2016 Posts

Below are links to all the 2016 weekly digest roundup posts grouped with high level seasonal themes along with summary highlights from each.


  • Week 46 (27.12.16): Developers as the new kingmakers, automation and the US economy, machine learning using TensorFlow, the excellent fast.ai practical deep learning MOOC, serverless, cloud, snowflakes, SSIDs and privacy, fear and uncertainty, 2016 takes another pop star in George Michael, carpe diem.
  • Week 45 (20.12.16): The problem of reasoning from induction, technological unemployment in the Age of Anger, neuroprosthetics, Mark Zuckerberg’s attempt to build Jarvis, deep learning, the top Python libraries of 2016, mobile is eating the world, the fall of Nokia and ageism in tech.
  • Week 44 (12.12.16): Edge computing, connectedness and the lesson from the past in Symbian OS, the AI awakening in big tech, where Alexa goes next, Amazon Video, AWS London, the challenge of digital transformation in a bank, Python web frameworks, side projects, tech company organisational structure, why there is no difference between online and offline and the rise of ressentiment in the age of anger.
  • Week 43 (04.12.16): AntiSocial media, Trump as Leviathan, AWS chooses MxNet for deep learning, prospects for voice control, private cloud considered pointless, Wardley on corporate strategy, why digital media agencies struggle with software, a Python library called curio, Python for data science, MOOC survey for 2016.
  • Week 42 (27.11.16):  The dangers in information imperialism, the leadership challenges that face the new populists, machine learning hype and reality, how Facebook and Google are remaking themselves into AI companies, the rise of system-on-chip computing, why Huawei are on top of the Android ecosystem, why serverless is the next big cloud paradigm shift, some notes on JavaScript programming and whether the universe itself is an alien.
  • Week 41 (18.11.16): What it means to live in a post-truth world, why resource depletion cannot be spun, Donald Trump and Prospect Theory, was Facebook complicit in his election, the adventures of an artist that uses machine learning, the remarkable momentum behind TensorFlow, Dunning Kruger redux, Hipsterism in Hoxton and Shoreditch and a sobering historical perspective on intolerance past from Reading Gaol.
  • Week 40 (09.11.16): The ramifications of a Donald Trump presidency and the dual role of technology as both an enabler for his campaign and as a bogeyman for disaffected blue collar workers plus the implications for the US tech industry, an Alexa fish, WeChat instant apps, a searing Black Mirror masterpiece called “Shut Up and Dance”, Amazon in China, Docker and Terraform, the decline of Xiaomi, Apple’s GPU,  OKRs, why most interviews are “useless”, a short intelligence test, and learning in the era of digital distraction.


  • Week 39 (31.10.16): Is democracy is in danger from the rise of authoritarianism and how technology may aid its cause, Moby on smartphone distraction, Bank of America’s AI chatbot, MonkeyLearn’s Tarsier tool for sentiment analysis of the US election, why Nokia failed, evidence that dating apps have hit the trough of disillusionment, what Agile is not, the Mirai botnet and ramifications for internet security, Dyson’s puzzle and behavioural interviewing, Brexit tea, jam and biscuits, and how consultancy firms are making money from austerity.
  • Week 38 (21.10.16): Signs from politicians in the US and UK that AI has arrived and notes of caution about it, DeepMind’s differential neural computers, gradient descent primer, serverless and function as a service or “FaaS”, data analytics at scale, the parable of the WiFi kettle, Samsung’s Nokia moment, Apple’s FPGA, the reality of Shanzai copy products in China, big Agile, software over-engineering, tech hype, Donald Trump’s ‘garbage heap’ soul and the growing sense that there isn’t anybody driving the bus on the UK’s journey out of the EU.
  • Week 37 (12.10.16)Google’s Pixel phone, Samsung acquire Viv, how the dead could live on via a chatbot, Nick Bostrom at IP Expo, the difficulty of programming a moral framework for AIs, brain emulation, Reinforcement Learning, data-led startups, an IoT launched DDOS, the importance of testing error handling, a Python visualisation library called Altair, the crazy world of JavaScript in 2016, engineering at Spotify, the Age of Tech, why being a Change Agent is tough,  kinetic art, modelling human memory,  Donald Trump’s disastrous week and some views on Brexit, foreign workers, zealotry and the UK’s unclear journey out of the EU.
