There is a strong case to be made that artificial intelligence (AI) is now the most central topic in technology. While the computer science that underpins AI has been in development since the 1950s, the rate of innovation has gone through multiple step changes in the last ten years.

The technological reasons for this are well understood: the advent of neural networks; an increase in semiconductor processing power; and a strategic shift away from AI systems that rely on parameter-driven algorithms towards self-reinforced and multiplicative learning, machines that get smarter the more data they are fed and scenarios they negotiate.

Development has been open and collaborative. The benefits of AI in process efficiency and, potentially, accuracy are clear. For this reason, R&D activity, pilots and commercial deployments stretch to virtually every sector of the economy from healthcare to automotive manufacturing to telecom networks. A recent Vodafone survey indicated a third of enterprises already use AI for business automation, with a further third planning to do so. Take-up on this scale, at this rate, could put AI on a level with prior epochal shifts of electricity, the combustion engine and personal computing.

Two sides to each coin
Whether that actually happens depends on how the technology is managed. I spend a lot of time talking with major telecom and technology companies. While it’s clear AI is a major point of interest to nearly everyone, the discussion is still pitched in generalities. Paraphrasing:

AI is the Fourth Industrial Revolution
We know AI is big and we want to do something with it, but we don’t know what
We’re moving to be an AI-first company
How can we win with AI?
We’re a far more efficient company because of AI
The ebullient tone is to be welcomed.

Far less talked about, however, are the ethical and legal implications that arise from trading off control for efficiency. It’s fairly clear that cognitive dissonance is at work – the benefits blind us to the risks.

How do you answer these?

A crucial faultline is the balance between programmed and interpretive bias. That is to say, how much are machines programmed to act based on the way humans want them to act (reflecting our value sets) versus their own learned ‘judgement’? This has a direct bearing on accountability.

To make this point, let’s pose a series of questions that draw on how AI is being used in different industries.

Autonomous vehicles
If a self-driving car faces the inevitability of a crash, how does it decide what or who to hit? If that same self-driving car is deemed to be at fault, who bears responsibility? The owner? The car manufacturer? A third-party AI developer (if the technology was outsourced)?

Criminal justice
If an algorithm is tasked with predicting the likelihood of reoffending among incarcerated individuals, what parameters should it use? If that same algorithm is found to have a predictive accuracy no better than a coin flip, who should bear responsibility for its use?

Social media
If Facebook develops an algorithm to screen fake news from its platform, what parameters should it use? If content subsequently served to people’s news feeds is deemed intentionally misleading or fabricated, does responsibility lie with the publisher or Facebook?

I chose these for a number of reasons. One, these are real examples rather than hypothetical musings. While they emanate from specific companies, the implications extend to any firm seeking to deploy AI. Second, they illustrate the difficulty in extracting sociological bias from algorithms designed to mimic human judgement. Third, they underline the fact that AI is advancing faster than regulations and laws can adapt, putting debate into the esoteric realms of moral philosophy. Modern legal systems are typically based on the accountability of specific individuals or entities (such as a company or government). But what happens when that individual is substituted for an inanimate machine?

No one really knows.

A question of trust
Putting aside the significant legal ramifications, there is an emerging story of the potential impact on trust. The rise of AI comes at a time when consumer trust in companies, democratic institutions and government is falling across the board. Combined with the ubiquity of social media and rising share of millennials in the overall population, the power of consumers has reached unprecedented levels.

There is an oft-made point that Google, Facebook and Amazon have an in-built advantage as AI takes hold because of the vast troves of consumer data they control. I would debunk this on two levels. First, AI is a horizontal science that can, and will, be used by everyone. The algorithm that benefits Facebook has no bearing on an algorithm that helps British Airways.

Second, the liability side of the data equation has crystallised in recent years with the Cambridge Analytica scandal and GDPR. This is reflected in what you might call the technology paradox: while people still trust the benevolence of the tech industry, far less faith is placed in its most famous children (see chart, below, click to enlarge).

[1]In an AI world, trust and the broader concept of social capital will move from CSR to boardroom priority, and potentially even a metric reported to investors.

