By Ross Fretten, Experience Director at OgilvyOne UK
This year has been a pivotal one for artificial intelligence, which has gone from an abstract and esoteric concept understood primarily through it’s portrayal in sci-fi movies such as 2001: A Space Odyssey and the Terminator series (no wonder it’s typically regarded as malevolent) to the centre-piece of todays zeitgeist as the singularity (the point when artificial intelligence becomes indistinguishable from human) draws near. Hyper personalisation and proactive intelligence have become expectations of software in 2016, and the big players in mobile and digital are striving for this through AI. The result is that we’re tipping into another fundamental paradigm shift akin to the explosion of mobile. Let’s take a look at some of the best of AI in 2016 and what we can expect in 2017 and beyond for industry, creativity and humanity.
Greek mythology talks about mechanical humans, robots and artificial intelligences and it was all the way back in 1951 that the first AI capable of playing chess was shown off courtesy of the university of Manchester. We all experience hands-on with AI every time we play a videogame and you’ve probably seen some of the huge advances in AI coming out of Google and DARPA — such as the iconic “BigDog” — the giant walking dog/cow thing designed to help troops haul large loads on foot and now the more humanoid “Atlas” which you can see here.
But 2016 has been significant and the reasons are three-fold;
- Machine learning is a thing now. We’ve tipped over from the majority of artificial intelligences being tools of systematic task-completion designed and controlled by humans (think AI that can play chess and do nothing else) into AI typically now being self-learning, artificial brains (think IBMs Watson, which can learn anything). This is the beginning of the evolution from artificial intelligence to artificial general intelligence, or AGI.
- We’re seeing a sea-change of willingness to hand over our data and welcome AI into the most intimate facets of our lives by baking AI into our most personal devices — our phones, vehicles and even our homes.
- AI is becoming more accessible. Whilst the big players are aggressively driving AI development and productising that R&D through integrating with great design in mainstream products (see Google Assistant), garage start-ups are hooking into an expanding open source community of AI developers and entrepeneurs.
These three things are fundamentally and radically changing the world and will continue to do so exponentially over the next few years. Every single part of our lives is going to be affected, even those we wouldn’t typically consider to be anything to do with artificial intelligence or even computers. Here’s how.
AI and Industry
Everything will be revolutionised by AI; personal computing, media consumption, health, diet, sex, romance you name it and this also applies to entire industries.
AI and marketing
If ad agencies want to remain the experts in marketing communications and design they’re going to have to adopt AI rapidly. We’ll see these ad agencies partnering with AI suppliers over the coming months as they try to remain relevant but as AI becomes more accessible and prevalent we’ll likely see some of the bigger, more capable agencies investing in developing their own proprietary AI technologies. They’ll likely monetise this R&D by licensing the platforms and providing access to facets of the generated data sets to smaller studios that will be delivering some of the more traditional work such as UI design and build. Blackwood Seven are one of the first agencies to be built around an AI capability, a rather rudimentary capability admittedly, but significant nonetheless.
It’s a matter of necessity really; art directors, copywriters, visual designers, creative directors, social media experts (if that’s a thing) and usability specialists are all going to be superseded by AI within our lifetimes and the same goes for planning which will have to become a more streamlined and human practice, tasked with extrapolating human insight from the data presented to them by an AI, at least until the AI can do the human insights bit too. And it’s not just the human factors that make this a necessity — if ad agencies don’t develop their own proprietary AI technology, their market share will plummet when the big boys in AI inevitably move into communications and design, similar to how some of the big business transformation experts such as Accenture have expanded into creative work of late. We’re already seeing the beginnings of another huge change in advertising spiralling out of mobile and led by AI beginning with how we reach and converse with consumers on their personal devices.
