Information Technology | Artificial Intelligence » The Artificial Intelligence Issue

Source: http://www.doksinet MAY 2016 THE ARTIFICIAL INTELLIGENCE ISSUE PA RT N E RS A N D R EWS A L D R I D G E : A R E P O RT O N T H E C H A L L E N G ES A N D O P P O RT U N I T I ES O F A I Source: http://www.doksinet 04 06 08 CO N T E N TS 14 18 22 THE EXPERT VIEW UP FRONT W e’ve all seen the films, read the books, maybe even had the dreams. Once-helpful robots gaining consciousness and ultimately attempting to overthrow the humans who created them is a haunting vision that is well represented in novels and movies. It happened in Terminator, it happened in iRobot, it happened in Ex Machina. Maybe it’s their fictional presentation, or maybe it’s the very real world predictions that AIs could one day do our jobs better than we can – but AI gets a bit of a bad press. Yet is this natural scepticism and apprehension fair? And beyond that, is it actively inhibiting the future? We thought it was about time to find out. After all, the origins of Artificial

Intelligence go right back to the 1950s, when Alan Turing published his infamous paper ‘Computing Machinery and Intelligence’. With the Turing Test, the idea that a machine could become genuinely ‘intelligent’ was born. And since then, the human race has been obsessed. Questions of humanity, creativity, independency, sensitivity and emotion all arise naturally from the AI debate and can make us question the very essence of what it is to be ‘human’. But what is plausible, what is likely and what possibilities, if any, should be keeping us up at night? We want to separate the fact from the fiction and get to the crux of the AI debate. We want to know if there is a place for AI in our everyday lives and whether that means there’s a space for them in our brand experience. And fundamentally, will their role complement or eventually overtake our own? And how do we go about mitigating any risks around this? This report brings you analysis, theories, the latest tech and even our

own experiments to get to the very heart of AI. So, whether you’re man or machine, sit back and read on. Alastair Cole, Chief Innovation Officer 02 PARTNERS ANDREWS ALDRIDGE | THE AI ISSUE CO N T R I BU TO RS TO T H I S I SSU E WHY SHOULD I CARE ABOUT AI? NEW, NOW AND NEXT WILL AI BE THE DEATH OF EMOTION? CAN ARTIFICIAL INTELLIGENCE BECOME TRULY CREATIVE? GETTING UP CLOSE AND PERSONAL WITH AI and their AI predictions Alastair Cole Chief Innovation Officer Forget robot armies taking over the world, AI is going to transform what brands – and humanity – can do. Szilvia Szabo Product Manager In the face of AI, the worth of human work and effort will increase and a ‘made by humans’ label will become a mark of value. Alexa Turnpenney Content Editor Opportunities and horror stories will still to go hand in hand, and AI will continue to be one of the most controversial topics of the century. Craig Lawrie Strategy Director The lines between robots and humans will continue

to blur – but whilst robots may become more human, ‘emotion’ will ultimately stop us becoming more robot. Gemma Ghelardi Strategist Robots are already creative – but they don’t yet own the creative process. That is, arguably, just a matter of time. Matt Williams Head of Creative Content Can’t this prediction be auto-generated by a robot? PARTNERS ANDREWS ALDRIDGE | THE AI ISSUE 03 Source: http://www.doksinet A year spent in artificial intelligence is enough to make one believe in God. Alan Perlis, Computer Scientist The power of AI technology is it can solve problems that scale to the whole planet. Mike Schroepfer, Chief Technology Officer, Facebook THE EXPERT VIEW I’m increasingly inclined to think that there should be some regulatory oversight, maybe at the national and international level, just to make sure that we don’t do something very foolish. I mean with artificial intelligence we’re summoning the demon. Elon Musk The development of full artificial

intelligence could spell the end of the human race It would take off on its own, and redesign itself at an ever increasing rate. Humans, who are limited by slow biological evolution, couldn’t compete, and would be superseded. Stephen Hawking 04 PARTNERS ANDREWS ALDRIDGE | THE AI ISSUE PARTNERS ANDREWS ALDRIDGE | THE AI ISSUE 05 Source: http://www.doksinet Why should I care about AI? Nowadays, the 6 o’clock news is full of self-driving cars, smart home devices and a mononym almost as well known as Madonna – Siri. By Alexa Turnpenney, Content Editor F or the past 50 years, Artificial Intelligence has mostly been the realm of science fiction writers and Hollywood producers. Although it was fictional, the dialogue has been intriguing and terrifying nonetheless. With it came whispers about how this fiction could infiltrate our everyday lives. Were huge companies, R&D departments and prestigious universities working on big machines? Could those machines eventually run

those organisations? Could they take over the world? Always an interesting train of thought, but in actuality, this had never really come all that close to the average consumer. Until recently. Nowadays, the 6 o’clock news is full of selfdriving cars, smart home devices and a mononym almost as well known as Madonna - Siri. So what has changed in the past one and a half years to take AI from fantasy and into our everyday lives? In a nutshell, tech compartments have gotten smaller and smaller, and crucially, cheaper and cheaper. Now components are so cheap and tiny that they can fit into most objects you can think of – and with that tiny compartment, an everyday object is transformed into a smart device. This change goes hand in hand with the increase of internet availability and speed, which contributes to big data, enables machine learning and consequently AI. Large companies like Google and Facebook soon realised that AI could be used to solve smaller, real world problems (and

