Data Science – Tech Insight

Hosted at the Goldsmith center Farringdon, AKQA took a look at Data Science with three guest speakers each with a variety backgrounds within the industry, with Dr Rebecca Pope talking about how machine learning can make Healthcare more human, Dr J.Rogel looking at how the discipline interacts with different areas of the business, and Robin Houston showing us his work at KILN where he brings data to life with data visualisations.

Dr Rebecca Pope was very easy to listen to, and I would recommend checking out one of her TED talks. She first began speaking about machine learning giving us an introduction into how how this has quickly developed over the decades showing examples of a rudimentary design that flips burgers to Google’s DeepMind  that built its own algorithm to get teh highest score playing a 1980’s video game.

A topic that was consistent throughout the talk was how media portrays that these machines will take human jobs, but Rebecca talked quickly to squash these, speaking about how our roles will change, but they will change by allowing us to do more interesting things rather than the mundane chores we have to do throughout the working day. This was the main topic of the talk, as we started to look at why this technology has taken traction in most industries with everyone having a Machine Learning tool in their pocket with their mobile phone, but has yet to take full traction within the Healthcare industry. There looked to be a few people within the room that were worried about how this could take jobs away putting further strain on the health care system, But Rebecca argued that Machine learning and AI would be used to sift through data, but you we wouldn’t be able to, or would want to teach them how to be emotive.

There are a few tools in the market that are looking to use the Data that comes through that allows users to treat themselves rather than having to wait to see the local GP such as Drayson []. The argument at the moment is that we have a very reactive healthcare system with issues only being looked at once an illness/ problem comes to the surface. Using data, the algorithms will be able to compare against similar people and find patterns, allowing some users to have preventative treatment whether this is through medication or lifestyle changes.

This led her explaining an example of how current GPs and opthamologist are overrun with very time intensive checks of looking at scans to see if there are any issues, and how people have to be treated on a first come first served basis not allowing for one person that might need treatment sooner than another. Machine Learning would be able go through the scans finding patterns and making sure that the people that needs the treatment can getthis right away reducing waiting lists, as well as freeing doctors time to allow them to do what they are good at, which is giving their human touch in treating people.

Robert Houston came to talk about his experiences of setting up his company KILN with his partner Duncan Clark. With their company they looked at how world is becoming more quantified and the importance of being able to interpret and display the results of the data we get. This is where would come in. Showing us how they have formulated data into amazing graphical pieces that can really get a point across. Showing something similar to what Han’s Rosling made famous.

An example can be seen in the Hans Rosling documentary for the BBC ‘The joy of stats’ which is well worth a watch. Where one graph shows the population size and compares against wealth and life expectancy across the world.

He then went on to showcase some of the work that their team have produced in the last 5 years, from showing the CO2 emissions from commercial shipping to flight paths over the last 100 years including some some gentle bits of humour from Edward Tuffy “The only thing worse than one bar chart is several.