Harry Potter and the Neural Network
Over the past few weeks I have been rewatching all eight Harry Potter films - something which has been both nostalgic, heart warming and left my search history filled with various forms of “how many witches and wizards are there in the magical land of Harry Potter?” - (the answer, in case you’re wondering is decidedly inconclusive and I would strongly encourage JK to get a tweet out asap so I can put this issue to bed)
For those of us who are in our mid-20s to mid 30s, Harry Potter was such a large part of our upbringing. Queueing to get the books, arguing with family members over spoiler alerts and prolonged periods of silence as we raced through the latest trial and tribulations of dear old HP.
For us Harry Potter was just so enchanting with every spell carefully crafted to inspire awe. However we were a pre-smartphone, pre-internet, encyclopedia owning, blue eye shadowed, selfie-less generation and I can’t help but feel that by the time we try and read Harry Potter to our own children it will be like force feeding them Enid Blyton.
So much of the magic has fallen to the remit of technology. Find my Friends is like Marauders Map 7.1, contactless payments surpass any sickle or knut and Blockchain proves far superior than a 3ft Goblin offering vault protection and if a muggle saw a flying car would they really be that shocked?
There are of course things that will remain magical. Anything which defies the Laws of Conservation of Energy such as magicing fire, any major force or whisking up a bouquet of flowers from a wand can remain firmly in Hogwarts territory.
Whilst it may seem that I’m dispelling the illusion and literally spoiling the magic, I would argue the opposite. I sometimes feel that technology creates its own chicken and egg situation. Does acknowledging our imaginations give us the license to make their contents a reality with technology? By taking our imaginations seriously can we literally make magic? I would bet that Elon Musk at some point in his childhood said “imagine if you could make a car fly”. Is find my friends in someway, inspired by the Marauder’s Map?!
Magic is defined as ‘The power of apparently influencing events by using mysterious or supernatural forces.’ Sub out the word supernatural and that almost sounds like the effect of machine learning and neural networks. I expect most reading this have experienced the almost alarmingly appropriate instagram advert which echoes the obscure conversation you’ve just been having with friends. Is the code behind this chain of events our modern day magic?
It makes me think how the Harry Potter books might evolve if they were to get a facelift for Generation iGen. No longer would the wisest wizards be the most well read, would Wizards, like muggles have google with knowledge at their every fingertip? Pub quizzes have had to adapt to the influx of iPhones so surely so would the Wizarding World. Would wands have a siri like feature in which you could ask for a spell that can turn a frog into a plant pot and, “bang”, the wand perform the appropriate spell for you? How would Voldemort exert his force - take down the UK’s WiFi?
I feel that wizards may have a new respect for muggles and the rapid advances we’ve made. It may even blur the lines between the wizarding and muggle world. But perhaps I’m taking this too seriously and should just accept that fiction sparks the imagination and we are increasingly equipped with the tools to turn imagination into reality.
I just hope that technology takes inspiration from the likes of Harry Potter and not the death eaters so we can make everyday life a little more magical. And failing that, if Uber could change “request Uber” to “accio car”, even for just one day it may make the 15 year heartbreak at the lack of Hogwarts Letter a slightly less bitter pill to swallow.
Mary Berry's Blockchain
Blockchain - it's about as mysterious as it is set to transform just about every industry.
The easiest way to explain this is to think of sending my friend Mary Berry a cake. Historically I might send this in the post. The issue is that I have no guarantee that the postman won’t eat the cake. This is where blockchain comes in.
Instead of trusting a third party, in this case the postman, I can send the cake in a distributed way. In other words, an open decentralised database.
This is like splitting the cake into slices and asking some trusted friends to look after each slice, verify that it is what I say it is before putting the slices back together and delivering to my friend Mary Berry.
However, if someone was to intercept one of the slices, say a hacker they would easily be able to guess that the slice is cake.
With the help of encryption, blockchain adds another layer of safety. Imagine if you could reverse the baking of the cake so that instead of sending slices to my trusted friends, I send the individual ingredients. That way, if a hacker detected say an egg instead of a slice they would have no way of knowing that that egg was from a cake, it could be an omelette, a meringue or a pie or a pancake.
These can be sent and the ingredients can be mixed to form the cake, which can now be enjoyed by my friend Mary Berry.