create new business opportunities). And since Silicon Valley’s copying machine is working fast, there are now a large number of start-ups working on all sorts of machine learning and AI applications in the vain hope of data collection. Because, let’s face it, data is the new currency. And as its presence in pop culture can attest to, AI is not just for engineers and sci-fi enthusiasts. Hundreds of thousands of people are satiating their curiosity by signing up for AI courses at top universities (like Stanford and the University of Washington), as well as to online courses. So, now that we are at the point where technology is finally catching up with theory, it’s time to start taking an interest in AI. Whether we like it or not, the future is now – and embracing AI could make all the difference to the development of that future. How our machines got smart – the history of Artificial Intelligence Are machines now intelligent? Turing Test Alan Turing, father of AI, introduces

the Turing Test, defining a standard for a machine to be called intelligent. 1943 WW2 triggers fresh thinking 1950 I Robot Isaac Asimov publishes a collection of thoughtWW2 brings together provoking and visionary scientists from a range short stories. Most notably of disciplines such he created the Three Laws of as neuroscience and Robotics, designed to stop computing. our creations turning on us. | THE AI ISSUE 06 PARTNERS ANDREWS ALDRIDGE Shakey the Robot 1969 Honda’s robot walks as fast as a human while delivering trays to customers in a restaurant. Sojourner A robot that combines locomotion, perception and problem solving is born. 1963 64 years after Turing published his test that would prove machine intelligence, chatbot Eugene Goostman finally passes. Asimo NASA deploys Sojourner, the first robot on Mars. 1980 1997 1998 2000 2011 Rancho Arm Alacrity 2.0 Furby Man vs machine Stanford University acquires Rancho Arm, a computercontrolled tool for the

handicapped with flexibility similar to a human arm. Alacrity 2.0 is released, the first commercial strategic advisory system based on the evolution of markets. Tiger electronics release the first successful attempt at AI for consumer audiences. IBM’s Watson took on the human brain on US quiz show Jeopardy. Watson trounced the shows best ever performers with ease. 2014 Rise of MEMS Ubiquitous computing is soon arriving with a shift towards microelectromechanical systems (MEMS). All daily objects will use unseen computers to collect data and improve their usage. 2020s AI personal assistants AI PAs like SIRI and Cortana will continue to develop by understanding colloquialisms, being able to hold conversations and eventually allowing us to proxy selves. | THE AI PARTNERS ANDREWS ALDRIDGEcreate ISSUE 07 Source: http://www.doksinet NEW, NOW AND NEXT By Szilvia Szabo, Product Manager We examine some of the biggest developments and stories in AI to give you the inside track on

where technology might be going next – and crucially, what it means for you. Cyber security Jobs Online financial service providers have a target on their backs when it comes to cyber attacks. To combat this, companies like PayPal are teaching your computer to be a detective. Allegedly, 5 million jobs will be taken over by AI by 2020. Experts are calling this the dawn of the 4th Industrial Revolution. Why it matters With machine learning and the application of AI, PayPal decreased the number of fraud cases below the industry norm. Their AI can tell you whether a foreign transaction on a US account is fraud or simply a travel purchase without ever making a single mistake – and all in real time, of course. According to the latest report by the World Economic Forum, white-collar office and administrative roles are expected to suffer the greatest job losses at the expense of AI. More specialised roles, such as computing, mathematics, architecture and engineering will also see an

increase in the number of ‘robots’ performing traditionally human roles. What’s next What’s next By sharing data repositories on cyberattacks and fraud, companies can benefit from an increasingly secure community. Australia, China, France, Germany, India, Italy, Japan, Britain and the United States have to be prepared to build a workforce with futureproof skills in order to cope with ever-growing inequality. We also need to examine ways to address the gender divide within the workforce to address disproportionate effects on men and women. Why it matters 08 PARTNERS ANDREWS ALDRIDGE | THE AI ISSUE Medical science Machine learning The brilliant Dr.House is back In the form of an AI Doctor called Doctor Watson. Chinese ‘Go’ has been long hailed as the pinnacle of human gameplay. Or it was, until Google pioneered AlphaGo, a machine who took on a human champion at the game – and won. Why it matters AI’s ability to sort through information and remember everything

it has ever seen is enourmously powerful. Computer-aided diagnosis can analyse way more factors than a human doctor. In an instant, it can review all of a patient’s history and weigh risk factors such as age, previous diseases and residence, to come up with a shortlist of possible diagnoses. It can even grade patients with a percentage rating that it’s disease X or syndrome Y. Why it matters Machines are much slower learners than humans. So to develop its own understanding of the game, AlphaGo had to play millions of matches. But that meant that by the time it played against a person, it had actually played more games of Go than any human possibly could within their lifetime. Making victory a doddle What’s next What’s next Robots won’t steal doctors’ jobs, but they will help overworked medical teams to avoid some of the dangerous fatigue that can lead to mistakes. Keep an eye on Alpha Go’s next challenge: playing Lee Se-dol, the world champion, in the ultimate man vs

