Syиphony 2016

In last June I started posting about conferences that I could attend. Syиphony 2016 is another chapter of that habit, but the only difference is it was organized for IFS employees. The conference was held on last Friday (18^th November) at OZO Colombo. It was my first time presenting in a conference with a software engineering audience and, I was much enjoyed and enlightened at many levels due to the diverse of presentations delivered.

The Syиphony 2016 is the annual conference of IFS’s Product Coordination and, Research & Strategy group. Each team and individuals belongs to the consolidated group whom got acceptance for their talk presents their products, advancements of technologies or any other business that makes a positive impact within the organization.

IFS’s Chief Technology Officer (CTO) Dan Matthews who is the director of Research & Strategy group was the chief guest and delivered the keynote speech. Also, Jonas Högberg, who is the director of Product Coordination group delivred another narrative/keynote speech. The event was full of talks ranging from psychology to continuous delivery to data analytics, but in here I would summarize few presentations that I was really impressed.

Kumudu Widyaratne, a lead software engineer from Product Coordination group presented about technical debt. His presentation was focusing on how developers make projects complicated and possible ways out. My fellow Business Intelligence team member, Isham Mohamed presented about stream analytics. This session was more of a practical one with a demo accompanied at the end, which is somewhat similar to Scott Hanselman’s presentation about IoT. But the most enjoyed talk was delivered by IFS Labs software architect Sukitha Magallege, which was named as “On Diapers and Things”. He articulated how to build a fault tolerance system by sharing his experiences being a novice father.

I presented about deep learning and some advances of it. Also, I was able to mention the work being done by Machine Intelligence Research Institute, Future of Humanity Institute and OpenAI in my presentation. Finally, the event was concluded by the closing note given by IFS Product Coordination and, Research & Strategy group director Indrajith Pradeepa. He explained his expeditions of learning about hiking and lessons we could learn for the betterment of both as an individual and an organization.

Syиphony 2016 Organizers. Image credit: Isham

There was an after party at the end of the day, where we got a chance to discuss about the talk and work being done within our group. But most importantly, both me and my colleague Isham got a chance to sit next to IFS CTO and have an inspiring open discussion about his career. The essence of all success we had at Syиphony 2016 was its organizing team and I’m really impressed about their work, kudos to the Syиphony 2016 organizers.

 

 

Are We Living in a Computer Simulation

Elon Musk thinks that is it almost certain that we are living in a computer simulation. He Says that humans are basically an advanced version of The Sims. Indeed this idea sounds pretty absurd. But people used to think all planets including sun orbits around the Earth and almost 2000 years ago Galileo proved that is it not. There for the reality could far different from what we are materializing.

If we are living in a simulation, there is a higher level being, but it’s some version of us

Musk is echoing a paper on this theory by Oxford professor Nick Bostrom. His argument goes like this. 40 years ago we had Pong, like two rectangles and a dot. 30 years later we got The Sims, and now 40 years later we have photorealistic 3D simulations with millions of people playing simultaneously and it is getting better every year.

So let’s go about 10000 years in to the future and it’s possible that when we get there, the entire civilization is gone because there is a celling to our advancement. It could be because of global warming or self replicating robots. Quoting Elon, “… if civilization stops advancing, that may be due to some calamitous event that erases civilization”. But another possibility is that if we keep advancing and assuming everything in the physical world can be simulated, eventually we will simulate ourselves.

Getting enough computer power to run billions of ancestor simulations could be a problem. But Bonstrom thinks we’d send tiny self replicating robots to other planets which would turn the planet into a huge computer and some of the simulations would start making their own simulation. In this scenario there are billions of universes that are indistinguishable from our own. That means chances are we are in one of the simulated universe. And given the other possibility, which is human civilization has an inevitable ceiling. Obviously it’s better that we are in a simulation.

Somehow there is another possibility. Maybe future humans don’t want to run ancestor simulations. Maybe because they think it’s unethical to run ancestor simulations. Because there is a tremendous amount of suffering in the world, and that suffering would still feel real to simulated humans. Or maybe they have other priorities.

However, there are three possibilities, (1) humans go extinct before we are able to run a simulation this big, (2) humans don’t run simulations because it is wrong or boring, (3) we are living in a simulation. Musk think there is only a tiny chance we are in scenario 1 or 2 and the Bostorm thinks it is more like a 20% chance that we are in a simulation but if you are not into futuristic predication,  Bostrom think the argument also provides other rewards. He says, “… it suggests naturalistic analogies to certain traditional religious conceptions”. In other words, if we are living in a simulation, there is a higher level being, but it’s some version of us.

