Pages

Pages

Tuesday, August 30, 2016

Thoughts on Jeremy Corbyn's "Digital Democracy Manifesto" launch

(Disclosure / intro for any politics-type people who don't usually read my stuff: I'm a technology & telecoms analyst & consultant. Much of what I do involves forecasting realities of new tech innovations & deployments, part to how the tech industry commercialises its products/services and part to policy and regulation of the Internet & telcos. 

While I'm "neutral" on tech futures aside from wanting to see the "next big cool thing" as fast as realistically possible, I'm not always neutral on policy; for example, I support Net Neutrality as a broad principle. I'm also not a Labour party supporter, and I'm definitely skeptical of Corbyn's ideology, policies and general competence).

This morning I attended the launch of the "Digital Democracy Manifesto" (link) by Jeremy Corbyn, current leader of the UK Labour Party, and associated team members / advisors including Eva Pascoe, a longstanding member of the UK's web establishment (see link here). Although in theory this event was part of the current Labour Party leadership campaign, given the high probability that Corbyn will remain leader it also gives an indication of future direction - and probably forms the basis of the technology bit of the Labour manifesto for the next UK General Election, which will likely be in 2020.




Amusingly, the event was held at a venue called Newspeak House (a "hub for political technologists") the name of which derives from George Orwell's 1984. I'm sure for general tech-inspired activism that has resonance - but perhaps less than ideal for a Corbyn speech given Wikipedia's note that "any form of thought alternative to the party's construct is classified as 'thoughtcrime'". 

I'm definitely guilty of "thoughtcrime" in this post.... starting by highlighting Corbyn's first name-drop of a tech company as fond memories of Amstrad, purveyor of 1980s/90s PCs. Amstrad stands for Alan M Sugar Trading. Mr Sugar's views on Mr Corbyn are somewhat less than reciprocally positive (link). Corbyn's second reference was to Skype - amusing given Pascoe's later rant about Microsoft during the event. I should also point out that the seating space in the venue was limited & some attendees had to stand - something of an irony given recent concerns about train capacity and seat availability.

 

There are 8 areas in the manifesto, which were covered in greater or lesser detail in the event, and which fit to varying degrees with my own coverage. I'm not going to talk about the "open knowledge library" of learning materials, and given that the "people's charter of digital liberties" will be driven by public consultation it's mostly too vague at this point anyway. I'm generally pro-privacy so this sounds a good idea in principle.

The main section I have thoughts on is the first, the "Universal Service Network". This goes considerably beyond current UK policies on broadband, and included a pledge for £25bn state investment in a "public sector backbone" to help deliver high speed broadband and mobile connectivity, everywhere in the UK from inner cities to remote Scottish islands. Not only that, but the manifesto commits Labour to ensuring that access is available "at the same low price without any data transfer cap". 

The relevant part of the speech itself was mostly about "equality" of network coverage. But implied in that statement is not just coverage, but also price controls and - essentially - infinite capacity. Leaving aside the competitive niceties of forcing identical pricing from multiple providers, the "no caps" promise is essentially unattainable, especially coupled to another seeming promise for (hard-ish) Net Neutrality. A reference to South Korea and its enviable broadband infrastructure was fair, but didn't attempt to explain why it is different to the UK (eg urban population density and less planning constraints).

I asked a question about whether Corbyn would be willing to relax planning regulations for cell towers, or rights-of-way for fibre installation, in order to fulfill this desire for ubiquity and unlimited capacity. He didn't answer personally, instead allowing Pascoe to address it. She didn't mention the cell-site problem and instead claimed that FTTH was mostly an unattainable goal (don't tell the Koreans) and that Google was developing wireless technology that could help fix things. 

However, FTTH everywhere (although a popular political topic) wasn't really my point. It was more about being able to put cell-sites everywhere (with fibre backhaul, including to small-cell locations) which is a huge practical constraint and cost for mobile operators, and the ability to put fibre elsewhere for trunk connections - and perhaps financial incentives for doing so.

Someone else asked about ownership, and the relationship with telco networks. Corbyn's answer was vague, but certainly didn't discount the possibility of wanting to re-nationise parts of the infrastructure. Given his recent spat with Richard Branson about trains, I wonder if Virgin Media is in his sights too.