  • Week 36 (04.10.16): Distraction sickness, how deep learning is pushing the boundaries of language translation, technological unemployment in the construction industry, Splunk 6.5 update, developer-friendly deep learning with Bonsai, Amazon Alexa update, software-defined cars, the rise and fall of Blackberry, three infrastructure mistakes you cannot afford to make, the best intro programming MOOCs, why SpaceX goes to Mars, the collapsing cost of solar power, the curious thing about leisure time, work and the Stupidity Paradox and counterfactual history with Donald Trump.
  • Week 35 (23.09.16): Youth versus experience revisited courtesy of Wardley Mapping, Acid House, the neural network zoo, Google Allo and privacy, the Stingray fake basestation, the consumerisation of IT, Android N, dotEverything physical push notifications, Go, the 2016 Python UK Conference and Bluetooth LE fun with a smart LED lamp, cognitive bias classification, more on unconscious bias, contrasting work cultures, a UK tech talent map and startups and shutdowns.
  • Week 34 (17.09.16): Dyson and his comments on youth vs. experience, black box models and the coming privacy catastrophe, Amazon Echo and push notifications, Amazon employee no. 1, Baidu deep learning framework, DeepMind’s WaveNet speech generator, technological unemployment and artisanal gig jobs, the Internet as force of polarisation and distraction, the NoPhone Air, kitchen sink development, unconscious bias, the state of Javascript frameworks, 20 Python libraries you don’t know but might want to, sea changes in linguistics and the lost world of 9-10 fifteen years on.
  • Week 33 (09.09.16)Apple Keynote and iPhone7 announcement, the Programmable Neon Kickstarter, why Amazon’s Alexa platform could be a major player in the post-mobile space, consumer IoT, the death of Project Ara, Deep Learning with Keras, product growth champions and elastic product teams, Amazon Drive, push notifications, algorithmic hiring, Trump again, Jeremy Corbyn and UB40.
  • Week 32 (31.08.16): The rise of ‘dataism’ and a third sphere of algorithmic culture, some pointers to the current status of deep learning-based speech recognition for chatbots and language translation, how Apple do AI, deep learning for text summarization, the organisational impact of shifting to cloud, a BBC Click special called Designed in China, dual-screen Android phones, the Programmer Olympics, Proxima b and the Fermi Paradox, Donald Trump and the Coen Brothers and the “death” of neoliberalism.
  • Week 31 (22.08.16):  The influence of Facebook and Are ‘Friends’ Electric?  the role of the human in AI, neuroprosthetics, deep learning cars, more on Google’s mysterious Fuchsia operating system, connected home opportunities, smartphone cameras are ‘eyes that can see’, odrive, customer journey analysis, the Olympics, why Brexit means Brexit but not much else right now plus some links on cultures of crisis.
  • Week 30 (15.08.16):  The nature of tech disruption, more Keras and Deep Learning using OpenAI Gym, connected cars and driver privacy, LIDAR on a chip, smart bulb security holes, Google’s Fuchsia, China as the Galapagos of tech innovation, Xiaomi seems to be crashing to earth while LeEco is rising, dealing with legacy software, agile cautionary tales, a couple of useful tools called httpie and smolder, Donald Trump’s “killing joke” and the political echo chamber of social media.
  • Week 29 (08.08.16): Perception and reality as advances in display technology further blur distinctions between the physical and real world, more steps with convolutional neural networks using OpenCV and Keras, a lightweight web framework called Chalice, Amazon’s dominance in IaaS, Samsung smartphone development insights, tracking users by battery stats, 10000 years into the future, why humility is helpful to programmers, pull requests, Trump the quantum candidate, primitive technologyand the world of Preppers.
  • Week 28 (01.08.16):  Truth and lies in the world of post-factual politics, advice on how to approach Machine Learning from a competition expert, a couple of TensorFlow tutorials, introduction to image processing with convolutions, mistakes managers can make with analytics, the rise of HTTPS, Twitter’s problem, Dollar Shave Club, Amazon’s healthy profit, programming language popularity, has Agile lost its way? and the Joy of Data.
  • Week 27 (25.07.16): What entrenched elites and would-be kings can learn from history, why almost nothing Donald Trump says or does can stop his appeal to those with little to lose in disrupting the status quo,  the acquisition of ARM by Softbank, the importance of high quality training data in building a high quality machine learning model, using the deep learning Keras library to build an AI FlappyBird, creative TensorFlow, on digital anonymity, more on Pokémon, Prisma on Android and 8 Chinese Internet startups worth watching.


