and not job losses;

3.  Tech advancements are unlikely to result in mass unemployment;

  4.  It will lead to more intense collaboration between humans and machines in a way that will change the focus of labour;

5.  The change will pervade business and industry;  influence how organisations and employees function;  automation of individual activities unlikely to replace majority of working roles and functions in play now;

6.  There will be increasingly tech driven workplaces;

7.  More routine tasks will be replaced by ML;

8.  Some old jobs may disappear, new jobs will emerge;

9.  Humans will work closely with machines and not replace them;

10. Union bosses are less fearful of new methods than about old technologies;

11. Equipment manufacturing, transportation, retailing, warehousing, and wholesale trade sectors likely to replace humans with AI and Robots.

               Some challenges cited are. :

1.   Computers are good at fast tracking tasks; retrieve information from databases, but lack capacity and sophistication to compete with human creative Intelligence;

2.   Computers also fall short in routine production of ideas and conceptual thinking;

3.   Humans do better in familiar noise risky working areas, such as halls, restaurants, and airports;

4.   Machines lack social and emotional capabilities although advances in AI suggest that gap is narrowing, the gap may be eliminated in 30 years;

5.   To be creative one has to make new combinations of familiar concepts which require a rich body of knowledge, Creative Intelligence is the key;

6.   Difficult to replicate in machines;

7.   Sweden hopes AI will be their competitive edge over low labour cost markets in Asia and Eastern Europe;

8.   80% of Swedes positive about idea of AI and Robotics;

9.   Move to AI powered function will be good for society and national economy in the long run;

               Sweden is one of the most progressive and forward looking country, and the findings may not be in full consonance with the views of many less developed nations where the internal dynamics in terms of Social features;  Political ideologies;  Economic strength/turmoil;  Employee preferences and relationships;  Regional differences;  can play a significant role in altering the general perceptions. But the positives from the study cannot be dismissed easily even in the presence of persistent differences in perspectives.

                AI will stay and march its way into the future.

 

” Dieu avec nous “

Monday,  May  21,  2018. –   5. 29 p.m. (IST).

Tidbit  :  ” Are we conceited enough to think that we get Intelligence from machines ?”.

 

 

 

 

Standard

” Artificial Intelligence – 101 ” _ Basics in Brief – May 20, 2018.

 

 

 

 

 

 

 

   

 

 

 

 

 

 

      1.  What is AI ?

Artificial Intelligence (AI) is a forward step in the progress of technological evolution embracing the extension of Automation Strategies for the benefit of enterprises, organisations, individuals, and the society.

AI concerns the ability of machines to mimic the cognitive functions of human minds as learning; reasoning; perceiving; problem solving; and exercising creativity.

 

          2.  History in a few lines :

The techniques used to break the German Enigma Codes by the UK Government Headquarters during/after World War II laid the foundation for Computing and AI. Military and Intelligence communities have been using machine learning (ML), and AI to process vast amounts of data which has not been possible earlier. These developments strengthened by advancements in computing power; enormous amount of data from sensors and mobile devices; plus algorhythmic advances have set the trend for mimicking the cognitive functions of humans in trying to solve real world business, sectoral economics, and social problems efficiently and effectively. Examples of tech that enable AI to solve business problems are : robotics, autonomous vehicles, computer vision, language, virtual agents and machine learning.

 

           3.  What  AI can achieve ?

(a) AI will free humans from the burdensome work of analysing huge amount of information. (b) The possibilities open to human will expand through use of robots, automatic driving cars etc. (c) Internet of Things will connect all people and things;  all knowledge and information will be shared;  and totally new value will be born. (d) Social issues will be overcome and humans will be liberated from various types of constraints.