So why should you care? Technology has led to the rise of the sharing economy, where trust is the global currency.
For a transaction that requires trust, whatever form it may have, the chances are block chain will play a part.
Block chain is in its infancy and will be as monumental as the introduction of the internet.
Techno Techno Technophobia - (untz untz)
Why we needn't fear the self-driving car....
Technophobia. The fear of technology - apparently number 84 in the World’s top 100 greatest phobias. At 84, it sits firmly between the fear of the future and - the fear of loud noises... Unsurprisingly, the fear of safety does not feature on this list. Self-driving is fundamentally motivated by safety - by the chance to reduce fatalities caused by human error and prevent as many as 94% of the deaths caused by dangerous driving. In my mind the question of should we fear self-driving, is already at odds with itself.
So why “fear”? It has become the defining emotion when contemplating a technologically advanced, increasingly automated future. A future in which hierarchies are potentially questioned and, sometimes reversed. As humans, we are naturally suspicious. We fear change. We don’t relish disruption. Yet science has always disrupted society and society has always adapted. And, society has always advanced, too.
We’re in the midst of a mobility revolution poised to bring more change in the next decade than we’ve seen in the past 50 years. Not only is the rise of self-driving modes exciting. It is inevitable.
As robotics, connectivity and the shared economy have become more prevalent, new players have entered the automotive playing field. Players that have disrupted the future of autonomy - the likes of Google, Uber,Dyson, Lyft.
Automotive companies have a responsibility not to fear driverless cars but use their expertise to shape the boundaries in which they function in order to deliver the many possible benefits.
FREE TIME. FOR FREE
With any new technological deployment we cannot truly predict all of the benefits to the customer due to the unpredictable way in which it may alter societal behaviour. Would snapchat have predicted their platform as a conversational tool? Or would camera phones have predicted the selfie?
The only thing we can be certain of is that self-driving cars will absorb tedious tasks and in doing so, offer users time. More free time. For free.
Research suggests that using money to buy more free time, such as employing a cleaner to remove some of your daily chores, improves wellbeing. In other words, time is the new luxury good.
Imagine turning up to work on top of your emails rather than frazzled from the traffic. Level 4 will begin to take on more and more of the driving task meaning that we will begin to spend our time whilst travelling instead of spending our time travelling.
Whilst Marketers will sell a romantic, utopian view of future mobility it does stand to save the UK economy £30bn annually in regaining some of the unproductive time lost to congestion. Increasing levels of ADAS are already delivering on removing the mundane elements of driving in traffic in preparation for the soft handover to Level 4, for example automated lanes on the motorway.
People will gain time. What else?
Skills. Throughout history, technology has been perceived to threaten jobs. The very word “sabotage” gained its meaning from the strike action of workmen throwing their wooden shoes – Sabots; into new machinery. However, history has shown us that every technological innovation has brought with it new jobs, new supply chains as well as upskilling. Self-driving will generate unprecedented revenue streams as we generate more data, more customisable experiences, more ownership models and change the way we interact with our vehicles.
But Fear. What do people fear?
It seems we fear the things we depend on but don’t seemingly have control over. Driverless cars are predicted to generate approximately 1 Tera Byte of data per hour – the equivalent of approximately 1 million iPhone pictures. What happens to this data? How vulnerable does the data make a user to hacking, let alone the possibility of hacking through any one of the sensor channels?
These are questions we cannot answer completely in such elementary phases of development. No software system can be tested to 100%. This is an issue that won’t be solved overnight or determined by one single automaker. Standardisation of testing, validation and homologation will require government guidance in the same way the automotive industry is legislated today.
User trust will be the key barrier to adoption. We are already seeing Level 2 technologies and can expect level 3 later this year. We accept technology when it is iterative which is exactly how autonomy will arrive; complemented by incorporating increasingly sophisticated Human Factors learning and always the option to resume control.
YOU, ME AND DUPREE (AND INDEED EVERYONE ELSE)
It is perhaps appropriate at this point to discuss the “trolley problem” - the most common way individuals and society articulate their fear of driverless cars. Ie will a car hit one person to save three?
The truth is, we cannot code for every situation or hypothetical, such as the “fridge from a bridge” example. There is no ethics requirement from a human driving test today; so are we obligated to tackle these dilemmas? If we do, are we making a concerted decision which otherwise would have been left to judgement and instinct?
