Is the city of the future going to be a cross between Tron, the horrible Colin Farrell remake of Total Recall and Metropolis? Yes and no.
While that might be the ultimate goal as to where humanity is heading, we'll still take a long time getting there. The building blocks of that vision of the future will be much less severe and people driven, however. In a nutshell, a metal robot will not be ruling over us for the near foreseeable future.
TRON wiki fandom
So, just how are things going to change? I started thinking about this very aggressively since my first son was born recently, wondering what sort of future he would grow up in. I spoke to a wide-range of experts in Dubai on the sidelines of the AI Everything summit, and one thing became very clear - depending on who you ask, the vision of the future varies drastically.
For the sake of understanding -- the ideas you read about below can be broken into different timelines: the short term, intermediate, intermediate to long term and the long term future. The first two will probably happen in our lifetime, there is some likelihood for the third happening in our lifetimes, but we definitely won't be here to see our long term future. For this article, the focus will be on short and intermediate only because that's imminent.
How we move around will be a key pillar around which cities will modernize and become smart. For one, our transport is looking at a major overhaul. Already, there are more flying car startups than food, and autonomous vehicles are getting better and better every day. At the same time, partially autonomous cars and fully autonomous cars are being aggressively tested.
Reuters
Cities like Dubai have clearly set themselves a goal to make at least 25 per cent of their total transportation autonomous by 2030, leading with metros, busses and taxis. Elsewhere in the world, Uber has been testing their self-driving taxis. On the border between Dubai and Abu Dhabi, the world's first commercial hyperloop is racing to meet its 2020 deadline.
Jobs is another thing that's looking at a major overhaul. In China and other parts of the globe, some smaller blue collar jobs are already getting affected.
A report by McKinsey Global Institute estimates that as many as 800 million jobs worldwide could be lost to automation and/or robots. Kai Fu Lee who is credited as an AI expert and VC in an interview claimed as many as 40 per cent of the current jobs will be affected due to the incoming AI-Robotics wave. At the head of this disruption will be jobs of people who drive for a living since AI in transportation is a major focus area for companies. Lee estimates, these jobs will be disrupted in the next 15-25 years (between 2034 and 2044).
Hyperloop
Financial experts have warned that banking industry could be 'ground zero' for job losses due to automation. China has already been testing robots who can write articles at a fraction of the time human beings take -- so SEO is about to become AI's bitch. Closer home, World Bank estimates as many as 69 per cent of the Indian job market and 77 per cent of the neighbourly China's job market are under threat, if automation isn't planned out.
There is however, a second side to this argument too. Speaking at AI Everything in Dubai last month, David Cox CTO of IBM-MIT lab said, "It won't be a job catastrophe as much as a job transformation." Lord Tim Clement Jones who's the chair of the UK's House of Lords committee on AI compares, the shift in jobs to "US' shift away from agriculture between 1910 and the turn of the century." "But it (the shift due to AI) doesn't mean people won't study Social Sciences," Dr Eng Ling Goh, VP and CTO High Performance & AI for HPE says.
The argument that these three gents are trying to make is that AI will take up the jobs humans don't want to do anyway and leave them to upskill and do other jobs. It's another matter that there will be a significant cost for upskilling such large swathes of population and it's unlikely that a chunk of these people will be able to afford education and upskilling. Which leaves the governments of the day in the respective countries and whether they are willing to foot the bill. Remember, jobs or lack of them are a major poll issue across countries right now from US to Europe and India, people are angsty about jobs going to immigrants, would they be equally pissed off when the same jobs go to a robot or an AI program?
Smart cities are smart in a way existing cities aren't, obviously. The scope and depth of a city's smartness is something that's widely open to interpretation right now. But fundamentally will the next set of smart cities be built from scratch or will it be built by transforming an existing city? Dr. Aisha Bin Bishr, Director General of Smart Dubai office thinks it's going to be a heady mix of both.
