It’s the end of the second week of 2022 and things are heating up. Getting back into the groove of things always takes some time and this week I’ve been trying to juggle multiple projects and establish a new routine.
With so much happening I’m going to start writing about what I’ve learnt each week. Courtesy of Tiago Forte and his ‘Building a Second Brain’ concept, I regularly write and collect notes on things I’m learning, ideas, and articles that inspire me. Here I shall review them and distil them for my own sake (but feel free to pry). Perhaps you’ll learn something or perhaps laugh at my previous ignorance. Either way, I imagine it will be good fun.
Coding and NLP
This week, as I continue on my coding journey, I have learnt the fundamentals of NLP (Natural Language Processing). This may mean nothing to you, but I guarantee you use NLP technology every day. As a field, NLP analyses human language and uses artificial intelligence and computer science in order to build sophisticated communication technology. You know when spellcheck saves your life? Or you ask Siri to tell you a joke? Or you’re reading the news and three other similar articles pop up as recommendations? All of these technologies can analyse the style, emotion, and topics discussed in your language. This makes our lives easier when we communicate with others (excluding when your phone corrects “crab dip” to “crap dip”). For more examples of the capabilities of NLP, check out Neil Patel’s article.
Much of what I learnt this week involves a variety of methodologies and NLP models for analysing language. Part of the importance of good models is the ability to understand the meaning of a sentence in the context of the document. For example, “I saw a cow under a tree with binoculars” is pretty easily interpreted by humans. But an NLP model could interpret this in a wide variety of ways. Does the cow, tree or I have the binoculars?
To combat these types of issues, models can break paragraphs into smaller units (sentences, phrases or individual words) and look at the frequency of those units. Grouping words into phrases helps models more accurately understand the context. When a model has been given enough training text (samples of text which it can run through in order to gauge the context), you can gather data on what the general consensus is and potentially generate text which mimics the tone and topic. The applications for this technology are endless and I’m fascinated by the variety of models used to influence how we communicate with each other. More on this to come as I continue learning…
Conservation and the environment
Bushfires have always played a big role in the cycle of Australia’s natural environment. Whilst fire is important for seed dispersal and vegetation regeneration, animal death is always the a major concern. In the 2019-2020 bushfires, it was estimated that 1 billion animals were killed due to fire or the dehydration and starvation following. It’s an astonishing number which breaks my heart.
But this week I came across an article, which investigated the effects of fire on animals in more detail. Using tracking collars, and data from field studies worldwide, they discovered that ⅔ of tracked animals survived the fire. Whilst this is a small sample of the total animals impacted by fire, it’s a promising outcome which indicates that many animals are more resilient than we think. Researchers suspect many animals, particularly in Australia, have some sneaky evolutionary adaptations.
However, there’s no denying that the aftermath of bushfire is not an easy time. With fresh vegetation turned to cinders and precious water sources gone, animals need all the support they can get in the weeks following a fire. With climate-exacerbated bushfires occurring more frequently, and the threat of megafires every Australian summer, the role that conservationists play in preparation and post-fire care is essential.
Problem solving and creativity
Despite maths and coding being considered quite ‘right-angled’ and theoretical, there is enormous demand for creativity and problem-solving in these fields. This week I came across the ‘Zwicky box’ as a method for thinking outside the square. Turns out Fritz Zwicky was a prolific 20th century scientist who, amongst other things, coined the term ‘supernova’. He is recounted as being capable of producing unique solutions to problems and used the Zwicky Box as a way to break down and categorise the components of a problem. Below is an example of a Zwicky box used for building an app (courtesy of Ness Labs):
By breaking a problem down into categories, we can think of unique examples for each category and jumble up the examples to create brand new combinations. People use Zwicky boxes for everyday decisions, content creation, brainstorming, and innovation. The applications are endless.
To read more on how to flesh out a Zwicky Box, check out the original article I read. I’m going to try it out over the next few weeks and see what interesting combinations of ideas I can create.