Nobody mentioned spectrum, at any point. Or how the new promises compare to existing UK government efforts to push mobile coverage. Or the questionable "success" of other national broadband infrastructure projects (eg Australia's NBN) or the practical limits of "local access cooperatives", especially when it comes to cellular networks.

Overall, I thought that the network part of the manifesto was pretty weak. Yes, rural areas need better networks, as do train lines. But it's hardly as if this hasn't been a focus in the past. But Labour's team doesn't - at first sight - seem to understand that both coverage and capacity incur costs. It is also unclear in explaining how planning rules and competition might fit with any government assistance. The role of state- or metropolitan-owned networks was not detailed, and the costing sounds ambitious/unrealistic when one considers the need for both remote regions and (presumably) in-building networks, plus vast increments to existing network capacities to satisfy a "no caps" pledge. Also, ongoing operations of any network would raise the cost & future commitment to expenditure much further. I suspect throwing fibre under the bus, in the hope of some future wireless tech alternative would have pleased the 5G lobby, had it actually been mentioned instead of some unspecific wireless innovation by Google. I'm assuming they don't expect Loon balloons over the Highlands any time soon.

The other area I asked a question about was open-source. Although social-media and press comment ahead of the event suggested that Corbyn would insist all "publicly funded" software and hardware be open-source, this was considerably toned down at the event. The outcome was that a future Labour government would have a "bias" towards open-source "where possible", and that government contributors to O-S projects would be be rewarded. 

I inquired how all this would fit with the ubiquity of proprietary software in things the government bought (say software in ambulance engines, x-ray machines... or political parties' campaign-management systems). I also noted the government supports/funds proprietary software through R&D programmes, support for the games industry, encouragement of IoT and so forth. The "where possible it should be open-source" seems like generic IT "activism" rather than analysis.

My takeout was that (like the network bit) none of this has been thought through properly. There seemed to be a general dislike of big (mostly US) software companies like Microsoft, but little awareness of how pervasive software is elsewhere. I'll be interested to see if all the software running the promised "public backbone" network is proposed as open-source too.

Perhaps Corbyn's first move should be to guarantee that all of the Labour Party's own software & hardware moves to open-source first. Judging by proprietary software experience required on its current open jobs (link) & previous ones (link) it's got a long way to go.

By contrast to applauding free software (especially when implemented by government), the manifesto very much wants people to pay for music and other creative output. There were promises to re-write copyright law to ensure that cash flows to the right people in the entertainment space. I got the sense that nobody had really recognised the ever-blurring boundaries between software, hardware, cloud, content, networks and the implied inconsistencies in the various manifesto pledges arising from this.


Other elements of the "digital democracy" pledges were:

- Platform Cooperatives, for which read "nationalised versions of Uber & Airbnb & TaskRabbit", plus pledges that anyone earning "most or some" money from "digital platforms" should be able to unionise and have an employment contract". (It was unclear whether the robots & AIs which would inevitably take larger roles would have the same rights as humans).
- Digital Citizen Passport, for which read "ID cards by the back door, but voluntary opt-in only and with lots of controls for privacy & who gets access to what data, honest!". It's unclear how this fits with the existing Gov.UK Verify project too.
- Community Media Freedom, which is a hotchpotch of things aimed mostly at the media industry, including education programmes for "analysing and making media". I look foward to diplomas in advanced trolling, speaking slots on RT & Iranian news TV, and YouTube/SnapChat editing. More bizarrely, this section also says "Ofcom will protect network neutrality from discrimination between data streams and manipulation of software algorithms for private gain". If generalised, the latter half of that sentence pretty much kills most of the businesses on the planet, given that software algorithms are in everything. (And yes, I know that Neutrality in its strictest sense doesn't necessarily "work", but various principles are realistic to implement, as BEREC has shown today - link)
- Massive Multi-Person Online Deliberation - this is actually potentially cool, with more participation from people in politics via web/apps/whatever to help design legislation. However there are lots of forums for this already (notably Twitter and Facebook) and the key problems are around partisan groupings and abuse/trolling. I'm not too sure about the Orwellian overtones of the this newspeak though: "The National Education Service will enlighten the British electorate with the theoretical knowledge and practical skills of digital citizenship". I'd rather get my practical skills about being a citizen from someone other than the people who want me to vote for them, thanks.