AI and personal computing
Apps and icons on the homescreens of our phones aren’t going to be a thing in the future and we can see this transition beginning now with Siri being able to transfer money and proactively plan our day for us on our lock screens and Google introducing instant apps (which allows users to use apps without installing them on their device). It’s conceivable and probable that in 5–10 years time we won’t have apps as we know them today, we’ll have a single AI personal assistant that exists with us on our phones, in our cars and in our homes that will invisibly be able to access “apps” in the cloud to perform tasks we ask it to and proactively present information to us how, where and when we need it. Think of the iPhone being renamed “Siri” and not having a screen, whilst maintaining all the same functionality.
In reality, Google are likely to beat Apple to this scenario- they’re already way ahead with instant apps as well as their deepmind AI and that’s just the beginning. Telling though, is that Apple — a company so hell bent on hardware first are increasingly talking about services as their major revenue stream of the future and coming out bullish about their work with machine learning and artificial intelligence at every opportunity. Tim Cook’s Apple are aware of a new world order in consumer electronics about to emerge and as things stand Apple just aren’t in contention to be anywhere near the top due to their infuriatingly bad cloud services and lack-lustre AI integration — the two technologies that will dominate the next 10 years along with the internet of things.
Conversational interfaces have exploded onto the scene in the latter half of 2016… and I’m convinced we’ll see a major conversational commerce platform go live within the next 6 months that will define the future of ecommerce.
We’re seeing the practice of user interface design telegraph this shift right now. Conversational interfaces have exploded onto the scene in the latter half of 2016, catalysed by Facebook allowing chatbots into their Messenger platform whilst Google and Amazon try to stake a claim in the home via their always-listening and voice-activated-only home devices, Google Home and Amazon Echo. Both devices are perfectly poised for the imminent battle for market share in the era of AI powered home automation and I’m convinced we’ll see a major conversational commerce initiative go live within the next 6 months that will define the future of ecommerce as a bi-product of this. I suspect the pioneers will be Burberry, who I’ve previously written about in an article discussing how no UI is the new UI and why I think Burberry are going to be at the front of that shift (which you can read here) followed by Amazon bringing it mainstream via their Alexa/echo platform.
AI and Creativity
AI has naturally been cast as a tool to augment our productivity because it’s relatively easy to do as the information is known, quantifiable, and digitally available and manipulatable. A primitive example would be Excel which can perform math on behalf of the human(s) and a more sophisticated example would be mapping software, such as Waze helping Uber drivers route plan considering traffic flows, time of day and reported incidents intelligently in the background.
2016 has produced some great examples of AI working alongside humans as equals in more “creative” ways than simple logistics and voice recognition.
The worlds first artificially creative director.
Contrary to what practitioners in creative fields might want you to believe creativity is a practice of logic just like anything else. There are rules, theories and best-practices for what constitutes “good” creative direction — think of the “rule of thirds” in photography and cinematography. There are even rules behind the abstractions and symbolism that creative minds typically work with within the fertile void and all these rules make creativity quantifiable, which means an AI can recognise, understand and produce creativity. In the case of McCann Japan, their AI Creative Director provides “logic-based creative direction” according to the McCann Japan press release. What this means is that it can process visual information, rationalise it and proactively suggest logical changes that would improve it — most likely in accordance with a number of rules and theories that have been taught to it through repeat exposure. Check out the first TV commercial to come from McCann’s AI CD here.
Anyone for a movie? Written by AI?
It’s this technique of repeat exposure that enables AI to recognise, interpret and learn from patterns. Ars Technica did exactly this with an AI to “teach” it how to write a film script by feeding it dozens of sci-fi scripts, which taught the AI to understand the anatomy of a sci-fi movie right down to the specific speech patterns most frequently observed in the genre. You can watch the 9 minute Sunspring, which is one of my favourite short films ever, here.
Take me to the diamond sky.
Probably the most creepy example is Daddy’s Car — a pop song composed by an AI at Sony Studios after being exposed to 13,000 songs and then asked to produce its own in the style of The Beatles. Daddy’s Car is the disquietingly sexual and implicitly pro-incest result. Before you go hating on the AI, it learnt from us remember.