Machine learning, neural networks and offline learning all include human input. In doing so it opens up the possibility to include an individual’s view on morality - introducing the risk of ultimate accountability. Machines are a product of their coders, we have seen recently accusations that Amazon’s Alexa is a left wing feminist.
What’s more is that machines learn from the past without any insight into the future. How therefore can we get a machine to “unlearn” events that are irrelevant to the future driving task? Automating driving tasks does not remove the power from the driver. We should start to think not in terms of artificial intelligence but augmented intelligence. An intelligence which serves as an extension of the user, allowing them to exist in their most efficient form.
Approach to morality therefore, cannot become a distinguisher between brands and will require unanimous alignment from ethicists, policy makers and automakers. What we can do is apply logic, mathematics and physics to calculate and assess situations, ensuring as far as possible that the car is never faced with a dilemma in which it must choose one life over another. Judgement requires humans and Level 4 will cater for this impetus.
However, the trolley problem is largely overridden by the opportunity to save lives. With self-driving we can begin reducing the number and severity of accidents currently caused by human error. Driver distraction, drowsiness, drunk drivers, elderly drivers, and recklessness to name just a few contributing factors to accidents, will begin to disappear. Safety of vehicles is and must remain the number one priority for OEMs.
Automating some or all of the drive democratises mobility. Elderly or disabled could be afforded the same freedoms granted to vehicle owners today. Coupled with changing ownership models and connectivity, users would benefit from increased access to transport options. Social isolation amongst the elderly is thought to increase mortality by 26%. Automated vehicles could provide first and last mile solutions to make better use of public transport solutions. Combating social isolation could significantly reduce strain on existing public services and dependent relatives.
DE-CLOGGING OUR CITIES
The benefits to society will extend to improving town and city environments. Increased green spaces, reduced congestion and polluting gases could all be realised by level 4.
At present 30% of the time in a car is spent looking for a parking space. Automated Valet – self parking services could allow cars to be parked away from built up areas. Connecting vehicles in the short term would allow cars to move off from traffic at the same time, reducing congestion and energy lost during idling. Shared mobility goes one step further and would enable a car to be in near constant use. Not only does this reduce the need for parking altogether but would mean a car’s end of life mileage is achieved approximately three times more quickly. As a result there would be a greater number of vehicles on the road with the most recent technology and inevitably greater efficiency.
In 2017, the government backed plans to investigate vehicle platooning between vehicles, creating “road trains”. Reduced interaction with other vehicle wakes improves rider comfort and offers substantial aerodynamic benefit with some studies showing a drag benefit of 20%.
Aside from the obvious congestion benefits, vehicles would not be impacted by drowsiness, irritability or prejudice. Roads would become a more pleasant environment for road users. Adherence to speed limits and limiting “bad” behaviour has been shown to increase throughput. TFL launched a study in 2016 which showed that using both sides of the escalator increased throughput by 31%. The findings have been transferred to a number of motorway trials which are investigating 50mph limits. Autonomy would not suffer from road rage or deviate from these findings.
These benefits could be doubly realised in the autonomous transportation of goods. Autonomous delivery of goods could mean road networks are used more effectively with more goods transportation at night and more passenger movement during the day. Autonomous haulage at night when there are fewer people could enable acceleration of Level 4 technologies on our roads as well as a plethora of yet more business models and revenue streams.
Diesel demonization and promises of 100% EVs by 2040 mean that Autonomy is likely to arrive on an electric platform. The benefits already outlined only stand to be exacerbated by a motor which has increased control and existing environmental credentials.
THE END OF MY WAFFLE
Self-driving modes will maximise our time. Make us greener. And above all, make us safer. However, there are many issues that we cannot begin to tackle with meaning in 10 minutes. Fear of autonomy is undeniable. Let’s not discount this fear.
It is a useful stimulus that pushes us to explore every aspect, facet and detail.
But, let’s not deny technology either. Society has always adjusted to such transformations, history tells us so over and over again.
New technology always brings questions and threats. Loopholes will be found to exploit. The internet for example has stimulated new forms of terrorism, bullying and addictions, and yet it has democratised information and education - allowed rapid response to disaster and even founded new forms of dating.
