An Introduction

Who am I
I am a software developer in London, I have been working in industry for almost 2 years after having graduated from a Masters course in Artificial Intelligence.
I am eager about Artificial Intelligence particularly the use of knowledge graphs to create semantic representations inside the machine.
The direction of my interest in Artificial Intelligence is focused on being able to be more expressive in the code that I write. More than building self driving cars, it's building tools which allow better interpretation of the authors desire for a program.
In my day to day I write Python code in vim.


Why I am writing a blog
As a dyslexic writing has never come naturally to me, however, as I go along I increasingly realise not just how important the skill of being able to convey ideas on paper is but that skills I used to think of as intractable for myself to achieve may actually, with a different approach to that used at school be within my grasp.
My most recent experience of this is my attempts at learning mandarin using Dualingo. Although the progress is slow, it is very much there. Not something I could have said of my days learning French and German in school.

I have also been persuaded by the number of sources I have read that talk about the benefits particularly for developers in writing a blog, in terms of developing your ideas and questioning your assumptions. These benefits are often talked about being present absent of a readership.



What I plan to write about
I have some ideas of future blog posts already. I hope to publish one each Sunday night.
I hope to be able to focus my writing on software development perhaps including some posts which are at best tangentially related to engineering.
So far I have some ideas for posts, including one on some models that I think have helped me explain the world to my engineering brain (Perhaps starting with a digression from software isn't the best idea).
Some classic blogger posts such as what I wish I knew when I left University about software engineering in industry and VIM tricks I find useful.
The other angle which I think might be more useful for me to write about, and might be more useful to a readership is about projects I either hope to work on or have been working on.
I think having a medium in which to communicate and really form these ideas would be most useful. The collecting of resources that I find, and problems I solve might also be useful for other people trying to work on similar things.
I might revisit some older projects which I have abandoned and talk about what I think could have been improved as well as what went wrong. One of my largest hobby projects was Avinders and I certainly had a lot more ideas of where to take this game than I ever implemented.
As part of my interest in knowledge graphs, I'm interested in if you could use some simple natural language techniques along with simple knowledge graph structures to produce more proficient chat bots. I find it interesting that the majority of chatbots, at least that I have come across, either use a large corps, perhaps with some substitution, or recurrent neural nets to product the output. I am intrigued as to whether a simple grammar parser could produce a reasonable chatbot, albeit with the proviso that the users input proper spelling and grammar. Things that might not make the chat bot all to successful in the wild but might provide interesting insight into the pitfalls of such a technique.