It's coming for design too... and quickly.
If you think “art” and “design” are distinct from “logical” and “rational thought” you’re very much wrong and artificial intelligence is how we’re going to come to understand how they both sit on a continuum. The Gridmight be one of the first examples of this. The Grid is an AI powered service that provides artificially intelligent web design and it does this by crawling the internet and observing what other sites are doing, recognising patterns and cataloguing those findings in its — sorry “her” (her name is Molly apparently) - weird non-physical brain.
Whilst The Grid isn’t something that can rival human capability when it comes to planning, strategy and execution of those things right now, Molly will have catalogued “more design decisions than stars in our galaxy” by the end of 2016 so it’s easy to see the potential of The Grid and similar AI in 5 years time and what that will mean for designers.
Ever heard of Prisma, the photo-filter app that was downloaded by just about everybody earlier this year? Well those insane filters are rendered in real time (during the triangle loading screen) by an AI that has been exposed to loads and loads of art, to teach it styles which it can then emulate when you apply a filter.
Here’s an image Prima’s AI rendered of my dog, Sailor, in the style of Hokusai’s — The Great Wave Off Kanagawa.
People love to think that robots, or AI are not “human” and therefore cannot be human, and people tend to equate being human versus not with having the capacity to be creative and emotional versus strictly rational. It may be romantic, but it’s a false dichotomy. Much like music is theoretical, logical, contextual and measurable — emotion is an observable, biochemical cocktail. It’s analogous to “natural talent” being art and “genetics” being the rationalisation of that talent.
AI is capable of processing peoples emotions and personality if it’s “brain” (or CPU) is powerful enough.
“Creativity” is often used as a synonym for “gut instinct” or “intuition” and “emotions” are often spoken about coming from the “soul” but ultimately everything humans perceive is data, and intelligence is the processing of that data. With that data, an AI is capable of processing peoples emotions and personality if it’s “brain” (or CPU) is powerful enough and it’s been trained to do so. Facial coding (computers analysing facial muscles to determine mood) has been around for years and we would be ignorant to think that’s the peak of AI’s capability to transact emotionally.
AI and Humanity
The short-term is going to see human capability augmented through AI as we’ve seen in the examples above but this will evolve and manifest through such applications as augmented reality (think Google Glass augmenting our informational lives) and prosthetics (think exoskeletons augmenting our physical lives). Our phones and software will become increasingly more powerful, augmenting our ability to manage our time, our jobs and our relationships in an innocuous blend of human desire and direction with AI data processing on the logistics side.
What if Russia and/or China speeds ahead of the West in the quest to reach and surpass the singularity?
But the conversation will change. The geopolitical ramifactions of artificial intelligence are probably the most significant. What if Russia and/or China speeds ahead of the West in the quest to reach and surpass the singularity? China are leading the way when it comes to bioengineering already, which is a field surely destined to merge with AI. To quote neuroscientist Sam Harris;
Imagine, for instance, that we build a computer that is no more intelligent than the average team of researchers at Stanford or MIT — but, because it functions on a digital timescale, it runs a million times faster than the minds that built it. Set it humming for a week, and it would perform 20,000 years of human-level intellectual work.
Maybe that AI could solve global warming, the dark matter mystery or something equally benign, but what if it belonged to a single nation, rather than say, a global alliance? How would that play out?
Do we diverge and co-exist alongside AI, as ants to humans in an effort to maintain our independence and essence as a species?
The reality is when we get to that point the more pressing issue is our relevance as a species. The question really boils down to how we proceed with our relationship with AI. We can either try to augment one another, which will have to involve nano-technology deployed into our blood coupled with neuro-implants to augment our cognitive abilities for us to retain intellectual relevance, or we diverge and co-exist alongside AI, as ants to humans in an effort to maintain our independence and essence as a species.
Luckily that’s still probably a hundred or more years off, but I’m sure AI historians will look back on 2016 as a major step towards that future.
You can read more from Ross at Medium here.