What we really think of Artificial Intelligence

by Round table discussion
07 Nov 2017
9 min read
Share this article:
What we really think of Artificial Intelligence

Artificial Intelligence (AI), robots, chatbots, the rise of the machines – this is one of the hottest topics of 2017 but, what does it all really mean, and what changes should we expect?

We wanted to get down to answering some of those curly questions so we gathered some experts from across the agency and asked them what they think about AI and where it’s heading. We heard from:

From left to right; Taryn, Greg, Jo, Edwin

Art Director, Taryn Sharman,

Taryn’s been on the creative team for nearly five years and from NBN to UNICEF she meticulously brings life to ideas in beautiful designs. When not snuggled up with her cats, Taryn is generally training for some sort of expedition or triathlon.

Digital strategist, Greg Calacouris

Greg has been with agency for just under a year and works across a range of clients to determine their strategic digital direction. When he’s not reading a trend report, Greg can be found getting through his New Year’s resolution of watching a film a day whilst no doubt teaching Google Home to do something fancy.

UX Designer, Jo Li

Jo gets into the mind of users whether they are using a chatbot or navigating an app, she’s usually found creating an intricate wireframe. When not penning up a map of the matrix, Jo can be found trying to keep up with her gorgeous greyhound who is almost the same size as her.

Senior developer / Business Analyst, Edwin Yosorahardjo

No one has known Adrenalin without Edwin, he’s like the grandfather of the agency having taken residency here for the past decade. Edwin is master of all things tech and when he’s not coding with a VR set on, he is generally educating the office on Indonesian culinary delights.

So with this crack team assembled, we decided it best to start at the beginning and find out what AI means to each of them. Using Einstein’s old proverb as the springboard, we asked each of them to describe how they would explain AI to a six year old.

 if you can’t explain it to a six-year-old, you don’t understand it yourself  ~Albert Einstein

Greg started by describing AI as “the way a computer can read data and make a decision or an output using that data.”

Jo explained it as “AI is teaching a computer in the same way you would teach a child something. So, this is A and this is B and this is why they are different. When they see a new object, they test it against A and B and see which one fits better. Being able to pattern match at a basic level.”

Edwin simplified the concept to “turning a computer into something like a human that can think like a human”

Lastly, Taryn a likened AI to “a robot where you can interact and communicate at a very basic level. Ask it to do stuff and get answers.”

The OXFORD dictionary defines the term as:

So once we were clear on the topic, we got down to the nitty gritty. As with most technological advancements, AI is often referred to in films, books and the media as a scary unknown that threatens our very existence. But getting past this doom and gloom outlook, we wanted to take a look at what our team find most interesting about what AI means for the future.

First, Edwin began by drilling down further into the definition of AI. He said “[the term AI] is quite blurry now as there are a lot of different areas of AI, one of them is deep learning, where AI can learn by itself. AI has been around for a while and is used in computer games where the enemy can find a way to attack the player. AI can start learning and then have curiosity logic so it can decide it would like to try something. But previously it was built in a method of brute force, for example ‘is this correct or wrong?’ and then it tried to find the best answer but it has evolved now so it can be curious.”

Following this idea of independent learning, Greg said what he finds most interesting about AI, and technology more broadly, is the unintended consequences. He went on to provide an example using Google image search:

“There are situations where you see it happening now. For example, if you Google image search “professional haircut” you get images of predominantly white people but if you search “unprofessional haircut” you get images of black people.

There’s an inherent racism in the Google algorithm that is likely an unexpected outcome. We could soon have AI do everything, such as drive a car, and there maybe unintended outcomes and we don’t know what they might be. It’s interesting and scary.”

Jo agreed that it’s because humans create AI that not only is it learning the things we want it to learn, but it is also learning our prejudices as well. She used the example of the board game Go and how when they taught a robot how to play it, they pushed it to play against professional players which is where it learnt from. She also gave another example of negative unintended consequencesMicrosoft did Tay as a Twitter account and it started learning from users and started spouting racial abuse and they had to shut it down.”

Everyone agrees it is a bad reflection of what people can do - Tay started off as a 16-year-old innocent teen and people kept Tweeting horrible things at it so it turned into a “Hitler loving racist”.

It didn’t take long before we were discussing the prospect of a dystopian future where robots take over. Taryn, suggested that “At the back of everyone’s mind with AI and machine learning there is probably the dissention topic where AI is going to completely outsmart human beings. People who first went up into space, first went up with less power than we have in our mobile phones now and the fact that we keep evolving. It is going to get to a point where we are the stupid beings on this planet. How do we control that, what does it mean for mankind?”

Jo added that this also raises the issue of responsibility “every turn of the century there is a new era where people are losing their jobs to machines and at what point do we stop replacing people with machines?”

However, Taryn countered by suggesting that that has always been the case throughout history but the current education system is in place to set up for success “Looking at the current education system compared to what we learnt in the UK 15 years ago, which was Microsoft Word, Excel, how to mail merge and things like that…a few years on they were teaching Flash and people were starting to learn how to code. It’s good that schools are now starting to teach children how to code from primary school age. So I think although people will lose jobs, it’s good that we are anticipating that to a level.

We are the first generation that’s actually moving with the times.

Previous generations had to be reactive but for upcoming generations, technology won’t be a hurdle. People will be empowered to create jobs in different ways.”

Greg agreed with Taryn, he suggested that over time blue-collar jobs have been increasingly replaced with‚ white collar‘ jobs which changed our world. “With AI many ‘thinking’ jobs will be replaced and long term we’ll probably see another big change in our world – hopefully for the better. Short term though we will see job loss and people will need to learn a new skill.”

The team agreed that although there are many ways AI is set to change our society, culture and our workplaces, they reached a general consensus that we are already adapting to that new future. As with any technology, the power and magic is in the creative application of the technology.

Here at Adrenalin we are constantly testing and pushing the boundaries of technology. We have an innovation lab, Adrenalin X, where we run rapid prototyping projects with the latest tech. Recently we have been using AI to build intelligent chatbots. Read more about our recent projects here.

If you would like to find out more about the future of AI, we are holding an event on 15th November “Rise of the Machines: What AI means for Brands” where we will host speakers from UNSW, UpWire and Gumtree to discuss their perspectives and insights on AI.

Spaces are limited, reserve your ticket here.

Share this article: