A bit of background on how Laura got to like AI so much.
Someone once expressed to me they saw AI as the “intersection of everything you do”. Having giving this some thought since, I do believe this to be true (sort-of). The “things I do” vaguely come under four categories— emerging technology, coding, data, people. These are wide buckets of things but if you were to Venn-diagram it up, AI would definitely touches all of these and therefore would be the intersection (more likely it would be one of the intersections).
It all started with data. If you know me well, you know I am a bit obsessed with data. I’ve been told I say data more than anyone else by a number of people, and no matter what I do for a job (or a hobby, to be honest) I always end up back to data.
My first proper exposure to data was during my Placement Year at a Big 4 company, where I got thrown in headfirst and nearly drowned in some data (I meant that as a joke, but it’s painfully true). My Placement Year was amazing and awful in equal measures, but it was honestly the making of me and I highly doubt I would be I am today without the skills and work ethic I learnt during it. My year mainly involved working on some high profile projects, including the financials for Lehman Brothers’ collapse, but I spent the majority of the year working on the data for the News International Phone Hacking court case. Very, very cool — doing work and seeing it in the news — very, very long hours and an extreme learning curve. I had never heard the term SQL before and I did not know there was magic behind Excel called VBA, but I got stuck in and picked it up. I worked on the entire data life cycle — including getting the data from printed PDF phone records into digital format, cleaning the data, analysing it, everything. One of my favourite parts was working on a “mis-spelling” analysis carried out on the financial records (this could explain my interest in chatbots and Natural Language Processing now).
After I finished off my degree, my plan was to come back to the Big 4 company, but I realised that lifestyle wasn’t for me and I wanted something different. It turned out what I wanted was a different type of data. Apparently my soul longed for data that could do more than just prove or disprove whether a crime was committed. What this meant was data I could genuinely get to know and get lost in. It sounds ridiculous, but even now with the qualitative interview data for my PhD (which I was stand-offish about at the beginning, before I understood it’s value) I know it inside out and I keep finding more beauty in it. Back when I was a fresh graduate, I bounced around jobs for a year as I just could not get it right. Analysing market research surveys in SPSS had potential but I missed coding, building complex models from BIG DATA in advertising had its draw but it lacked meaning… I was just a bit lost.
Surprising to me, I found my place in a massive corporate company’s Data team. The company was a global company which processed something like 40% of the UK’s card payments and operated in over 200 countries, so lots of data. Data about people, data with meaning. Data I could pull interesting trends and statistics from, data I could make pretty visualisations from. Basic analytics soon got boring — I loved automating processes and creating models which they reluctantly let me put towards an ambitious trend analysis / anomaly detection project (using SQL, R, Tableau and Salesforce). This project involved the customer facing Relationship Managers getting their customers’ data everyday, and meant I had to understand their needs and train them in how to use the end product. I loved this part of it, it was the missing people part of the puzzle.
Another project I worked on was tracking Apple Pay and Android Pay when they went live in the UK. I created a dashboard to send to important stakeholders at the company showing the trends during the launch. One of these stakeholders was the head of the “Technology Innovation” team. They were a team who looked at new technology and how it could be used or impact the Payments Industry. When Apple Pay came out they went on a “field trip” to the pub to test it out… Now, I could not believe this was a real job. I got to know the team (through a shared interest in tech and karaoke), eventually an opening in their team came up and they asked me to apply. I got the job (much to my surprise, as my imposter syndrome made me cancel the interview twice).
I spent 4 years building cool proof of concepts, writing whitepapers, working with start ups and loving the tech evangelical side of my work (talks, hackathons, conferences). I coded in everything from Java (cries) to Unity to Scratch, and everything inbetween. I was the team’s resident ethics and privacy advocate. They’d come up with cool ideas and turn to me to check how creepy or unethical it was (sometimes I despaired). Eventually data crept back in — product managing the launch of an awesome app was surprisingly data intensive. I worked on a project to protect small merchants by predicting when they might be in trouble (inspired by the challenger banks and the telecomms industry). I had data, coding, education, people and, now another puzzle piece, emerging technology.
I was in my element, but longed for something more meaningful than “payments”. Two project which really captured me were building a mock up of how the Industrial Internet of Things could work with Raspberry Pis and writing a paper on what happens when robots can make payments legally. These caused a small obsession with robots and AI.
This small obsession caused a random application to a PhD programme 100 miles away which I surprised myself by getting into (only because I forgot to cancel the interview so did it via Skype out of politeness — imposter syndrome is tons of fun). My Masters project, by choice, involved lots of data — on designing tiny, tiny particles to deliver chemotherapy drugs to cancer cells. Data which I cleaned, wrangled, sorted, analysed, visualised and built neural networks from. I enjoyed it, cleaning messy data is a weird process which brings me much joy, and I got to make some delightful graphs. The Machine Learning part was a great opportunity to flex my coding skills. It wasn’t quite right though. The data and the work itself were missing the people part of my Venn diagram (I got to play with tiny robots from time to time which kept the emerging tech part satisfied).
After some more bumps and doubts (and starting a social enterprise, and being made redundant from my Tech Innovation job...), I finally feel like I’ve found “my thing”. My current PhD project is understanding how education can be used to help people not be left behind as Robotics and AI advance. Somehow combining education and emerging technology (namely AI), in terms of how these things work (or don’t) for people and society, (which involves data and coding in its own way) seems to make my Venn diagram feel complete (but hard to visualise, see my attempt below).
I wanted to start this blog as a way to explore my interests not currently covered by my PhD work (privacy, ethics, coding, data, emerging tech like IoT). The rest of the posts WON’T be about me, they will be much more interesting.