The debate we are having is too accusatory, too negative, too fearful.
Saving lives. Does that make for a debate? Don’t fear the mass adoption of driverless cars. Fear the questions that need to be answered along the way.
Let them fuel the debate, so that we address them frankly and answer them innovatively - so that we will realise a socially responsible - safer future.
The Perks and Perils of Remote Working
This week a mother of two returned to work in our team following two full years of full pay maternity leave for her two children. She has come back to work exactly where she left off with a flexible arrangement to work three days a week from home.
This is incredible.
She is proof to the rest of the team that if your employer is accommodating, then flexible and remote working can be made to work for both the benefit of the parent and the employer. I fully recognise just how lucky I am to be working in an organisation that will go the extra mile to look after its working parents.
Her arrangement is largely enabled by internal chat, live shared documents, conference and video calling. Many hail such moves as the start of the completely remote workforce, predicting the office will soon become a dated concept. However I have to say that such a notion makes me feel just a little bit sick.
Last month I moved house and now my commute is a bit more meaty so I often work a day a week from home. Google working from home and you’ll be bombarded with pictures of people with a laptop open on their lap, sat on the sofa, cup of tea in hand and donning some John Lewis pyjamas. My reality is somewhat less Centre Parc Leaflet friendly. Yes I’m in my pjs, but I wore them last night and I got up when I needed to start working so I have elected not to shower. I’ve had breakfast but I’m near my kitchen and so I had it twice. And typing from the sofa makes you hunch your shoulders and is frankly less than comfortable.
Despite my large appetite and poor personal hygiene, this skims the surface of my issues with remote working. Quite simply, I miss the idle chit-chat, the facial expressions that convey what a person is really thinking and the nuances that make working relationships with colleagues that much more effective. It’s also bizarre finishing your working day and no one says “have a nice evening”. I miss how ideas escalate, akin to going out with your friends and suggesting a big night out and before you know it you’ve all booked flight to Mykonos. Creativity from home is just that little bit more stale.
Radio 4 recently looked at the benefits of emojis and how they have helped convey tonality and prevented misinterpretation of the tone of emails. But does a smiley face emoji really replace the power of a face-to-face exchange?
In 1986 the Space Shuttle, Challenger exploded 73 seconds after take-off killing all seven crew members. The O-rings were the cause of the failure due to the air temperature on the day of launch. The suppliers of the O-rings were not comfortable with the decision to launch. However the debate between NASA, the Government and the suppliers over whether to go ahead and launch was conducted over several conference calls. Each party was not privy to the concern stricken faces let alone the panicked conversations that would have occurred between conference calls. Instead each could only hear the pressure and eagerness to stick to the planned launch date - a conference call which ultimately led to the death of 7 people.
In my field of work my phone calls or emails will only ever piss someone off and I’m thankful that’s the extent of the damage I can cause. However, I can’t help but feel this remote-ness gives us a layer of cowardice. We can be more brazen in our decision making by hiding behind our screens from the comfort of our homes. The rise of job terminations via email is astonishing and I genuinely think GenZ consider dumping over text a standard course of action.
That said, there is a lot to be said for remote working. If office workers didn’t have to commute congestion and by consequence air quality would significantly improve. Reduction in need for childcare may allow more parents to return to work. And of course, sometimes I do relish the opportunity to simply crack on with work, albeit whilst smelly and bloated.
In spite of all of this what concerns me the most is how remote working could eliminate the opportunity to check in and simply ask “how are you?” and properly interpret the “fine thanks, how are you?”. I work on the phone lines for a charity and I am amazed at the lack of support so many people have and the pure desperation many have for someone to ask the simple questions.
We often think that “how are you” is a pointless segway into the conversation we actually want to have, but if no-one is asking then its importance cannot be underestimated.
Perhaps I am old fashioned and should cherish the flexibility I’m afforded in my job and it really is fantastic to be able to work a day a week remotely. But we do need to exercise caution - technology may make us more flexible, give us more freedom, allow us to be connected everywhere but it could also lead us to feeling lonely everywhere. We must be cognisant of the importance of face-to-face human interaction.
In short, be mindful of the perks and the perils and for the meantime I sincerely apologise to my boyfriend as I almost certainly still won’t have showered when he gets home.