  • Week 26 (17.07.16): The role of technology in disrupting truth and facilitating global conflict, more analysis of possible Brexit impact on the UK economy, why the Labour Party seems doomed to split, fun and games with the MonkeyLearning text analysis service, GrowthBot, CTRL-F for the real world, Pokémon Go, Prisma, Silent Circle, Smart Home go home, why Amazon is becoming a subscription company and why the next Facebook will be an eCommerce company, tech founder hazards, Apollo 11 source code and a David Deutsch TED talk.
  • Week 25 (09.07.16): Brexit – attempting to understand why the country appears so divided, who the winners and losers are, what the result means for UK tech and digital services and a sentiment analysis of Brexit tweets.  The other week’s links include: technological unemployment portents and prospects, Deep Learning resources, chatbot state of the nation, Bob Cringely on Big Data and practical big data analytics using Splunk.
  • Week 24 (30.06.16): More Brexit preoccupations notably the bizarre post-vote political machinations better suited to Macbeth or Brideshead Revisited, more on why it happened and what happens next.  The other week’s links include: why it’s hard to predict how AI will eventually be embedded in consumer tech, the importance of Siri, implementing an RNN using Tensorflow, what Hillary Clinton thinks about technological unemployment, Android Nougat, WileyFox Spark and “quantum team management”.
  • Week 23 (24.06.16): UK voting for Brexit and the country changing overnight.  Will the “radical + bold” UK2.0 include a paid individual opt-in subscription model for EU membership for the millions who voted for Remain?   The other week’s links include: Trump, technological unemployment is coming, more ‘machine learning is maths’ links, Jeff Bezos interview, WeChat, Scala as the ‘golden child’ of programming languages, the end of reflection and why shifting iPhone production to the US will add cost but not actually create any new jobs there.
  • Week 22 (16.06.16): More of an extended piece than usual focussed on the EU referendum and US election campaigns and reflections on the connections and differences between them.  The other week’s links include: the brain either is or isn’t a computer, speech recognition state of play, Amazon Echo prospects, machine learning in the Enterprise, WWDC meh, Microsoft acquiring LinkedIn, the rise of app-pathy and of paranoia.
  • Week 21 (09.06.16): Why machine learning is more maths than computing, why your brain is not a computer, new virtual assistant propositions Ozlo and Pana, BladeRunner Redux, Sophia the robot, LoRA and IoT consortia, Mary Meeker’s Internet Trends, Amazon Echo, Twilio, Agile is a Mindset, non-conformism, the Trump University debacle and its pale reflection in comparison with the life of Muhammad Ali.
  • Week 20 (02.06.16): Elements of Machine Learning – spaCy and natural language processing, the advent of Enterprise AI, Apple to ‘open up Siri’, a Russian military robot, Android’s update lag problem, why WeChat success has been difficult to replicate outside of China, India’s space shuttle, passive WiFi, Trello, Dropbox, the downsides of mobile, travels in hyperreality, why software may change but maths is immutable, overwork is bad for you and some thoughts on Thiel vs Gawker, Trump, authoritarianism and Hitler comparisons.
  • Week 19 (25.05.16): Elements of Machine Learning – sentiment analysis, insights into Google announcements at IO, tensor processing units, the future of news, Nokia’s re-entry into mobile, Huawei ambitions, Intel culture, an Orwellian app called FindFace, extremist behaviour online, Ethereum, Pinipa, the cargo cult of excessive collaboration, gun violence in the US and a minimum wage machine.
  • Week 18 (18.05.16): Elements of Machine Learning – word vectors, AI poetry and the End of Code, more chatbot fever, transcription and translation updates, client side deep learning, the robotics revolution from a Japanese perspective, McKinsey on digital transformation and organisational structure, more IoT security fail, blockchain, paper IoT, a Raspberry Pi valve amp, Python decorators, the case for a 3-day week, more Trump darkness visible, sleeping, aging and the dying art of skywriting.
  • Week 17 (10.05.16): Elements of Machine Learning – automated classifier search with a Python module called TPOT, the “Uber for X” landscape, TensorFlow in the cloud, autoencoding and neural networks, how Apple Music ‘steals’ your music and what it means, Dyson investment, the Disunited Nations, troublemaker antipatterns, McKinsey on company-wide agile and digital transformation, developer salary analysis, how computer technology is being used by criminals, Leicester City’s incredible triumphant week, and the dark storm clouds swirling around the other big victory of the week which put Donald Trump one step closer to becoming US President.