machine grudge match. Automotive Driverless cars are big news, with companies like Google actively promoting their pioneering work in the field. But there are some fierce opposers to these new innovations. Why it matters BMW and Mercedes-Benz are among the high-end car makers looking to make their vehicles driverless – but they seem to be the exceptions. Porsche boss, Oliver Blume, said he has no plans to take away the pleasure of driving a Porsche from anyone and that ‘an iPhone belongs in your pocket, not on the road.’ However Porsche aren’t averse to innovation and plan to offer hybrid versions of all its models in the future, to lower emissions across its fleet. What’s next Driverless cars may not be the nextbest thing in the luxury world, but we’ll definitely be seeing them in the mainstream market. Uber says its fleet is likely to be driverless by 2030, Elon Musk expects the first vehicle to be approved in 2021, while Audi says its vehicles will be capable of fully

autonomous driving by 2017. Connected Home Your fridge may just be the biggest threat to national security. Why it matters James Clapper, Director of National Intelligence USA, talked about the danger of household connected devices for the first time at CES earlier this year. His biggest worry is that they can be used as surveillence tools. “In the future, intelligence services might use the [Internet of Things] for identification, surveillance, monitoring, location tracking, and targeting for recruitment, or to gain access to networks or user credentials,” What’s next The main problem is not the exsitence of the IoT, but the fact that the majority of these devices are not secure. Companies will have to make security a priority at the start of the manufacturing process. PARTNERS ANDREWS ALDRIDGE | THE AI ISSUE 09 Source: http://www.doksinet What does AI mean for your business? By Will Sharples, Account Executive Businesses from all sectors are embracing Artificial

Intelligence. Processing power is cheaper than it’s ever been, so businesses can use AI knowing the costs of computing are affordable. The coinciding emergence of the Cloud allows businesses to access storage and computing power on demand, utilising a pay-as-you-go cost model. MOORE’S LAW PROCESSING POWER FOR COMPUTERS WILL DOUBLE EVERY TWO YEARS AT THE SAME COST. AI is creating opportunities 80% Many types of enterprises are utilising the power of AI to help analyse data and use it to make effective business decisions. A new survey from Narrative Science reveals how top enterprise leaders perceive and use AI in their business. OF ENTERPRISE EXECS SAY AI MAKES WORKERS MORE PRODUCTIVE AND CREATES JOBS ENTERPRISE LEADERS DEFINE AI AS TECH THAT Thinks and acts like humans – 31% 2000 2005 2010 29% Learns to do things better over time – 25% 2015 Can understand language – 7% Can answer questions – 4% All industries benefit from AI OF RESPONDENTS NAME PREDICTIVE

ANALYTICS AS THE TOP PRIORITY THEIR ORGANISATION COULD ADDRESS WITH AI-POWERED SOLUTIONS Passes the Turing Test – 1.5% All of the above – 28.5% AI is now part of our everyday professional and personal lives. Its use has infiltrated all aspects of the business world and has a part to play in all brand stories on some level. Other – 3% How companies are utilising AI Most widely used Artificial Intelligence solutions by enterprise The main reasons why enterprises currently use AI Access to data business can use to make effective decisions HEALTHCARE CONSUMER TECH TRANSPORT Zotec Partners leverages natural language generation (NLG) to automate written, datadriven narratives used to make decisions and improve customer care Companies like Facebook, Google and Twitter invest millions in AI to better understand Big Data to better connect users with people, info and ads Google builds self-driving cars, which have so far clocked-up over 1 million miles and are currently out on

the streets of California and Texas Access to data consumers can use to make effective decisions Automation that eliminates manual and repetitive tasks Monitoring and alerts about the health of the business Voice recognition and response solutions Automated datadriven reporting Machine-learning Virtual personal assistants All of the above System used for decision support 10 TOYS MARKETING AGRICULTURE Hello Barbie doll, Wi-Fi enabled, can hold conversations with children in any language Marketers use data analytics to crunch the value of each new customer, factoring in hundreds of variables Combine harvesters are driven by robots, with data gathered on crop yield used to determine fertiliser usage PARTNERS ANDREWS ALDRIDGE | THE AI ISSUE Automated written reporting Robotics All of the above Other 0 10 20 30 40 PARTNERS ANDREWS ALDRIDGE | THE AI ISSUE 50 11 Source: http://www.doksinet AI: FACT VS FICTION Computers will have no place in our lives Fiction In 1943,

IBM’s president Thomas Watson predicted that there would only be a world market for five computers. Reality In 2015, there are approximately 2 billion PCs in use worldwide. Whilst this market has seen a slight decline of 9.5% due to the increase in smartphones and tablet sales, the number is still impressive. Food for thought Although fewer people are buying PCs nowadays, they still remain the number one tool for ‘getting things done’. Machines are taking over the world Fiction In the 2001 movie Space Odyssey, HAL was designed to control the Discovery One spacecraft. However HAL starts its own mission - and that mission doesn’t involve humans. Knowing me, knowing you: the death of brand experience? Reality In reality the closest we have come to a supercomputer is Watson, which ‘understands’ natural language and is able to read millions of books. Watson is currently working alongside humans, helping in the research departments of large corporations like P&G and Coca