 

           4.  What is key to growth of AI ?

AI, in workplace specially, is in its infancy now although its potential is limitless. It reminds everyone of the early days of Internet (1990s). Answers to questions of improving productivity; customer service; finding, analysing, and predicting from data patterns;  and improving efficiencies at work and interactions will mark the ‘high’ that AI can reach. It is important to understand that AI is not a simple programming exercise (with instructions) of a computer, but learning and training of the computer itself.

 

           5.  What is the status of AI now ?

It appears that as at present “repetitive” tasks as well as tasks that require humans to process large amounts of information to make decisions are most suited for machine language (ML).   (a)  Say,  you are studying a crowd;  many people have gathered;  all are in some kind of motion;  your goal is to identify/recognise the faces of persons wanted :  ML  and  AI  can help spot the individual in such situations.   (b)  Say, you have developed a cybersecurity software;  the goal is to identify/recognise threats that attack the system you are working on;  it also needs to stop the malware/virus/glitch squeaking through the system  :  Again,  MLand  AI  can help you achieve the goal.

In both cases one needs a huge database in terms of individuals with characteristics and threat features. It also requires the ML training in terms of identification on several cases for effective recognition. So an approach more sophisticated than the traditional ML is called for. It goes with the terminology,  ” High Fidelity ML”.  This can perform tests before and after the threat feature enters the system making the counterattack to eviscerate the glitches effectively.

AI models are now designed to help employees, scientists, individuals, to do better instead of planning to move out or replace people.

 

          6.  Two  AI  examples here :

(a)  AI  in  Municipality :  City of Boston CIO,  Jasach Franklin Hodge, focuses on what AI  can do for municipal employees and constituents. He wants to make humans working with AI tools to make them “Superhumans” without replacing them. He wants to focus on (i) predicting when and where accidents will happen; (ii)  identifying where a food safety violation is likely; and  (iii) finding the individual in a group of opioid users who will be responsive to intervention. In each case he would like to target the remedial course of action more effectively.

(b)  AI in Agriculture :  Microsoft is investigating ways to enable data driven farming techniques such as,  “precision agriculture” where crops are watered or fertilised with exactly the amount they need and no more. Mission is to use AI to deal with looming Global Food Crisis.  Farmers have to double food production to sustain the projected 9 billion people that will inhabit the planet by 2050.

AI  will create new organisations and groups within the companies around data management;  data integrity;  and AI best practices. These are ultimately expected to stimulate growth of companies.

 

          7.  What are the hurdles to AI growth ?

Lack of AI skills in enterprises and the ability to frame an effective AI strategy are the hurdles that experts hold at this juncture. However, there are also a few limitations like :  (a) whether we can comprehend and substantiate why the algorithms are predicting what they are predicting;  (b)  whether they are transparent enough;  (c) whether there is any bias in the data;  etc. However, with time and more use case developments these constraints will also be overcome, foresee analysts and experts.

 

           8.  Some  AI based applications in enterprises/industries :

Sales forecasting in retail;  supply chain optimisation in logistics;  robot advisors in financial services;  transportation route optimisation;  processing medical trial outcomes;  healthcare image recognition for safety management;  medical image analysis such as screening images of retina for diabetic retinopathy.

 

           9.  To conclude :

 AI  is expected to transform the manner in which efficiency and effectiveness combined with accuracy, reliability, and swiftness can be obtained through the available tools for increased throughput and productivity;  improved safety;  higher quality;  reduced waste in industry. It will also help reduce operating and administrative costs for public and private sectors translating into improved operating margins for business. Social benefits cannot be overlooked.

 

       “Dieu  avec  nous”

       Sunday,  May  20,  2018  –  8. 29 p.m. (IST).

        Tidbit  :  Forrester Research :  ”  AI  will kill 17 % of jobs;  it will also add about  10% jobs.  So net decline will be 7%.”

 

 

 

 

 

 

 

 

 

 

 

 

Standard

” 5G Wireless Tech – 101 ” Basics in Brief _ May 19, 2018.