This point is of heightened importance for telecom and tech companies given their central role in providing the infrastructure for a data-driven economy. Perhaps it is not surprising, then, that Google, Telefonica and Vodafone are among a vanguard seeking to proactively lay down a set of guiding principles for AI rooted in the values of transparency, fairness and human advancement. The open question, given the ethical questions posed above, is how actions will be tracked and, if necessary, corrected. Big questions, no easy answers.

– Tim Hatt – head of research, GSMA Intelligence

The editorial views expressed in this article are solely those of the author and will not necessarily reflect the views of the GSMA, its Members or Associate Members.

[1] https://www.mobileworldlive.com/wp-content/uploads/2019/01/Jan-31-GSMA-AI-ethics-blog-Jan19-v2.docx-Chart.png

Depending on your perspective, CES either feels like it took place eons ago or it’s fresh in your memory. Personally, it seems pretty recent. I’m still seeing reviews of the show in the media. I haven’t finished logging all of my CES meetings in Salesforce. And, we’ve got an extensive, survey-based report coming up that, while not informed by CES, bears witness to a lot of what we saw there.

So, what’s the report about? Drones? Autonomous transport? Robots? The future of self-cleaning cat toilets? Nope, none of the above. It’s about Edge computing. That’s right, call it what you will, Edge cloud; Distributed Compute; Distributed Edge Cloud, there was plenty of Edge-related stuff going on in Las Vegas earlier this month. It might not always have been positioned as such, but it was there if you were looking.

If you’re not familiar with the Distributed Edge Cloud concept, it’s fairly straightforward and very powerful. At its simplest, it’s about siting compute and applications (including network functions) closer to the user. Doing so positions it as a way to deliver low latency communications where a specific use case requires them. And where edge nodes are able to host workloads from various players, the concept opens up opportunities to expose applications in the way public cloud players do, all while lowering backhaul burdens. Of course, it’s also positioned as a space where operators and public cloud players will battle to deliver value to the enterprise.

So, what did we learn in Vegas?

The Edge comes in many shapes and sizes
If you compare U2’s lead guitarist (The Edge) circa the release of albums The Joshua Tree, Achtung Baby and Songs of Experience, you’ll see he’s changed over the years, but we always know who he is.

The same can’t be said for Distributed Edge Cloud. Talking to people (operators and vendors) for our report, it was clear that everyone had a different definition of where the edge of the network was. Within an operator’s network. In the enterprise. In a user device. This definitional tension is about more than just semantics: it sets out issues of ownership and monetisation (in other words, who will benefit). It was also front and centre at CES with lots of different vendors positioning themselves as edge players, whether than means delivering home gateways, IoT gateways or high-end phones.

My favourite: an instantiation of Amazon’s Greengrass (extending AWS to edge devices) on a robot, putting the “mobile” into mobile edge computing.

It’s not just about nodes
Given its presence in smartphones, IoT and computing devices of all sorts, Arm was omnipresent at CES. But, rather than connect around something sexy like drones, wearables or artificial intelligence (AI), I took time to catch up with it about Fog Computing. While sometimes used interchangeably with Edge Computing, the two are not the same.

As former OpenFog Consortium chairman Helder Antunes put it: “Fog computing is an end-to-end horizontal architecture that distributes computing, storage, control, and networking functions closer to users along the cloud-to-thing continuum.”

Key here is the concept of an end-to-end “architecture.” We can quibble over the differences between Edge and Fog, but there’s an important reminder here that placing compute closer to users involves more than just nodes. It requires sites for those nodes, applications to run on them and management systems to get those applications deployed. Again, this is more than just semantics: different participants in the edge ecosystem will deliver different components. Where our study, for example, saw operators deploying the majority of nodes, it suggested webscale companies as deploying the majority of workloads.

It’s not just about enterprise
Part of the massive buzz around edge computing is the potential it holds for helping operators (and others) enable the digital enterprise: nearly 45 per cent of the operators we surveyed saw the enterprise sector as generating the most value from distributed edge clouds. CES, however, highlighted a clear role for supporting consumers. To be fair, commonly cited use cases like AR/VR, connected car, and mobile gaming all imply a consumer component. The same, however, holds for home IoT gateways, which do more than facilitate sensor connectivity. None of this is a revelation, but with a tight focus on the enterprise (and new operator revenues), it’s important to recall the consumer value proposition.