Cola. Food for thought By Alexa Turnpenney, Content Editor I t’s Thursday afternoon and you’ve been eating healthily all week, so you’re in the newsagents picking out a chocolate bar. A pretty simple task, yes? Or is it? Are you really making the best choice? Are you indeed even the best qualified to make that choice? Or could a machine do it better? In essence, can a machine actually know what you want, better than you do? Think about it – a machine can track ever y thing about your chocolate bar preferences. It can access your smartwatch or fitness tracker, your calendar, your emails, your social media, even your spending. It has a bet ter idea of the composition, flavour, and nutritional value of each chocolate bar. And it can overlay that information onto your sugar levels, your budget, your mood. So it could, in theory, 12 PARTNERS ANDREWS ALDRIDGE | THE AI ISSUE predict which chocolate bar (out of the hundreds out there) you would most like to eat on a Thursday

afternoon. That might work sometimes, but would you trust an AI to really know these things? Or are there some factors that simply cannot be quantified? I f t h e a n s we r i s ye s , t h e n b r a n d experience is safe. At least for now. One day though, we could be faced with an AI so smart that it knows your preferences better than you do. Of course, that’s dependent on a circle of trust developing between us and our AIs. Because the more we trust them, the more we’ll tell them – and the more we tell them, the better their choices will be. And arguably, the better our experiences will be off the back of that. But that’s something to worry about in the future. In the meantime, brands can learn a little something from AIs. By hypertailoring services to customers’ needs, brands can reach people in ways that are genuinely transformative. Using algorithms to convert data into useful analysis, insight and output means that we can work more accurately and usefully. A union

between people and machines in pursuit of the perfect brand experience. So the next time brands think about a campaign, they should ask how AI could help. Is there a way that scheduled social posting can ease a team’s workload? Can we use AI to customise tone of voice to speak directly to the individual consumer in CRM? Rather than being the death of brand experience, AI might just be its saviour. If one of the biggest and most advanced AI machines still needs human guidance, then we shouldn’t be afraid of small task automation, personal assistants or AI powered productivity tools. Machines will destroy us Fiction According to the film franchise, the T-800 Terminator robot was programmed to kill the human race on behalf of Skynet. But the Terminator developed selfawareness and eventually helped to save humankind Fact The US military uses Atlas robots, which looks similar to the Terminator, for search and rescue. These robots haven’t been developed for warfare and the UN said

that no weapon should be operated without human control. Food for thought If the UN says ‘no’ to weapons for robots, will it say ‘no’ to driverless cars? If humans give up control of driving the car, we also give up control of any ethical dilemmas that may arise whilst in transit. PARTNERS ANDREWS ALDRIDGE | THE AI ISSUE 13 Source: http://www.doksinet WILL AI BE THE DEATH OF EMOTION? By Craig Lawrie, Strategy Director T he press recently had a field day when a new lobbying group, ‘Campaign Against Sex Robots’ was created. In the words of The Inquirer, the group was seemingly directed at ’people with money to burn and an itch to scratch’. US lad’s mag Maxim understatedly wrote, ‘it’s pretty unlikely that many people will form an emotional connection with an animatronic m a n n e q u i n th a t m a ke s m e c h a n i c a l moaning noises and can’t blink’. Yet whilst ‘Roxxxy’ (retailing for £7,000 and coming in a variety of emotional attachment

types) is a bit far off the vision laid out in the film, Ex Machina, it’s worth asking – has the team behind ‘Campaign A gainst Sex Robots’ cot toned on to something? In an age where more everyday tasks and jobs are becoming automated, is this having an affect on our own emotions? Does Roxxxy mark the start of the human retreat towards primal instinct over the power of emotion? And how does this affect the future of sentient devices with the power to emote, as well as rationalise? In an age where more everyday tasks and jobs are becoming automated, is this having an affect on our own emotions? 14 PARTNERS ANDREWS ALDRIDGE | THE AI ISSUE Actions are made on instinct; behaviours are taken on emotion Two fundamental factors separate human behaviour from the likes of Roxxxy. • Roxxxy may have been created as a sex toy, but she has no built-in need to have sex. As a robot, she is devoid of any sense of instinct - even if she is engineered to express ‘emotion’.

• Although we too are machines (albeit biological ones), we make choices based on emotional as well as rational factors. Humans are unique because we are able to differentiate between the two. And it’s this ability to separate the emotional and the rational that makes humans “intelligent”. As it currently stands, humans are the only species (biological or otherwise) whose behaviour is ruled by emotion. Although emotions by themselves are simple to feel, they are actually rather complicated and little understood. We just have to look at the human brain to understand why: each emotional experience is triggered by more than a billion neurons in the brain (made up of 100 trillion connections), resulting in a sensation personally felt and totally unique. Emotions are personal, intense and unique Even now, scientists have been unable to identify what emotion is or indeed how many emotions the average human genuinely feels. Given that much behaviour is learned, as humans and

intelligent devices start to inter-relate, what scientists can’t yet decide is whether humans will start taking on ever more mechanical behavioural characteristics or whether emotion will be the one point of difference between biological and synthetic machines. Will humans increasingly used to co-existing with robots in ever automated lives live of f default behaviour and instinct, or are emotions so deep rooted into each and every one of us that emotions will continue to separate man from machine? Is Roxxxy there to be abused at will or loved unconditionally? PARTNERS ANDREWS ALDRIDGE | THE AI ISSUE 15 Source: http://www.doksinet Manufactured happiness through imaginary friends Can you automate wellbeing? The tech industry is certainly trying. Meet ‘EmoSpark’, dubbed the ‘world’s first emotional AI’. At first glance, this 90mm box is nothing to behold. However, by analysing real time user facial expressions as well as what it knows about its owners from smartphone