  • Week 16 (03.05.16): Elements of Machine Learning – Regression, DeepMind moving to TensorFlow, word embedding, truck drivers and AI, the security risk of chatbot transcripts, SirinLabs, Fairphone’s OS, Will.i.am’s Dial smartwatch, a Raspberry Pi powered GSM base station, the rules of pricing software, the limitations of Swift in the browser, Hacking Team, a massive cyber bank heist, signs of decline for Apple, tech disillusion, on running, more Donald Trump, how the media makes mountains out of molehills, hipsters and Buddhism
  • Week 15 (25.04.16)The humans at the heart of the AI bot machine, the many challenges facing Facebook’s digital assistant M, an EU crackdown on Android, how big companies can make big mistakes, a Tensorflow playground, a tuhao with eight robotic maids, the spurious rigour of regular password changes, a shapeshifting display from MIT, the art of fixing broken things, DNA as a storage medium, content moderation, post-capitalism, HBR on embracing Agile, Elon Musk’s CV, why Generation X had the worst of the recession, the Centennial Minecraft Generation, snooker’s Steve Davis as an unlikely techno DJ and the sadly departed Prince as a guitar genius.
  • Week 14 (18.04.16):  The connection between serverless and IoT, Yann LeCun on the future of AI, a TensorFlow simplification library, AI and the bible, the state of long-form voice transcription, face detection camouflage, privacy and trust in IoT, Facebook and the future, Microsoft’s updated developer play, Tesla Autopilot miles, another rm -rf * hoax, Swift in the cloud, Julia, technical literacy and politics, the Fermi Paradox and aliens, Panama and how the 1% ‘corrupt democracy’, male bias found in Hollywood screenplays and whatever happened to the 70’s ideal of the Good Life.


  • Week 13 (11.04.16)Open and closed domain chatbots, X.ai’s Amy and Microsoft’s Tay, sklearn, application analytics engines, Nest shutdown of their Resolv home automation hub, a Scarlett Johansson lookalike robot, Tesla Model 3, the Mossack Fonseca breach and its wider impact on society, politics and IT security, MediaTek’s Helio chip, why learning to code won’t save your job, Joy Division in Python, laughter doesn’t scale, thoughts on courage and leadership.
  • Week 12 (03.04.16): Video games and machine learning, technological unemployment, the prospect of precrime, Netflix, LinkedIn gets ridiculed, Zucking up to China, why HTTPS is hard, Windows 10 will integrate Bash, Android Factory Reset Protection, the retro rise of dumbphones, shift register explanation, why the Tesla 3 launch is the moment of truth for electric vehicles, US unicorns are more likely than not to be founded by an immigrant, American manufacturing and runaway inequality and Donald Trump’s misogyny.
  • Week 11 (26.03.16)More thoughts on human dysfunction and artificial intelligence, the meaning of AlphaGo, universal basic income, Domino’s robot pizza delivery proposition, IoT security, how the dead live on Facebook, Fjord design trends 2016, AWS at 10, phone forensics specialists Cellebrite may be helping the FBI in lieu of Apple, hacking a Tesla S, why you shouldn’t use 1234 as a PIN code, real-time video capture and ‘re-enactment’, StackOverflow 2016 developer survey shows the field remains overwhelmingly male, patriarchal tech, the making of modern China, Powa run out of juice, the Simpsons and Donald Trump, death by gentrification in San Francisco.
  • Week 10 (18.03.16): AlphaGo wins 4-1, the path to the Singularity, New Order’s Singularity, why a Donald Trump tweetbot might be funny but he is no joke, Scott Adams moist robot, Google robot arms are learning from each other, Google Cloud Platform vs. AWS, Android N and Android Wear updates, chat messaging perspectives from Kik and Layer, the mobile phone and Blood and Earth, fooling a fingerprint sensor, IoT security shockers, we can’t all hire above average staff and the Guardian on the disenfranchisement of Millennials.
  • Week 9 (11.03.16): Why Deep Learning could melt the ice rink, AlphaGo in Seoul, DeepMind in the popular media again, an essential Rolling Stone two-parter on the AI revolution, why you may just be a 4000 feature vector in Facebook, technological unemployment and Donald Trump, Snapchat is confusing olds and courting the kids,  3D printing and the Chemical Brothers, holacracy growing pains at Zappos,  why it pays to look for the Third Door, more thoughts on why Donald Trump is winning, evidence that engineers are disproportionately more likely to engage in violent extremism, detecting terrorists from their V-signs using machine learning and the link between Brexit and immigration.