Overall, it was mostly standard "digital engagement & inclusion" waffle you'd expect from politicians - some of which makes sense but is largely happening anyway - plus lots of incoherent digs at private businesses "we hate Microsoft - except for Skype" and "we hate Google - unless it does wireless networks" and "we hate Uber - so we'll do it ourselves". 

There was very little on "proper technology" and little awareness of the realities of networks, software development, IoT, Internet architecture or existing initiatives.The pledges for uncapped and ubiquitous broadband look utterly unrealistic - even with a one-off investment of £25bn.

What was also missing from the event was the presence of anyone else in Labour who actually acts as a shadow minister for telecoms, IT and the so-called "digital" economy. Corbyn uses social media but clearly isn't a technologist himself - so is reliant on the people around him to fill in the details, which is OK in principle. But the other speakers at the event: Pascoe, an associate with a dubious history (link), and someone who runs campaigning for Labour's Momentum supporters' fan-club weren't exactly deep techies either, especially on networking issues. It's all very well addressing the "social" side of the Internet - employment by Uber, musicians' rights and "massive on-line deliberation" but unless it's underpinned with proper understanding of how networks and software and hardware and IoT work, it's just fluff.

Monday, August 22, 2016

TelcoFuturism: Initial thoughts on Blockchain

An area I'm currently researching, as part of my ongoing analysis of telecoms/futurism intersections is Blockchain. (See here for an intro to what I mean by telcofuturism)

For the uninitiated, Blockchain (abbreviated here to BC) is the technology underpinning Bitcoin and other cryptocurrencies. It's a way to create distributed, secure, unchangeable, peer-to-peer databases for "trust" and transactions/applications which require it. It removes the need for central coordination and storage to prove that you own/bought/sold/transferred things, and stops the "double-spend" problem of digital copying of things like e-money. This is potentially great for finance, where a lot of cumbersome back-office processes could be made hugely more efficient.

When more extensive definitions of "trust" are considered (eg authentication of documents or relationships), it potentially has the ability to dis-intermediate all sorts of other existing businesses and even government functions, beyond just banking and digital money. There are tons of books, conferences and think-pieces about BC, from everyone from FinTech disruptors to governments to major IT and auditing companies. It's definitely a "thing" at the moment.

However, it is debatable whether some of the more far-fetched concepts presented are truly visionary - or sci-fi hype peddled by people who are less-than-objective wishful thinkers. It could become as important and ubiquitous as electricity, semiconductors or the Internet - or else it could just be an interesting platform for diverse applications, but not really a global "game-changer". 

An historian from the year 2100 might point to Blockchain as the most pivotal enabler of the restructuring of global business and society - or else it might be a minor footnote to the much-larger impact of other innovations around AI, CRISPR gene-editing and nanotechnology.

I'm trying to look at Blockchain through the lens of telecoms, networks, communications and cloud platforms. I can't really comment on the full impact on banking or manufacturing or property markets.... but I think I have an idea of the practicalities for telcos to deploy blockchains, and also the realities of networks as they might apply to other use-cases.

I haven't yet reached firm conclusions about the most important use-cases for blockchains in telcos & other communications infrastructure, or the probable timelines, but I'm starting to develop some initial hypotheses. There's some good arguments about BC's use in billing systems, IoT/network registration and control, vertical-market services in finance and healthcare, and perhaps integral network/OSS functions. I can also see it dovetailing with eSIM, cloud/PaaS platforms and numerous other niches in telcos, enterprise comms/UC domains and beyond.

I'm cautiously positive about the technology, rather being than a full-on religious convert and evangelist, as some Blockchain advocates seem to be. I don't buy into the notion that it's going to magically remove all intermediaries from all areas of human interaction, and lead to some anti-capitalist utopia/dystopia (delete according to taste) where middlemen no longer have roles to play.