 

         

 

 

 

                  1.  What is 5G ?     5G is wireless tech touted as the ‘paradigm shift’  (next generation network)  in communication with speed and power,  making advantageous inroads into sectors hitherto not reached by 2G,  3G,  or even 4G  earlier.  5G is the backbone of digital economies and societies to say simply.  New applications, new developments,  and new business models will emerge benefiting everyone.

” If  ‘LTE’  has already made its prominent display and secured its place,  5G  will mark the transformation in wireless networks in 3 to 5 years”  –  opine analysts and experts.

 

               2. What is  5G  Wireless Tech’s  Promise ? :   (a)  Lightning fast speed;  (b)   Incredibly low latency;  and  (c)  Capacity to carry massive number of connections simultaneously. 

Today’s  LTE  networks give downloads of 2GB per second.  5G  can give speeds of 20 GB per second.

Average internet speed in our households now is less than 5 mb/second.  At the mobile international conference in 2017 Samsung displayed a 5G component capable of 500 mb/second.

In January 2018, Qualcomm – which is at the Vanguard of 5G tech –  in partnership with 6 Chinese manufacturers  (Lenovo, Oppo, Vivo,  Xiaomi, ZTE, and Wingtech )  began work on 5G rollout in 2019.

 

               3.  What are 5G’s  distinct  categories ?   They are two,  (a)  low and mid band 5G focused on spectrum below 5.9 GHz;.  and  (b)  high band 5G focused on spectrum above 6 GHz.  At this juncture they can augment existing LTE networks but not replace them.

 

4.   What are the challenges to 5G use ?    First, the advantage :  5G will be in the forefront of new (next generation) applications and business models enabling streaming of 4K movies on mobile easy, reliable, providing an enjoyable experience without buffering.  5G  will also enable connected and self driving cars;  smart cities;  AR;  Real-time biometrics like facial recognition/voice recognition;  AI and IoT applications;  Robotic surgery;  and drones.  Fast connections that can bring the power of a supercomputer to a device held in hand is what we, customers, will experience.

 

Simply put,  5G can unlock the potential of IoT by enabling upto 1 m/sqkm  connections at low power.

Challenges  :   Economics of 5G roll out is still being debated; but it’s ability to deliver at low latencies ( 1 millisecond )  beneficially and successfully in time critical contexts as healthcare, utilities, and in vehicle safety systems is definitely not in doubt, aver experts.

That cell towers constitute the chief platform of ‘digital economy’  is well known. They are similar to rail road tracks in the case of rail traffic;  and toll roads in the case of highway traffic.

While the 5G low bands portray similarities with LTE in terms of cell sites density,  5G high bands require a completely different architecture with much denser networks,  i.e.  15 to 20 cell sites/sqkm  against the usual  2 to 5  cell sites in specially highly populated areas now. Also, the cost of deployment is 4 to 6 times higher than LTE.  Sites also take 6 months or more to go live. This means cost and time are big hurdles for 5G,  more so at high spectrum (above 6 GHz).

Hence, it is important for digitally developing countries like India to consider the above factors critically while planning for 5G introduction. Will the economics encourage introduction in selected regions/markets or even more prudently to build an overlay to the already existing tech as a major approach before a move to nationally introduce the network is worth pondering.

  Factors critical to high band 5G deployment like cell tower locations;  fibre backhaul;  and Street Infrastructure,  additionally,   need to be kept in mind. One way to overcome this hurdle is ‘network sharing’  by wireless operators. And not the least, a positive relationship with the regulator is a top priority for the wireless operators entering this horizon.

 

5.  What to conclude ?    5G is here to stay and effect the digital transformation (the next generation networks) with exciting killer uses in store for development, application, use, and benefits in several sectors.

 

     ” Dieu  avec  nous “

Saturday,   May   19,   2018  –   4. 19  p.m.  (IST).

Tidbit  :   ”  India  has  5G  in  its  DNA “.

 

 

 

 

 

 

 

 

 

 

 

Standard