  • Week 8 (01.03.16):  Artificial Intelligence and humanity, indicators and signs everywhere that our AI future is already here, what to make of that last episode of The Brain, brain wearables, technological unemployment and the Basic Income argument, why we may all have reason to regret pushing Atlas around, MWC roundup, Mark Zuckerberg interview, serverless is going mainstream, Spotify moving to cloud, the quad-core Raspberry Pi 3, ten UK-based AI startups, what Google’s quest to build the perfect team found, the politics of attention and why everyone wants some of yours, the irresistible rise of Donald Trump (“Brace yourself America. It’s really happening”) and Noam Chomsky’s sombre reminder from history.
  • Week 7 (23.02.16): What the past can tell us about the future, why cheap sensors and AI are fuelling the next wave of major tech innovation and what Andy Rubin has to do with it, cyborg rat, machine learning that kills people, Apple vs the FBI, a UK government science paper on Blockchain, Twitter backlash, MWC impressions, the CAT S60 phone, Sense home hub, smart vs. Swiss watches, Nest, understanding innovation and competition in China and deleting unwanted memories.
  • Week 6 (16.02.16): Deep Learning and monkey brains, DeepMind and the rush to AGI, the origin of the 10000 steps a day Fitbit meme, the IoT Expo in London, Dotti, how to Snapchat like a teen, polar perspectives on blockchain, IMSI catchers, a mesh VPN called Tinc, Hangman in 3 lines of Python, LIGO and the discovery of gravitational waves and the pivotal role Python played in it, on procrastination, abstinence in general and some future developments that should be cause for forboding including the expensive chaos that would follow a Trump victory.
  • Week 5 (09.02.16): Xiaoice and the state of the art in chatbot tech, signs that the Babelfish is coming, VR and interior design, Alphabet has become the world’s most valuable company and the standard bearer for applied Machine Learning, a bot that plays golf, of eagles and drones, the last days of HitchBOT, WhatsApp, Visa, a game with no extra lives, the ‘truth’ about internet access in China, surveillance privacy and encryption, product owners, why coding isn’t easy and suggesting it is doesn’t help anyone, SETI and Bowie’s top 100 books.
  • Week 4 (31.01.16): My write up of Andrew Ng’s Coursera Machine Learning course including a worked example of Principle Component Analysis, DeepMind and their AlphaGo milestone, Roy Batty’s inception happened already, military drones presage the age of ‘warbots’, robotic Rubik’s cube solver, Facebook’s decision to shutdown Parse as a sign their focus lies elsewhere, Facebook anxiety and fairweather friends, Twitter’s struggles, why Donald Trump is dead wrong about Apple, WindowsPhone is dead but Windows 95 lives on in the browser, Hofstaders Law, what trolls and the Bullingdon Club have in common and some views on the shortness of life.
  • Week 3 (25.01.16): On the convergence of messaging and AI, neural nets and regular expressions, why IoT is really applied Machine Learning, Atlas as a really expensive household robot, spermbots, a real life Iron Man, the rise and fall of Microsoft chronicled in one graphic, behind the scenes of the Google autonomous car program, serverless architectures and the Ocean pocket server, big companies and the Stack Fallacy, contracting as a lifestyle choice, a TV series called The Brain, another strange mummification story, plenty of fish in the sea but by 2050 they will be outweighed by plastic, the Indian diaspora, more perspectives on Bowie’s death and the story of the ★ logo of his final album.
  • Week 2 (18.01.16): Thoughts on the life and death of David Bowie, Udacity’s MOOC guarantee, peach.cool, the art of designing chatbots, signs the future is happening around you right now, why the Internet of Things is a security nightmare, Square’s SPADE decision-making framework, Startup L Jackson’s guide to getting rich in tech, evidence that social media serves as an echo chamber of confirmation bias, how the world has changed since 2010, London youth slang, the UK curry crisis and the self-mummifying monks of Japan.
  • Week 1 (08.01.16): Mark Zuckerberg is going to build a digital assistant, CES themes, Android version distribution data for different parts of the world, LastPass upgrade, UK digital, is 2016 the year we start to Ask The Fridge?  Fitbit Blaze smartwatch, Raspberry Pi magic mirror, sense hat of a different kind, thoughts on AI and capitalism, Fleye personal flying robot, Drinky the drinking robot buddy, is your company on the road to failure? Iran’s blogfather on the state of the internet and thoughts on whether something ‘extraordinary’ is going on or not..