I see a few major problem areas and "gotchas" emerging, that lie between the vision and possible reality, especially in the medium term:
  • Often, blockchain is suggested as the "missing piece of the puzzle", after which a new low-friction process, or entire new industry can be born. Yet in many cases, it isn't transaction cost, or cumbersome trust arrangements that are the "gating factor" stopping deployment adoption today. There are other practicalities involved too - perhaps regulation, business model, customer preferences and loyalties, or 100 other factors. For instance, the idea that everything to do with IoT just needs a sprinkling of Blockchain pixie-dust, for trillions of dollars of value to be released, on interoperable open-source style platforms, is pure hyperbole. It might be desirable - even necessary - but it's certainly not sufficient for many things to take off.
  • A fair amount of envisaged blockchain use-cases require perfect, ubiquitous connectivity. That might be OK for banks and fintech companies using multiple data-centres and redundant fibre links, but it doesn't work well for mobile/wireless which is not going to be ubiquitous any time soon (if ever). Unless the applications have some way of dealing with "offline mode", or patchy/intermittent connections, that's a major obstacle.
  • Some blockchain architectures have significant time-lags involved, due to processing for verification and permanent storage/encryption ("mining" etc.) That's fine for things which operate on a scale of minutes/hours/days - say transfers of property deeds - but not ideal for network operations that involve subs-second decisions. As we move towards 5G and "millisecond latency" critical applications, this becomes even more imperative.
  • Blockchain applications will need to "play nicely" with all sorts of other technology trends, such as distributed clouds, telecoms NFV/SDN architectures, third-party PaaS, integration with legacy systems, security gateways, policy-managed networks (can they spot, block or prioritise BC data flows?), AI and machine learning creating unpredictable or novel transactions/interactions and plenty more.
  • Many suggested blockchain use-cases ignore what might be termed "immovable obstacles". It's all very well having a BC-based wireless mesh network, but if it's ignored by the companies owning big chunks of licenced spectrum, and creating non-BC back-office functions, it's a bit of a waste of time. The same thing applies to regulations, taxation and assorted other slow-moving areas of bureaucracy. It's all very well suggesting that your house can act as an autonomous business, renting out the WiFi to passers-by all by itself when you're out, and using the payments to pay the utility bill - but that my well cause consternation among people who tax and regulate such things. Other ideas - such as micropayments for IoT sensors selling weather data by themselves - sound great until people realise that the billing systems only support 2 decimal places, or ask the user to click "OK" or answer a captcha. There are many, many devils in the detail.
  • A lot of the rhetoric seems to suggest that everyone wants a completely peer-to-peer, decentralised, no-intermediary, no-brand economy. However the evidence seems to suggest that humans actually quite like intermediaries for many things and are prepared to pay for them - Apple running an appstore, curators for a museum, editors and brand for a news service and so on. Add in the clear need for designers as the new uber-class of intermediaries, and the over-riding importance of UX in any situation, and the "fully automated world" seems even less plausible.
None of this means that blockchain based services are a bad idea, or irrelevant to telcos and network/software vendors. There are, undoubtedly, many important and possibly huge opportunities in blockchain-based telcofuturism.

But at the same time there is a lot of hype. Outside of financial services, we're still mostly at the napkin-diagram/Powerpoint/very-early prototype stage of telecom/BC use-cases. I'm hoping to get some more clarity over the next few weeks - and will try to assemble a realistic timeline that blends vision and pragmatism for comms and network applications.

Please get in touch with me if you'd like to discuss Telecoms + Blockchain combinations. I can be reached via information AT disruptive-analysis DOT com, or via Twitter or LinkedIn.

Wednesday, August 03, 2016

NEW: eSIM Status and Forecast report published

Beyond M2M: eSIM Status & Forecasts
Overcoming practical & economic issues for mid-term consumer-market eSIM adoption


Disruptive Analysis has published a 36-page report on the emerging technology of eSIM and SIM remote-provisioning. The focus is on the use-cases, practicalities, drivers and obstacles for bringing eSIM-based devices to market, alongside suitable mobile data plans or subscriptions.

The report addresses both the motivations (lower costs, higher revenues, better experience) and problems (business-case, user journey, regulation, transition) that will be experienced by operators (MNOs) and device vendors (OEMs).

