2015 End of Year roundup
[avatar user=”malm” size=”small” align=”left” link=”file” /]
At the end of 2014 I posted my themes for that year which covered Android, Wearables and the Internet of Things, Smart Retail, Automation and Inequality. For the last roundup of 2015, I thought I’d look back at some of the key themes covered in the blog over the past year. Looking ahead, it seems they will all remain very much at the forefront of 2016’s links namely: i) Artificial Intelligence, ii) Machine Learning, iii) Uberisation, iv) Digitalisation, iv) Health Wearables:
- Artificial Intelligence has become a dominant theme in 2015 and will remain firmly in the tech zeitgeist as well as mainstream in 2016. Celebrated AI thinker Calum Chance has outlined his predictions for 2016 which include a DeepMind (of which more later) AI version of Go capable of beating the best human as well as VR becoming a very big deal.
- Machine Learning is widely seen as a specific aspect of AI focussing on self-improving algorithms. The algorithms that really sparked mainstream imagination in 2015 covered Deep Learning which are built around multiple hidden layer neural networks. The growing applicability of Machine Learning and its close relative predictive analytics within the corporate environment together with growing concern over technological unemployment will remain key themes in 2016. It helps explain the desperate rush by many leading companies to get machine learning competence on board.
- Uberization and its shadow has loomed large over 2015 both positive and negative. They are mentioned in two of the five tech predictions Kevin Rose makes for 2016 and their influence on the tech industry will remain a key theme. In particular the seamless fusing of human and machine-driven intelligence in the form of HAAI (about which more later) will be a key 2016 development.
- Digitalisation of the Enterprise has become an ever more urgent issue across every industry vertical. McKinsey have tracked progress over the year and will remain a key source of insight in 2016 too. Their paper of the year covers raising corporate Digital Quotient (DQ) and why that is important. This graph provides a neat summary:
- Health wearables have gone from strength to strength in 2015 notably Fitbit who remain the undisputed leader of the field following their successful mid-year IPO, Apple Watch notwithstanding. They ended the year at the top of the app charts ensuring they had a bumper Christmas to round their best ever year off. Where they go next with their proposition will be a key tech story in 2016.
To finish off on a personal note, I updated and re-ran my Docker instance holding my end of year word count for my weekly posts across the year and here it is. Once again, if you’re interested, the source code that generated this and the other graphs is available here:
Manufacturers and Devices
- A detailed initial review of the Qiku Q Terra phone is in from PhoneRadar and it’s very good. The reason this phone is a bigger deal than most coming out of China is the impressive partnership behind it. Ultimately though, for PhoneRadar the Q Terra score is mainly down to camera and display:
QiKU [is] a brand born with the partnership of Qihoo 360 and Coolpad. [You] could say this is an ideal collaboration as former is the global internet giant with expertise’s (sic) in software technologies while the latter is a master in smartphone manufacturing.
- Samsung doesn’t ‘get’ software but is at least trying to change things according to this fascinating Reuters piece which suggests it is in a very similar place (albeit with Android rather than Symbian) to Nokia when they attempted their infamous and ultimately unsuccessful switch from being a hardware-first to services-first brand. It’s just as unclear whether they will make it:
“There’s a lot of distrust of top executives who can actually implement stuff that is more of a software and services offering,” said one person familiar with the company’s inner workings. “It’s still ‘we know how to sell boxes, we sell boxes’.” … Growth in handset sales is slowing as the smartphone market matures, and without its own distinctive software, content and services, Samsung has little to differentiate itself from other Android phone makers selling similar devices at lower prices.
- Tomi Ahonen wrote another monster “state of the nation” post on why mobile is the nexus of the new global economy and represents a trillion dollar opportunity that will affect you “for the rest of your life” so you should make it your New Year Resolution to develop a mobile strategy now. It’s entertaining (if long) reading. David Doherty at mHealthInsight published a handy infographic summarising the main points. And Ewan Spence linked to it in a Forbes post likewise suggesting the ‘secrets of success’ in 2016 lie in the smartphone. He references a New Years Day post that really struck a chord with me – I too got zero Happy New Year SMS:
How many SMS did you get last night? Zero? … This used to be the most profitable night of the year for the mobile operators globally. (SMS costs almost zero) … Keep ahead of the game and don’t let yourself become irrelevant in 16
Apps and Services
- In an entirely predictable development, a Chinese anti-terrorism law has now been introduced that ‘weakens encryption’:
If you were hoping that you could regularly rely on encrypted messaging services to keep your discussions private while you’re in China, you’re about to be disappointed. The country has passed an anti-terrorism law that requires companies to hand over encryption keys when officials want to spy on someone’s communications.
