Saturday, February 18th, 2017

Today marks my last day as a teenager. I figure that may be enough of a reason to look back at my childhood, but to be honest, there really isn’t much to discuss. I finished going to school, and then proceeded to begin living on my own. I’ve held a few jobs, but I’ve got enough debt right now to ruin my life, so all I’m trying to do is save up as much money as possible. I think I can save roughly ten thousand per year so that I can afford a down-payment of a house by the time I turn 25. I’m still trying to decide where I should live, but the problem mainly focuses on me wanting to live in the north-west United States, which isn’t too cheap.

If this whole military thing works out, I should be set by 25. If it ends up not happening, I think I can look forward to an interesting bit of time working off my debt. So far the standard day to day progress has been going well. I’ve been researching multiple subjects that I find interesting, and the countless array of websites that offer help and knowledge is staggering. One topic that I find myself constantly going back to is that of machine learning.

Machine learning is a type of artificial intelligence (AI) that gives computers the ability to learn without being specifically programmed to do certain tasks. It primarily focuses on the development of computer programs that change and improve when exposed to new data, like that akin to data mining. Think of it like a computer that can read and grow in intelligence in the same way a human could. The work that has been done in this field is often fascinating and is rapidly advancing. I may be a novice to the subject, but I’m attempting to improve my understanding so that I could become someone who could report advancements to the field. I really think this is going to seriously change how we interact with machines in the next few decades.

My research started by visiting OpenAI. Founded in 2015 (with the help of everyone’s favorite eccentric billionaire, Elon Musk), this company’s main goal is to help further the advancement of artificial intelligence without the threat of stockholder intervention. They aim to offer progression by setting open standards in the hope of developing an AI that won’t threaten humanity. That fear is held by many, who imagine a Terminator-like scenario, including several key leaders in the discussion and advancement of AI like Stuart Russel and Steven Hawking. After looking more into that fear for a few hours (including a few humorous videos and posts discussing that topic), I stumbled upon a rather helpful resource.

Looking into a GitHub repository, I found an addition made by Zack Chase Lipton, who is a grad student working with the AI group at University of California, San Diego. This led to me finding a useful blog entitled ApproximatelyCorrect, which helps break down various issues entrenched within the deep learning subject. While a few of the structures are a bit too advanced for me to follow (seriously, try tackling this presentation), the gist is quite standard and easily understood. I’m glad that I was able to find a resource that can give me the breakdown for the future of high-level computer interaction.

Finally I managed to access a digital textbook written by three leaders within this field – Ian Goodfellow, Yoshua Bengio, and Aaron Courville. You can visit it here or purchase a physical copy on Amazon here. After working my way through a few chapters, my interest is only becoming more enthusiastic. I figure more study will allow me to become adept with the advancing trends, so I look forward to additional research.