Forecasts are given for annual shipments of eSIM-enabled devices (phones, wearables, M2M, tablets), and for the installed base that will be a target for after-market eSIM provisioning.

Key findings:
  • There are numerous use-cases for “remote provisioning” of SIMs with mobile operator “profiles”, especially where the SIM hardware is built-into devices
  • eSIM adoption will have a slow start. 2016-17 consumer deployment will mostly be early concepts, allowing MNOs and OEMs to gain practical eSIM experience and refine implementation and processes. eSIM phones will emerge very gradually.
  • Adoption should ramp up in 2019-2021 as cost, industry value-chain and user-experience problems are progressively solved.
  • Apple and Samsung are unlikely to use eSIM to become MVNOs / carriers. Neither will they aggressively push eSIM into their flagship products.
  • For many M2M/IoT devices, the eSIM decision is secondary to justifying the extra cost, space and power needs of the cellular radio itself. 
  • eSIM is "necessary but not sufficient" to drive adoption of cellular M2M. It is unlikely to change the competitive dynamics vs. LPWAN technologies like SigFox or LoRa.
  • There remain unanswered questions about regulation, customer-support and business model for eSIM. Although some projected cost-savings are attractive for operators, it is unclear that it will help OEMs generate extra revenues/loyalty. 
  • There will other approaches to remote provisioning beyond GSMA's vision of eSIM. Some OEMs may adopt proprietary versions, while standards-body ETSI is intending to develop specifications which go beyond just mobile use of chip-cards 
  • By 2021, 630m mobile & IoT devices will ship with embedded SIMs annually, driven mostly by smartphones, although vehicles and tablets show growth earlier.
  • By end-2021, the installed base of eSIM-enabled devices will exceed 1 billion 
  • While significant, this only represents around 10% of total cellular connections
In a nutshell: eSIM is an important evolution for some use-cases, but it is neither an outright "game-changer" nor a major risk to traditional cellular business models.


To purchase the report, see below



Report Contents

Executive Summary
Introduction & Outline
   The Potential
   What is eSIM / eUICC?
   New uses for eSIM & other programmable-SIM technologies
   A device-centric view of SIM provisioning             
   A growing variety of “SIM evolution” options
The Practicalities             
   Economics and demand
   SIM/eSIM irrelevant if radio module costs too high          
   Operational issues          
   User experience              
   Retail and channel management              
   Maintenance and lifecycle-management               
   Security               
   Transition issues: the need for hybrid SIM + eSIM devices             
   Regulatory considerations           
   Ecological considerations: fit with other telecoms trends
The Phones        
   Low-end vs. high-end phones
   Apple-specific considerations
   Conclusions and Forecasts          
   Forecasts            
About Disruptive Analysis            

Figure 1: Understanding the definition & semantics of “eSIM”     
Figure 2: Advantages of “programmability” vs. “embeddability” varies by device 
Figure 3: SIM evolution – multiple variants are emerging, not just GSMA eSIM     
Figure 4: SIM evolution – costs and key stakeholders       
Figure 5: Few handsets’ gross margins can sustain extra BoM cost from eSIM       
Table 6: Forecast eSIM shipments, by device category, 2016-2021             
Figure 7: eSIM shipments, by device category, 2016-2021             
Figure 8: eSIM device shipments, hybrid SIM/eSIM vs. eSIM-only
Table 9: eSIM active installed base, by device category, 2016-2021           
Figure 10: eSIM installed base, by device category, 2016-2021     
Figure 11: Overall SIM & eSIM active installed base, end-2021     
 

Ordering & payment


The report (delivered as a PDF) costs:
  • US$900 for a 1-3 user licence
  • US$1500 for a corporate-wide licence + a free 1-hour conference-call discussion
  • (plus VAT in UK/EU as appropriate)

Payment is via credit-card and Paypal (see below), or where a purchase-order and invoicing details are submitted by email to information at disruptive-analysis dot com. The report will be emailed to you within 24 hours of receipt of payment.

[Note: Sometimes Paypal's credit-card transaction process is a little variable, especially with corporate cards. Please drop me an email if you have problems]

eSIM Report, 1-3 users




eSIM Report, Corporate