It’s not just about latency
The top business driver for Edge computing per our survey of operators and vendors? Application latency.

On a scale of one to five (the latter being “extremely important”) the overall rating was 4.2. This aligns well with a focus on edge computing support for use cases including critical communications and AR/VR, and it makes sense when considering the second most-cited business driver, user experience.

The problem? Latency often captures all of the attention, detracting from everything else we want to accomplish with edge computing. Operators, for example, will be looking at more than an improved user experience thanks to latency improvements. They’ll be looking for transport efficiencies and potentially regulatory compliance around how and where data is handled. And, on the user experience front, low latency is only one part of the story. Think about all of the low-power IoT devices launched at CES: if forced to do lots of processing, battery life and application performance will be compromised. But if that processing can be pushed up to (offloaded to) an edge node, app performance and battery life should benefit.

It’s not just about operators
For many people, the distributed edge cloud concept is inextricably linked to operators: mobile operators, in particular.

Long before 5G networks got trialled, we talked about Mobile Edge Compute, which was born from early efforts at integrating compute with RAN platforms. That evolved into something more holistic (multi-access edge compute), but the mobile and operator bias remains. 5G, for example, dominated as the chosen “most relevant” access technology in our survey, with Wi-Fi and fixed options bringing up the rear. Meanwhile, almost none of the operators we surveyed expect webscale players to generate the greatest economic value from edge compute.

Cue Baidu: in tandem with CES, the Chinese behemoth announced OpenEdge, an open-source edge compute platform, highlighting edge as a critical component of its AI, Big Data, and Cloud strategy.

Whether or not we need another edge platform (open source or otherwise) isn’t the point. That cloud players are actively targeting the edge and putting development (and outreach) efforts behind it, is.

– Peter Jarich – head of GSMA Intelligence

The editorial views expressed in this article are solely those of the author and will not necessarily reflect the views of the GSMA, its Members or Associate Members.

For an IoT industry analyst, CES is a real treat.

[1]Naturally, some connected devices on show left me scratching my head (connected underwear, anyone?) But I look at it through the lens of how we segment IoT at GSMA Intelligence: consumer vs. industrial or enterprise opportunities. We see both being massive markets (see chart, click to enlarge), but the latter growing at a faster rate. However, after wandering through the CES halls for a few days, I came to a few new conclusions about consumer tech. More than just consumer tech, though, they tell us something about IoT as a whole.

IoT is meaningless without data and analytics. Ginni Rometty, IBM CEO, dubbed AI as the “world’s greatest natural resource” during her keynote, and a technology that will empower revolutions across multiple IoT segments from smart cities to healthcare, transportation to robotics. There’s a lot to digest there… and I fully agree. We often highlight the fact that connecting devices alone isn’t what IoT is about. The real value comes from IoT data translated into actionable insights. So I was glad to hear that event organiser CTA recognised that as well, announcing a new megatrend at CES Unveiled: the data age. An age where consumer choices or business decisions are reached and supported by data. This was an underlying narrative during the show – how to translate sensor intelligence into tangible results. As one example, John Deere, a CES first time exhibitor, showcased just that. It brought its connected harvester to the showfloor and I even managed to hitch a ride in a self-driving tractor. Pretty impressive. The reasons behind implementing precision farming are clear, go beyond cost saving and yield increase to feeding our ever growing population. John Deere achieves this via a mix of connectivity and data analytics.

Voice brings the smart home together. Google’s presence at CES was overwhelming. “Hey Google” was plastered all over the monorail, a Google “fun ride” was just in front of CES’ central hall, and a slew of device manufacturers added the “Works with Google Assistant” sticker. Google also had people in lots of partner booths. Why’s that important? It flags that Google is spending some serious bucks on integrating with partners. Fear not, Amazon Alexa was not forgotten and added to new products, quite often alongside Google. As a result of increased reach, according to GSMA Intelligence’s new Consumer Survey, smart speakers’ household adoption has doubled to 15-20% across the US/Western Europe [2]. Although (typically) absent from the show, Apple made its way into the spotlight as well through new Home Kit partnerships. Why is this important? One simple reason: smart speakers have emerged as the prominent voice-control computing platform for smart homes [3], leading to the ‘resurrection’ of voice as a user interface. An open ecosystem offers the ability to scale up and extend market reach.