- A couple of weeks ago, the blog linked a few posts examining the rise of conversational UIs powered by HAAI (Human Assisted Artificial Intelligence) suggesting it was a key growth area for messaging apps over the near term future. Now it’s been suggested that Google have been working on their own one ‘for a year’ presumably aimed at countering Facebook M. As predicted above, HAAI will be a key battleground in 2016:
with both of these companies working to create messaging apps that don’t restrict people to communicating with other humans, the combined force could help messaging services become the central hubs of consumers’ digital lives.
- In the ultimate vindication of the view that OSS is merely a tool that can be applied to emancipate or enslave in equal measure depending on context, researchers have outlined how a version of Linux developed in North Korea has a range of technical feature used to suppress freedom. And in Iran, OSS technology is being used to build an increasingly sophisticated protectionist ecosystem increasingly separate from the rest of the global Internet in a manner akin to China:
In regard to social media, Iran has been moving in two parallel directions. It has mastered data-mining techniques, enabling it to find potential troublemakers who use the web as a tool for stirring political unrest. This expertise has proven effective, as attested by the tracing and capturing of several cells of online political activists in past years. Iran has also established several organizations to produce online content and fight the political opposition’s activities on the web.
- Meanwhile in Bolivia, OSS is at the core of efforts by some to secure its digital sovereignty from US technology giants.
- Blockchain technology seems poised for a breakthrough 2016 within the digital enterprise. Money transfer, auditing and supply chain management are just a few of the many potential areas of application:
An example of a startup here is Thingchain, which is applying a bitcoin-inspired cryptosystem to multiple use cases, including proving the provenance of goods and who owns them.
Consciousness and Artificial Intelligence
- Admittedly impenetrable in large parts, this research paper on consciousness from Max Tegmark at MIT is nevertheless worth checking out. In it he outlines his research on consciousness as a state of matter distinct from solid, liquid and gas. He calls it ‘perceptronium’ with ‘distinctive information processing’ abilities including the ‘dynamic integration of information’.
The key assumption in this paper is that consciousness is a property of certain physical systems, with no “secret sauce” or non-physical elements. Instead of starting with the hard problem of why an arrangement of particles can feel conscious, we will start with the hard fact that some arrangement of particles (such as your brain) do feel conscious while others (such as your pillow) do not, and ask what properties of the particle arrangement make the difference
- At the other end of the spectrum, John Searle’s Google Tech Talk on consciousness in AI suggests such a thing isn’t remotely possible today. Searle is famed for his ‘Chinese room’ concept against strong AI which posits that ‘simulation is not duplication’. Namely that a system capable of manipulating Chinese symbols to a degree can pass the Turing Test from an external observer’s perspective using syntactic manipulation (code) without actually having any semantic context whatsoever as to what is going on. Searle’s view is that consciousness is ‘intrinsic’ in the human brain in an ‘ontologically subjective’ way that we simply don’t understand today. It is therefore impossible to say today whether we could duplicate what we mean by ‘thinking’ in another material. There’s an interesting exchange 38 minutes in with Google’s Ray Kurzweil where Searle essentially contrasts the messy behaviour of brain synapses under chemical stimulants (specifically cocaine) with the relatively perfectly well-behaved sigmoid activation function in a syntactic neural network:
- A great Nautilus read from earlier in 2015 on Walter Pitts, who rose from a tough upbringing to a place at the top of the mid 20th century MIT Cybernetic group along with the likes of John von Neumann and Norman Weiner. He was driven by a desire to build a universal model of symbolic abstraction. In his own lifetime, his views fell out of fashion and he died in obscure circumstances at the tail end of the 1960’s. However his work arguably underlay future subsequent developments in neural networks:
there was a moment in history when neuroscience, psychiatry, computer science, mathematical logic, and artificial intelligence were all one thing, following an idea first glimpsed by Leibniz—that man, machine, number, and mind all use information as a universal currency. What appeared on the surface to be very different ingredients of the world—hunks of metal, lumps of gray matter, scratches of ink on a page—were profoundly interchangeable.