data and Facebook information, it continually changes its emotional behaviour to match that of its ‘keeper’. In the most basic terms, it alters its personality depending on whether the owner is happy or unhappy. Whilst this device has the feel of an ‘affordable shrink’, in reality it behaves more like a butler. By understanding its owner’s emotions, it can use this data to play appropriate films or music, tell jokes, find out relevant information through tools such as Wikipedia and even ghost write emotionally relevant posts on Facebook. EmoSpark is still ver y early in its development, but there have already been warnings as to its actual capability to create genuine human happiness through artificial emotional intelligence. Discussing whether machines need to ‘learn etiquette’ 16 PARTNERS ANDREWS ALDRIDGE | THE AI ISSUE as humans start living in increasingly computerised worlds, the New Scientist is already forewarning of instances such as ‘coming home from a

funeral to find your AI itching to tell you all about the latest Facebook cat videos.’ Grief will become more common We m e n t i o n e d a b ove t h a t h u m a n s instinctively project their own image on to non-human things. Humanising objects builds a stronger level of attachment and builds a need to be loved. And with the possibility of more artificial pets in our lives, there will be more real-life grief - if, or more likely, when, they have been preprogrammed to ‘die’. This isn’t a trend that is being welcomed by everybody: Nicky Treverrow of charity Cats Protection has hit out, saying, ’programmed affection is no substitute for the real two-way bond between a pet and its owner. It’s sad to think a lump of metal could take the place of a sentient being with a unique personality.’ Treverrow’s point of view assumes that robopets are nothing more than c o m p u t e r i s e d to y s . R a t h e r, s o c i a l intelligence is being engineered into robopets so that

they learn the classic traits of any pet: companionship, love, obedience and dependence. As robopets become more like the real thing, so we’ll increasingly start to project our view of love on to them. As we move further away from the halcyon days of an analogue world, attachments will be made by whatever, rather than whoever, we have a connection with. With more and more robopets replacing the real things, grief will be even harder, once we are unable to differentiate between the death of a real animal and one that has been made in a factory. and emotional, they are also the only beings consciously able to know the difference between right and wrong. It is this fundamental principle that is being engineered in to AI systems to ensure that machines behave in a knowingly ethical way. The ethics of AI are certainly nothing new. Even as the Hot War was turning ever Cold, scientists, philosophers and lawyers started to seriously consider the threat of autocratic regimes run by robotic

beings. Asimov’s ‘3 Laws of Robotics’ – that robots can’t hurt a human, must obey human orders and must The death of emotion? Rather than a world where machines are the property of Man to be abused and subjugated on the whim of primal carnal desires, is it more likely that Roxxxy’s descendants turn on their erstwhile masters to enslave the human population and rule by fear? This scenario, whilst possible, is also rather unlikely. Emotions may not control us, but they do play out as consequences following actions that we make. And whilst humans are the only beings able to dif ferentiate between the rational Attachments will be made by whatever, rather than whoever, we have a connection with protect its own existence - have been treated as the blueprint for how sentient machines should behave in a co-existent world with humans. The fundamental problem with Asimov’s laws is that they were written for his novel I Robot and so were never intended to be applied for a real

world situation. Which begs the question: will sentient machines that had previously learnt to choose via rationality, begin to learn emotion through the injustice of being a second class ‘citizen’? Could the surprise of being different turn to sadness that leads to anger? Will machines only ever be ‘happy’ when in the company of other devices? Will anger be a double edged emotion, suffered by both humans and machines alike? Even in an automated world, it seems unlikely that AI will be the death of ‘emotion’. But as humans and machines start to live side by side, it’s the intensity of those emotions that is still an unknown factor. The question is, whether abuse suffered from the likes of Roxxxy, will be the catalyst for a new, raw and as yet undiscovered emotion that destroys the laws for it that Asimov once created. The end of the brand experience? So what then does all of this mean for brands and the people that look after them? Firstly, whilst robots may become more

human, it’s unlikely that humans will become more robot. Whilst rational behaviour may drive more of our decisions as the world becomes more automated, emotion will still determine who we are, how we are feeling and how we relate to others (humans, animals or robots). Emotion, will therefore, be even more important to drive a connection with a consumer, who will increasingly see the sentient as well as the salient in everything. Indeed, the role of the brand should play an even more pertinent place in people’s lives as we build deeper relationship with the synthetic. And as the likes of Facebook M aim to surprise and delight us by knowing more about us than we know about ourselves, the future will not just be one based on an emotional connection with brands, but one built on emotional dependency instead. This article was first published in the IAB’s Future Trends Report Volume 14: The Future of Robots. PARTNERS ANDREWS ALDRIDGE | THE AI ISSUE 17 Source: http://www.doksinet W

e’ve all read at least one article about Artificial Intelligence and the impact it will have on the jobmarket over the next ten years. ‘Who will be next’, ‘the death of advertising’, ‘copywriting to be automated’ – journalists, or the algorithms who replaced them, certainly know how to write an impactful headline. And if you read on, you will already know that AI is going to replace – if any – the kind of content that looks like it was written by a robot anyway. However, the fact that we get to keep our jobs for a little while longer doesn’t imply that creative advertising and quality content are immune from the influence of AI. Saatchi & Saatchi, for instance, have put out a prototype of an artificially intelligent ad that adjusts its output to the reactions of bystanders, so as to measure success But what if computers managed to replicate human creativity? Computational creativity, a simultaneously vague yet interesting research field, puts together p h i