References:

Are You Living In a Computer Simulation? Nick Bostrom. Philosophical Quarterly, 2003, Vol. 53, No. 211, pp. 243-255.

Full interview of Elon Musk at Code Conference 2016

International Conference on IoT for Smart Living

I’ve just returned from International Conference on IoT for Smart Living, which was held in Colombo, Sri Lanka. It was my first time attending to an IoT conference, and I was very pleasantly impressed on many levels including conference organization and breadth of the talks delivered.

In day to day life as a software engineering practitioners we tend to think in terms of solving the problem and getting results in the best possible way. In that process, we choose the best technologies we could utilize, and the most promising software stack that could help us to simplify the lives of both people and us (developers). The hardest part of this process is not implementing the solution as a product, but give a solution that simplify the problem. This is something really important thing I learned from most of the presentations we had.

The keynote speech of the conference was delivered by Mr. Wasantha Deshapriya, who is the secretary of the Ministry of Telecommunication and Digital Infrastructure. His speech was on “IoT Country Readiness in Digital Sri Lanka”. Throughout his speech, Wasantha highlighted how Sri Lankan is introducing eGovernance to improve the efficiency and impact of public sector, and the need of national level policy (or rather a roadmap) to leverage IoT to accelerate the current work being done.

Morning session had three more speeches other than the keynote speech. Dr. Rishi Bhatnagar, who is the president of Aeris Communication India, Mr. Shivananda Koteshwar (Shivu) who is the Director of Technology at Media Tech India were delivering speeches of the evolution of IoT , Global prospectives of IoT and Smart Cities. Both of them were able to share some insights of IoT in India, where most of the focus is on utilizing IoT to take good care of the senior population, mitigating the problems of urbanization, and climate change. Also, Dr. Rishi highlighted the dimensions of cost, security, right skillset, mindset and cultures readiness, regulations and some technical aspects such as interoperability of devices and long lasting batteries.

Mr. Ajit Ashok Shenvi, who is the Director of Big Data & Analytics at Philips Innovation Campus, Bangalore was also delivered a speech in the morning. He was more focused on IoT in Healthcare. In his speech one thing that I noticed is the new business models that emerging with IoT such as “charge per service, not for the product” (Philips MRI scanners are available as a service in India as I heard).

We had four speakers delivering speeches in the evening. Mr. Pradeep De Almeida, Group CTO of Dialog Axiata delivered an insightful talk on IoT Infrastructure Development in Developing Country Perspective. The speech contained details about protocols that could leverage in developing country’s context of IoT. But the most important part is a use case that he presented complex problem and solved it without any technological infrastructure (I have attached a video below on that). His point was, “If you see a problem, simplify it and solve with minimal complexity. Later move for a more sophisticate solution peacefully”.

Mr. Wellington Perera from Microsoft and Mr. M. I. Deen fro Sri Lanka Telecom also presented in the evening session. The second most important speech of the evening session was delivered by Dr. Chathura De Silva, who is the Head of the Department of Computer Science and Engineering at University of Moratuwa. Where he demonstrated how he has done home automation to wire up things at his home. This presentation was one of exciting one we had.

In sum, IoT haven’t been in a level that would motivate me to attend to a conference, but this participation was refreshing and worth the time I spent there.

Understanding Softmax Regression

This is going to be short... Very short!

Reason being this post is too short is it's Pokemon Go time :)

Logistic regression is where we simply classify a set of values into two known classes. For example predicting whether a grid of pixels intensities represent a “0” or “1” digit. Now let’s consider a situation where we have to predict which digit a grid of pixel intensities are representing. This is exactly a generalization of Logistic regression, where we have to perform a multi-class classification. Hence we could also refer this as “Multinomial Logistic Regression”, which is also referred as the “Softmax Regression”.

Given an input x, our hypothesis will estimate the probability of the class label of the K different possible values and output a K-dimensional vector resulting K estimated probabilities.

Data Flow Graphs

In data flow graphs, computation is programmed as a directed graph. This graph enables the data flow across operations, where each node in the graph represents an operation or a computation. A node may send, receive or can send a response to a message that they received.

In the above data flow graph, the left most node generates the integer values from 1 to 10 and passes them to two successor nodes. One of the successors squares each value it receives and passes the result downstream. The second successor cubes each value it received and passed the result downstream. The right most node received values from both of the middle nodes. As it receives each value, it adds it to a running sum of values. When the application is run to completion, the value of sum will be equal to the sum of the sequence of squares and cubes from 1 to 10.

Given below is a pseudocode that represents above computations. Each function in the pseudocode is equivalent to a node in the graph.