- Quartz published this Quora response from an evolutionary biologist on why Elon Musk is right about the threat of AI but very wrong with his suggested remedy of providing “open access to machine learning technologies“. Today the most advanced forms of the tech are largely confined to relatively few commercial organisations. By opening out the field, we threaten to escalate their Darwinian selection processes beyond our control:
If a small number of ruling AIs were to attain world domination, it’s just possible they might be able to coordinate to avoid mutual destruction. But when millions of AIs are fighting for market share, the coordination problems involved in long-term sustainability become intractable.
- One leading example of a commercial organisation working at the forefront of AI is Google’s DeepMind team. This recent public lecture from Demis Hassabis provides an approachable introduction to the current state of the art on artificial general intelligence (AGI). There is some great footage in here of the recent general system DeepMind built to beat a variety of Atari games and some revealing insights at the end – notably that DeepMind remains an entirely UK output and that they have some ‘very big announcements to make on Go’ in 2016:
Wearables and the Internet of Things
- Independent Kickstarter fulfillment report suggests that 9% of all projects on average fail with a distinct skew by raised amount:
- More adventures with Raspberry Pi and OpenCV from Adrian Rosebrock. This time he looks at how to increase the frames per second (FPS) from the Raspberry Pi camera basically by using threads.
- Nice short introduction to writing RESTful Python client interfaces using Python requests together with a few other tricks.
- Class Central’s survey of the state of MOOCs in 2015 suggests that although they remain in a period of exponential growth, business models are undergoing revision with free certificates being phased out and self-paced courses becoming more popular. Subject choices remain widely dispersed though over half of the total courses remain STEM-based.
- HBR post suggesting only a very small number of leaders good at both execution and strategy:
- HBR again this time on why Mindfulness is an important tool for calming your brain during workplace conflict:
Mindfulness is the perfect awareness technique to employ when a conflict arises — whether it’s at work or home. It allows us to override the conditioned nervous system with conscious awareness. Instead of attacking or recoiling, and later justifying our reactions, we can learn to stay present, participate in regulating our own nervous system, and eventually, develop new, more free and helpful ways of interacting.
- Salutary tale from the protected and messy endgame for Good Technology prior to its acquisition by Blackberry in September that when a unicorn startup stumbles, it’s the employees and not the executives that get to reap what is left of the rewards:
- Here’s how unequal rewards for being ‘in it together’ worked out for a couple of monkeys:
- Closely linked is this Nature research piece entitled simply, “Hierarchy is Detrimental for Human Cooperation“. It is likewise built on primate studies presumably similar to the above:
Compared to a condition lacking hierarchy, cooperation declined in the presence of a hierarchy due to a decrease in investment by lower ranked individuals. Furthermore, hierarchy was detrimental to cooperation regardless of whether it was earned or arbitrary.
Society and Culture
- Reflective post on what seems to be the growing toxicity of much online debate suggests that dealing with the issue of ‘why can’t everyone be nice online?’ requires us all to play a role. It’s important for sure but how effective it is depends in some part on whether you view concepts like Godwin’s Law as fundamental:
the real reason we can’t have nice things online–or anywhere–is because there is less distinction between us and them than many of us would care to admit. … We are all responsible for what the internet becomes. We are all responsible for how we act, what we say, the content we choose to amplify, and the content we choose to ignore.
- Quartz profile the humble origins and history of Louis Vuitton now one of the world’s most iconic and copied luxury brands.
- Demonstrating an ability to adapt with the times, their 2016 collection is being modelled by a Final Fantasy character:
- This post explains the politics of Star Wars through the connections between the New Republic, the Resistance and the New Order. It also makes a sobering and depressing point that the supposedly ‘decisive’ victory achieved at Endor (Return of the Jedi) largely unravelled as a result of in-fighting and disorder within 30 years.