l o s o p h y, m e c h a n i c s , c o g n i t i v e psychology and art in order to try and understand how exactly the creative process works. The application of any insight remains to be seen, but some of the results achieved by the discipline over the past 20 years are absolutely stunning. BRUTUS, for instance, is a piece of software that can create stories with a certain plot complexity – unfortunately they are not yet available to the public, presumably because they aren’t yet very captivating from a linguistic point of view. JAPE is programmed to invent funny puns, ASPERA produces fresh new poems if fed Well, that depends on the definition of creativity we believe in. Is it a state of mind, a spark of genius or does it follow certain sets of rules which could eventually be translated into a self-sustainable code, able to generate touching ideas and spotless executions? And more importantly, what is the role of all the software that helps us create content beyond the

possibilities offered by our human capabilities, such as making a portrait even if we can’t draw, recording a guitar solo when we’ve never played an actual guitar or suggesting words and associations when we are a bit short on ideas? Crucially – does AI just help to develop an idea that was already there or does it actively participate in the process? And if it’s the latter, does AI claim some ownership of Whilst a computer can mimic a Picasso, ideation and that raw moment of pure creation are currently only accomplished by humans. Can Artificial Intelligence become truly creative? in re a l-tim e. T h e s of t w a re ch o os e s combinations of images and text from a database, assesses the response it generates from those looking at it, then promotes the copy and design which performs best and eliminates the rest. Despite the experiment generating a lot of attention for Saatchis, it also showed how important quality content is for the future of advertising. If the results

offered by this Darwinian tool were to be taken seriously, we’d end up covering the city with kittens and hearts just because people react better to them. Behavioural economics, though, has largely proved that persuasion occurs on a deeper level and that most visual ‘nudges’ have a short lifespan. In other words, once the novelty effect of ‘the ad that knows what you’re thinking’ has passed, no one will remember the experience. with a set of words and examples and IAMUS wrote a beautiful score that was played by the London Symphony Orchestra. Google has also released DeepDream, an opensource software which detects faces and patterns in images, enhances them via algorithmic pareidolia and finally produces truly spectacular psychedelic dreams which blur the border between art, emotions and computers. Still, you might argue, a machine can’t learn to replicate the genius of, say, Picasso or Magritte. Well, wrong A group of German researchers developed a machine that can

create new work in the style of whatever artist you might like. If you’ve ever wondered how Picasso would have painted a portrait of President Obama, you now know. Each one of us may find computational c re a t i v i t y ’s a d v a n c e s s c a r y, b a n a l , pioneering or irrelevant for content creation. the creative process? Admittedly, there are more questions than answers, but we can assume that creativity always starts with a spark, an idea, an emotion, a desire to reach someone else and communicate a state to them. Whilst a computer can mimic a Picasso, ideation and that raw moment of pure creation are currently only accomplished by humans. Perhaps in a hundred years from now androids will have perfect human brains and they’ll be able to ‘feel’ and understand emotions. Until then, creativit y and communication will remain a task in which humans perform better than machines, regardless of the quantity of data and associations the latter can produce. By Gemma

Ghelardi, Content Strategist 18 PARTNERS ANDREWS ALDRIDGE | THE AI ISSUE PARTNERS ANDREWS ALDRIDGE | THE AI ISSUE 19 Source: http://www.doksinet WHY IT MIGHT NOT YET BE TIME TO RAGE AGAINST THE MACHINE By Matt Williams, Head of Creative Content F lick back to the ‘Expert View’ section on page 4 of this issue and you’ll find some pretty terrifying words of warning on artificial intelligence from some of the world’s most knowledgeable and respected figures. Yes, when Elon Musk reckons that “with artificial intelligence we’re summoning the demon,” you’d be forgiven for being a little bit nervous. And when Stephen Hawking backs that up by stating that AI “could spell the end of the human race,” then you’d think it might be time to break into full-blown panic. Indeed, much of this entire publication is focused on the dangers and fears that AI brings. If you were expecting a celebration of all things robot, you’d be very much mistaken. The theme of artificial

intelligence simply comes with too many caveats to ever feel fully confident in its future. So why then, are people investing so much money in artificial intelligence? And not just any old people – we’re talking the 20 PARTNERS ANDREWS ALDRIDGE | THE AI ISSUE world’s largest and smartest organisations, the biggest research and education facilities, and some of the most forward thinking entrepreneurs on the planet (including, ironically, Mr Musk). Part of the argument is that, like it or not, you’re never going to stop the emergence of AI. So you might as well do what you can to control its trajectory. This is Musk’s view, which is why he’s making all his organisation’s research public. This way there’s at least some sort of effort to regulate the progress being made, “so that we don’t do something very foolish.” But amidst all the doom-mongering and impressions of robot armies overthrowing the government, the counter argument is that AI could actually be

incredibly beneficial to so many strands of life. And we’re not just talking here about the immediate brand and business values that AI can provide and are covered in other articles in this issue. We’r e t a l k i n g s p a c e e x p l o r a t i o n . We’re t a lk in g m in in g. We’re t a lk in g experimentation that has either been too dangerous or too complicated for humans. That’s where AI comes into its own. There are new territories to explore. New planetary atmospheres to discover. For humans, even Mars is impossible to visit. It takes time – a period that we’re not sure the human mind can cope with. It would require food, water and oxygen – luxuries that aren’t currently readily available on the planet. And it would put strain on the human body that would be impossible to survive - be it space radiation, micro gravity or something else entirely. For artificial intelligence, it’s a fair bit easier than that. And that’s just to explore our second-nearest