Ageing baby boomers get connected. Independent/assisted living solutions were on full display at CES this year and went beyond just alerts based on basic connectivity. The ultimate goal is to allow seniors to maintain independence for as long as possible, using a combination of wearables, beacons, sensors, and cameras with an AI overlay. Using AI powered platforms alongside sensors adds value and helps to manage daily activities. For example, Hive Link Connected Care learns a seniors’ routine using multiple connected devices and alerts the caregiver when an anomaly and/or inactivity occurs. Similarly, CarePredict monitors ADL (Activities of Daily Living), training neural nets based on individual behaviour and enabling preventative actions once changes in daily activities have been detected. Such services can also be an add-on to the carer’s existing smart home security service, e.g. Wellcam, an Alarm.com service. This is still a nascent market with “a trial and error” approach to service provision. The addressable market is rather large – according to the United Nations there are almost 1 billion people aged 60+ globally and this will double by 2050 to almost 2.1 billion, equivalent to 13% and 21%, respectively, of the total population.

Wearables get healthier. Omron is a great example. The company essentially miniaturised a traditional blood pressure monitor and turned it into a wrist watch, Omron HeartGuide. It isn’t pretty but that isn’t its purpose. Managing medical conditions using medical grade devices and analysing data via a secure platform with a built-in AI engine, which then reports to a physician for a diagnosis and treatment, is an ultimate goal. This can result in reduction in healthcare spend (which accounted for 18% of GDP in the US in 2017). Net net, it delivers a real and measurable value. Understandably, the US is an outlier here – private insurance offers more opportunity for health device adoption to manage costs. Other barriers to adoption include behaviour (patient and clinician) as well as reimbursement structure and regulation. But we’re seeing a real definitive shift towards a more preventative approach to healthcare – and wearables can help.

My favourite new tagline. #LikeaBosch is the tagline of a new global image campaign to highlight the company’s capabilities to deliver connected experiences. It was funny. It also drove home a message that there are still things that can be connected and interconnected to deliver better experiences. Should they be? Well that’s another question. Probably not everyone needs a connected coffee machine or lawnmower.

– Sylwia Kechiche – Principal Analyst, GSMA Intelligence

The editorial views expressed in this article are solely those of the author and will not necessarily reflect the views of the GSMA, its Members or Associate Members.

[1] https://www.mobileworldlive.com/wp-content/uploads/2019/01/IoT.jpg
[2] https://www.gsmaintelligence.com/research/2019/01/future-of-devices-smartphones-ai-immersion-and-beyond/717/
[3] https://www.gsmaintelligence.com/research/2018/08/smart-home-from-niche-to-mainstream-by-2025/687/

Last year, I kicked off a weekly column series from GSMA Intelligence and I’m excited to bring it back for 2019. Over the next 12 months, there will be columns from me and some from across my team of analysts (the ones from my team will likely be more insightful and better written – you’re welcome).

In any event, we’re now in the middle of the season when people (including analysts) take stock of the year that’s come and gone, make predictions about the one just beginning, and maybe even throw in a few resolutions. In past years, I’ve dedicated columns to recaps and predictions. I’m not sure why I’ve skipped resolutions. But since resolutions are some of the only things I can actually control in the industry, I’ll kick this year off with my mobile and communications-related resolutions for 2019. Think aspirations/commitments for how I want to work with the industry – things I hope to do in the New Year.

Here we go…

5G: Worry Less
This is going to be a big year for 5G. As a certain Head of State might say, it’s going to be HUGE. This is the year we’ll see commercial services (mobile and fixed) begin to ramp up and get the first crop of 5G smartphones and other devices. We’ll get a better understanding of how services will be marketed and priced. And, as the kinks get worked out of service offers and marketing plans – as well as networks – we’ll see lots of critiques. Devices will be called out as too expensive. Services too. In some cases, I’m sure the initial user experience will be deemed lacklustre while focusing too much on mobile broadband and not enough on innovative services. Some of this might be valid. None of it will matter much in the long-run. In the midst of this, I plan to ignore the worrying, recognising these initial launches for what they are – market testing and optimisation exercises that will set the stage for mature services and networks in 2020 and beyond.