She cites the example of Dubai and compares the landscape to green and brown fields. Green fields are completely barren empty lands where the seeds of the smart city are being sown (like Amaravati in India) and brown fields are existing cities old cities like London, Paris or even New Delhi which will undergo a transformation using AI to become smart and overall happy. The first step of this is being undertaken by her department to make the city run completely on blockchain by 2020 (first in the world).
Shutterstock
The very next year, the government aims to go completely paperless. You could of course argue that it's easier to implement these measures in a smaller city (by population) than some of the bigger cities and countries, but that's why there's now a 'AI Principles' and Governance document, the city is inviting others to follow in governing AI and limiting the sort of control it might exert on human beings.
According to Omar Sultan Al Olama, UAE's 29-year-old minister for AI there is not enough emphasis on the "Tangible impact of AI" but on fear mongering.
"If you (govts and society) really are worried about AI then do something about it, take some steps. If there are concerns let there be legislations and policies deployed to govern it. And show people the tangible impact AI could have on their lives."
"There are a lot of positive implications of AI. For example, in healthcare and diagnosis - AI could diagnose diseases better than the best doctor, and there would be a human doctor in the loop to help in the therapy process to help people heal. AI also has a lot of promise in logistics, infrastructure, minerals etc. There are negative implications of AI in autonomous weapons, etc but we aren't looking at that at all, we are looking at AI as a tool to serve humans and increase their happiness, the value they are getting from their govt," he says.
Robot doctor
Cities like Dubai have already started training their officers to use hoverbikes, hydrogen powered drone is already up to monitor hard to reach areas and they have also launched a high speed pseudo satellite which can go as high as 65,000 ft (the highest certified altitude for a passenger plane was the Concorde at 60,000 ft). But a clearer, more obvious implication of AI in the smart city if you go by Dubai Police plans is driverless patrol cars reading number plates, drones, robots and automated vehicles issuing fines. They are also looking at equipping close to 10,000 CCTV cameras with facial recognition software to analyse behaviour and detect offences.
The CCTV cameras equipped with the facial recognition as a topic, if you've been reading is something which has come in for quite some stick, especially due to the moral and ethical concerns on it. Major Dr Rashed Alghafri, Chairman of the Dubai Police Scientist Council, maintains that while there will be challenges in AI, the key lies behind whether or not the state would be creating a database of all citizens.
Reuters
While policing is unlikely to go into the Minority Report space when AI is predicting crimes before they happen on cue balls, there will almost certainly be a role of AI in predicting/identifying crime prone areas. This combined with proactive autonomous vehicle policing may just be the new normal for our cities.
It does however, threw up the second big concern after the lack of privacy -- false positives. Just last month, an 18-year-old is suing Apple for falsely recognizing him in a burglary, at an Apple store. AI is far from perfect and very often inherits the biases of its coders. But at the same time, it's also clear that there isn't much getting away from using AI to govern/police the AI-led smart city.??
AI is already being used in border security as countries gear up to handle more and more tourists, students and refugees. And it might be a lot less offensive and a lot more cost efficient than building a wall.
They may and they may not. China has already tried out a rating based system (Zhiva rating) which integrates with a governmental blacklist and has an impact on everything from loans to perception. While a rating system might be a little bit less extreme, dumbed down versions of it might already be operating -- for instance Uber drivers give rider rating based on how much a rider made a driver wait, how he/she was to interact with, all it needs right now is for this to start interacting with your other social media like Facebook and Instagram and see how many people unlike your posts to come to the conclusion that you are not a likeable person.
Netflix
You could then be de-prioritised on Tinder, if flights get in on that, they could use the information to decide who to give an upgrade to etc etc. This could also connect to a pecking order in jobs or even visas -- USA has already made it mandatory to share social media account details for visas. Translating likeability or other attributes into building the new haves and have-nots isn't really a big leap.
Clearly, there's a lot of potential for AI to improve our lives and make our society and cities better. But while we're busy building that future, we mustn't forget certain AI principles. Here's what's at stake:?
AI Ethics