planet. If we need to go further still, it’s AI that will be safely leading the way. And indeed contrary to all the ‘robots rise up’ stories, safety could be viewed as a primary long-term benefit of AI. We can dig for fuel in new places, mine in more dangerous and hard-to-reach areas. We can enter perilous places to save lives. South Korea’s Hoya Robot is already a great example of a machine that’s built to enter burning buildings, withstanding temperatures up to 320 degrees Fahrenheit to transmit image, sound and data to firefighters. As technology progresses, imagine how much more advanced these life-saving robots can be. Why put real lives at risk if a robot can eventually run into a collapsing building and physically carry out those caught up? Finally, of course, there’s the more cynical side to consider. Some of the scenarios mentioned are ideal for AI because it’s unsafe or even impossible for humans to negotiate themselves. But these – and many other – issues

are also better to leave in the hands of AI because they’d do it cheaper, quicker and with less fuss. Humans need feeding. They need sleep They need remunerating for their efforts. They sometimes get sick. They sometimes get demoralised. They have days where their performance levels are better than others. The beauty of AI is the constant it offers. It’ll work for days, weeks, months on end. It won’t complain or demand better wages. Mundane tasks are not seen as such. It won’t get sick (although may require maintenance, which is where humans can still come in handy), and if it does eventually burn out or break, you can toss it away and replace it with another one. Sound inhumane? Yes, but that’s the point. There’s no emotion involved Which makes tasks where emotion is seen as having a detrimental effect on performance much more palatable. Of course, we’re well on the way with this already. There are currently more than a million working robots in the world. Many are used

to sort through landfill sites, or work in assembly lines. A whopping 80% of all car production is now completed by machines. Robots like ‘Baxter’ have been created to work safely alongside people in factories, and are able to be programmed with new skills in less than half an hour. In the long-term these offerings will only improve. Things will become even more efficient. It’ll mean humans will be required for different jobs than the ones they currently do now, but surely that’s what progress is all about? And whilst there are of course plenty of issues that do need to be addressed with AI – Hawking, Musk and friends aren’t going to be twitchy about an issue for no reason – the benefits on offer in the mean time can be huge. PARTNERS ANDREWS ALDRIDGE | THE AI ISSUE 21 Source: http://www.doksinet Getting up close and R personal with AI HAZEL – your digital housekeeper Tested by Gemma Ghelardi, Content Strategist WHAT THEY SAY Hazel is the Mac housekeeper that

organises, filters and manages your files and desktop. She creates rules that can keep your folders in order – if a file type or name meets a pre-set condition, an action occurs based on that condition. WHAT WE SAY umour has it that Artificial Intelligence might one day make our jobs redundant. And there are currently plenty of AI apps out there that claim to be making our jobs and everyday tasks easier. A smart idea, but the free version just causes the window to run slower than usual and setting up rules for each and every folder on your computer is a lot of work. The idea of the app is great, but Hazel doesn’t do quite enough of the work for you. To take it to the next level, she needs to suggest sorting options and learn from your choices. THE BOTTOM LINE Hazel is based on a good premise, but the practice doesn’t live up to it – yet. 3/10 So as AI gets closer and more personal, we wanted to try out the latest and greatest AI apps. We wanted to know: these apps are

truly ‘AI’ or just some clever 01 Ifmachine learning If they can change the way we interact with one 02 another or genuinely make our everyday tasks easier 03 If they’ll ever be able to take our jobs KONO.AI – your mobile scheduler SLEEP CYCLE – your sleep tracker Tested by Greg Machray, Project Manager Tested by Lauren Sudworth, Content Strategist WHAT THEY SAY WHAT THEY SAY The AI mobile scheduler that makes creating events incredibly easy. Sleep Cycle is the app that tracks your sleep patterns and wakes you up at a natural time. It can help you find your natural rhythm. WHAT WE SAY WHAT WE SAY POCKET – save your articles for later SIRI – your personal assistant Kono.ai promises a lot – scheduling your meetings at the best time and in the perfect location – but it can’t deliver. As Kono can’t see other people’s calendars, it can’t guarantee their availability and you have to manually follow up. Venue suggestions are based on your location,

but they don’t take into account local knowledge – like that little café you know round the corner, rather than the busy Starbucks with no seats on Oxford Street. Tested by Will Sharples, Account Executive Tested by Gemma Ghelardi, Content Strategist THE BOTTOM LINE THE BOTTOM LINE Tempting idea, but the data isn’t smart enough to replace a real Project Manager! Next steps would be to get a proprietary version running on a shared Outlook server, so Kono gets access to good data. Smart and thought-provoking. If nothing else, it gets you thinking about your sleep cycle so that you can make tweaks. We conducted an agency wide experiment for one week. Here’s how six of the most cutting edge stacked up. WHAT THEY SAY WHAT THEY SAY The extension that helps you store and categorise information you find interesting, Pocket makes it easy to view articles later. Siri is iPhone’s personal assistant. She really requires no introduction, as we’ve all tormented her with

questions like ‘what’s the meaning of life’ and ‘can you make me a sandwich’. But jokes aside, can she actually make my job easier? WHAT WE SAY Pocket lets you create your own tags for categorising articles – which puts the onus on you to organise properly. Articles fall into ‘My List’ and ‘Archive’ and as long as you are tagging correctly, it’s easy to find what you’re looking for. Tagging can be a bit of a pain and it’s better to have a strategy in place before you start THE BOTTOM LINE Revolutionises the commute, but content could be displayed more clearly. 8/10 WHAT WE SAY Well, not quite yet. Siri is great at timing reminders and fetching unread emails. She’s not so good at long sentences or ‘work words’ – like ‘brief’, which she hears as ‘breeze’ – so you essentially end up rewriting the email she was supposed to handle. Plus, she can only sync tasks that are on Google Calendar, whilst other organisational tools are out. Not ideal.