4G: Worry More
Where 5G gets the spotlight this year, there’s a real risk that 4G innovation – at the service, device and network levels – could get ignored. Consider the case of operators looking to move people onto their 5G networks. It would make sense for them to prioritise 5G device innovation and direct capex into delivering the best, most competitive 5G experience. Right? Sort of. LTE will continue to carry the bulk of mobile data traffic for the foreseeable future; it will remain the primary mobile broadband asset for most operators. In markets where 5G hasn’t been rolled out, it will be the best option. In markets where 5G has been deployed, LTE will need to deliver a compelling fall back, evolving to integrate features (higher throughput, lower latencies) promised by 5G. Development of 5G innovations and ecosystems cannot come at the expense of 4G. I can’t do much to drive the direction of LTE R&D. But, I do plan to be vigilant – looking for instances of LTE innovation (product and service) and highlighting them as much as possible in order to make it clear that betting on 5G isn’t an “all or nothing” proposition.

AI: Build Something
By this point, we’ve all come to grips with the fact that Artificial Intelligence (AI, in its many forms) will fundamentally impact the way fixed and mobile services are consumed and delivered. From the camera settings on your smartphone, to the service package offered at retail, to the placement and maintenance of network assets, AI will likely be involved. The AI revolution, however, is less about the sophistication of the technology and more about its accessibility. As AI capabilities have become easier and easier to integrate into existing software and hardware solutions, companies of all stripes (including operators and network vendors) can put it to use for their own needs – evolving its application and potential in the process. Standing on the side lines, of course, is no way to understand this dynamic. For my part, then, I’m going to put AI to use within GSMA Intelligence. No, I’m not sure how. But trying to understand and deliver insights about a major industry shift (potentially the major industry shift) by simply talking about it just doesn’t make sense.

Multi-play: Ignore the Big Boys
We officially launched our multi-play (fixed and video/content, alongside mobile) coverage in 2018 and it was a good year to do so. After all, it was the year that AT&T acquired Time Warner, Vodafone picked up a variety of Liberty Global assets, and a comedy series from Amazon (not HBO or Netflix) took home five primetime Emmy awards. There was a lot more that happened, but you get the idea; 2018 provided plenty of examples of how the fixed, mobile and content industries were co-mingling and evolving – creating a new content and communications landscape. And yet, as it often does, the headline news focused on a small sub-set of players. The big ones. The ones shaping that new landscape. And everyone else? The content and access strategies of smaller operators? The specialist content providers? They can easily get overlooked. Obviously, there’s no way to ignore the big players, but there’s only room in the market for so many carriers like AT&T and Vodafone or content providers like Amazon and Netflix. To that end, I plan to spend more time this year focused on the smaller players – for lessons and insights that impact a broader set of market participants.

IoT: Focus on Users
In the same way that multi-play conversations often revolve around the machinations of a few big companies, IoT conversations often get derailed by a focus on a few valid but short-sighted (big picture) considerations. How many connections do we have? Which technologies are getting used? How much revenue are they generating? Who’s getting their share of that revenue? This all overlooks a very fundamental reality; everything begins with the use case. Whether for consumers or enterprises, the broader market success of IoT (those big picture metrics) depends on how it’s getting used and what it delivers – in line with the specific requirements of a given enterprise vertical or consumer segment. Companies that understand this are the ones who stand to benefit. We’ve spent plenty of time at GSMA Intelligence delivering insights into the size of the IoT market and what’s driving it. In 2019, I want to take us back to basics, with a deeper focus on the “how” of IoT vs. the “how big.”

Some of these will be easier to stick to than others. On the IoT front, for example, we’re wrapping up a massive survey of enterprises which will deliver insights into how IoT is getting used and the demands of specific verticals. Regardless, you can hold me to all of them. As we roll into the final stretches of the year, feel free to check in on how I did (Twitter [1] or LinkedIn [2] – they’re good ways to find me). And if you want to send me your resolutions before that, I’m all ears. You might even convince me to add some to my list!

– Peter Jarich – head of GSMA Intelligence

The editorial views expressed in this article are solely those of the author and will not necessarily reflect the views of the GSMA, its Members or Associate Members.

[1] https://twitter.com/pnjarich
[2] https://www.linkedin.com/in/peterjarich/