THE BOTTOM LINE Siri can handle some of what you throw at her, but a lot of it gets lost in translation. 5/10 The app definitely makes you aware of how much sleep you need to function well. It can help you to redefine your habits, by getting you out of the negative cycle of sleeping late on the weekends and then starting the week with a ‘sleep deficit’. The only downside is that the alarm isn’t very useful - you need to play around with your own ‘sounds’ to find something that works for you. 7/10 2/10 CRYSTAL KNOWS – the best way to communicate Tested by Cat Melville, Account Executive WHAT THEY SAY Crystal Knows is a Chrome add in that aims to help people understand and communicate with each other better at work and in life. WHAT WE SAY The fact that Crystal can help you alter your messaging depending on contact and relationship can feel pretty invasive and is hard to get used to. The calendar sync allegedly gives you an insight into the personalities of people who

are going to be in your meeting – but unfortunately this felt more funny than useful. THE BOTTOM LINE A bit of a spooky one. Lots of potential, but fundamentally was not useful enough to justify the extra effort. 5/10 22 PARTNERS ANDREWS ALDRIDGE | THE AI ISSUE PARTNERS ANDREWS ALDRIDGE | THE AI ISSUE 23 T H E L AST WO R D By Alastair Cole, Chief Innovation Officer A I is everywhere. It’s in our homes, our offices, our bedtime reading. It’s in our algorithmic social media timelines and our one-hour delivery drones. Whether you’re excited or sceptical, AI will undoubtedly be one of the biggest talking points of our lifetime. And what makes it so is its uncertainty - we don’t have all the answers. As much as we humanise individual machines with talk of ‘intelligence’, we have to admit that the field of AI is much like a living organism – it’s 24 PARTNERS ANDREWS ALDRIDGE | THE AI ISSUE growing, changing and unpredictable. Whilst we can all speculate on

what might happen next, the honest truth is that its direction is uncertain. Will machines help us become the best versions of ourselves or will they destroy us? Will the impact be much less than that? Could it be even more? Our experiments proved that, if anything, the practice is still a long way away from the theory. The potential of AI could be huge, but the technology has a way to go first. Perhaps even more importantly, we have a long way to go too. Because without our whole-hearted backing, AI won’t develop. There needs to be enough appetite for new technologies and enough people willing to test them out. Otherwise, AI can’t develop much beyond an idea in somebody’s head. So with this in mind, we’ve developed a few AI inspired products to help brands push their boundaries. If you’re interested in these or anything like them, get in touch. We can promise you more than an automated reply. At least for now AI solutions Source: http://www.doksinet Our personal and

professional lives are being served by a growing number of systems and entities - many powered by Artificial Intelligence. Below are three products we offer our clients to help demystify this fascinating field and to add value to their businesses. Lunch and Learn: 01. Introduction to AI Over lunch, our team will walk through the latest developments, examples and trends within Artificial Intelligence. We look at how these are influencing customer experience today and which brands are investing for the future. We’ll also share the latest findings from our internal thought leadership programme on this topic. T h e s e s e s si o n s a re d e sig n e d to b e interactive and relevant to your business. We need at least one hour to cover the material and allow for questions and discussion. We bring at least two team members and ask that you invite a maximum of ten attendees. Price on application. Automation 02. Audit Our team will spend time in your business, speaking to key digital

stakeholders and platform owners. We will assess the levels of marketing and advertising technology automation, and gauge how your teams are utilising tools that provide machine learning capability. These sessions are designed to look at how your automation tools are configured and where quick wins could be made. We’ll review the frequency and format of data collection from these tools, and assess how this storage procedure prepares your organisation for future capabilities in Artificial Intelligence and predictive analysis. It’s also important for us to assess the human skills and capabilities necessary to support the growing infrastructure of automated technology. As this increases, your people will need to develop new strengths and your organisation will need to manage and retain new knowledge. Price on application. 03.  ecision support D Recommendations Our recommendations reports outline how new tools and practices could be used to improve automated decision support

inside your business. We organise these into quick wins, medium term improvements and longer term visions. Example areas in which decision support could add value to your business: • Audience profiling and segmentation • Customer acquisition through predictive prospecting • Automatic targeting and contextual content recommendations • Technologies and platforms that enable rapid and remote ‘test and learn’ • Computer vision (image recognition & machine learning) • Natural language analysis (text and voice) The findings will focus on the benefits changes would bring to your people, operational efficiency and bottom line. We will describe how they could impact and improve the everyday working lives of your people in cost effective ways. Price on application Get in touch If you’re interested in any of these products or if you’d like to talk to us about AI, contact alastair.cole@madebypartnerscom PARTNERS ANDREWS ALDRIDGE | THE AI ISSUE 25 Source:

http://www.doksinet alastair.cole@madebypartnerscom