How competing software, hardware and political ideology could accelerate existing divisions in humanity into the future
It’s been great to get people’s thoughts and feedback on the last article on “The iPhone 20”. Some of your responses considered that given Apple’s business model is effectively a walled garden, this makes any integration with the human body very unlikely in the future.
That’s why in this article, rather than focusing on Apple, I’ll look to explore some of the ways tech companies and organisations (including government bodies) will compete on software, hardware and protocols that will shape humanity’s journey to 2029, which may accelerate our existing divisions that stifle collaboration and splinters our future societies.
Software and hardware ecosystems
We already experience almost a technological xenophobia against people not using the same software platforms as us. iOS users are annoyed about other people who’s texts don’t appear blue on iMessages. The same goes for fast file transfers using AirDrop, where the alternative is to share it via WhatsApp at low resolution, or uploading it to a DropBox/Google Drive for the other person to download after sharing them a link somehow. Oh it’s a large video? Sorry, you’ll have to upgrade your file storage service to fit it in. Other third party file transfer apps would charge you money for the premium version which could actually work cross-platform. So the next best thing would be to share it via a Facebook or Messenger, right? Oh wait, the other person followed the #deletefacebook movement and now you’ll have to install and send it to them via SnapChat. And then suddenly, you’re playing with face swap…
You then have to rock up to a work meeting that you’ve been invited to last minute, where you’re asked to present. Great. You hope the venue has an Mini DisplayPort cable instead of only HDMI to connect to your MacBook, or failing that, hope they have an Apple TV and not a Chromecast or those buggy WiFi-connected displays. Your iPhone’s also running out of battery so without a charger, you’re also hoping they have a spare Lightning Cable and not a USB-C.
It was an awkward start to the meeting, but eventually you get the technology to work. Afterwards, you’ve been tasked to look for a Christmas present for your little nephew, so you buy them an Apple Homepod at JB Hifi. A couple of days later, you end up getting re-gifted the same present at a secret Santa gathering because they wanted an Alexa that integrates better with Spotify instead.
I’m thinking about where should I post this article later… Medium or Linkedin? You get the point. These tech companies want to lock you in their walled gardens and encourage you to stay in their ecosystem and to convince others that you are either “with us” or you’re “against us”.
But the most interesting platform war is being fought in cloud services and artificial intelligence (AI) development.
For Enterprise, cloud market share among the likes of Amazon Web Services (AWS), Microsoft Azure, Google Cloud Platform (GCP), and IBM Cloud is as follows:
Open-source cloud platforms like OpenStack also offer a good alternative to the mega-corporations who are incentivised to lock you in to their ecosystems. The challenge is the complexity in implemetation and available skillsets.
The fragmentation in cloud platforms has created greater reliance on third party containerisation and orchestration tools like Docker and Kubernetes, which have to deal with stability issues to support interoperability.
However, what the different cloud platforms have also created is a fragmentation in the deep learning frameworks adopted by data scientists and AI researchers.
In AI development, the deep learning framework you adopt will depend on whether you’re:
- Using Microsoft Azure — it’s easier to integrate with the Microsoft Cognitive Toolkit, CNTK.
- Using AWS — it’s easier to integrate with Apache MXNet
- Using GCP — it’s easier to integrate with Google Tensorflow. (Google DeepMind also uses TensorFlow, plus an open source library called Sonnet for research. Based on DeepMind’s AlphaStar performance against professional StarCraft 2 players, this is likely to be the cutting edge of AI research.)
- A Facebook AI researcher — in which case you’re likely to be using Pytorch and Caffe2
- Applying OpenAI’s open source toolkits — in which case you’re likely using OpenAI Gym for comparing reinforcement learning frameworks. (OpenAI may have lost its 5v5 match in August 2018 at the Dota 2 International, but its progress since then has likely surpassed this milestone today.)
- An IBM or Intel AI researcher — in which case you’re likely using Chainer
- A Java developer — it’s easier to use DL4J.
I won’t explore the different coding languages and developer frameworks, but here’s a good analogy in comparing coding languages based on Lord of the Rings characters:
In an attempt to resolve the fragmentation in the deep learning community, Amazon, Facebook, Microsoft and others have got behind an open format to represent deep learning models, called the Open Neural Network Exchange (ONNX).
ONNX enables interoperability between deep learning frameworks such as Apache MXNet, Caffe2, CNTK and PyTorch via the ONNX Model Zoo. How well it gets supported in the future will likely depend on what’s at stake between the major tech companies…
Distributed ledger technology protocols
There’s also another battle happening which may determine the very future of the internet as we know it. The story so far:
- The World Wide Web was originally designed in 1991 by Tim Berners-Lee while he was a contractor at CERN. It is all the Web pages, pictures, videos and other online content that can be accessed via a Web browser. The Internet (built upon ARPANET), in contrast, is the underlying network connection that allows us to send email and access the Web. ARPANET took advantage of sending information in small units called packets that could be routed on different paths and reconstructed at their destination. The development of the TCP/IP protocols in the 1970s made it possible to expand the size of the network, which now had become a network of networks, in an orderly way on the Internet.
- Web 2.0 was the continued evolution of the Web since the original design to include interactive (social) media and user-generated content that requires little to no technical skills, such as blogs and wikis. Arguably, Web 2.0 has been won by social media companies like Facebook and Tencent’s WeChat.
- Web 3.0 is evolving today, including the Semantic Web described by Tim Berners-Lee. The battle is raging on for distributed ledger technology (DLT) protocol will be used to create the “new new internet”. Currently, the blockchain protocol, Ethereum has the largest developer community working on Web3 as it enters the next hard fork in development.
For an overview of the DLT protocols and cryptocurrencies out there, here’s a high-level competitive landscape, courtesy of an Australian DLT startup, XCredits.
However, there is already a de-facto fork in the way users experience the internet around the world, particularly in countries with higher content censorship. China has banned all cryptocurrencies but are considering backing various blockchain initiatives for trade and finance. This may eventually result in Chinese government-backed programmable money that will reinforce its controls over transactions in society and the economy.
As at the date of this article, the Trade War between China and the US is still at large. Reminiscent of the beginnings of the Cold War between the US and Soviet Union, some have compared the Marshall Plan to China’s One Belt One Road initiative. Existing communication barriers in language, culture and political ideologies between these two global superpowers, who are on an economic collision course, is likely breeding mistrust among the uninformed, and stifling collaboration for business and scientific research.
This is exacerbated by the current censorship restrictions in place in China on services from Google, Facebook and other US tech companies which the Chinese government cannot control, as well as by the current hawkish stance from Western countries led by the US on 5G deployment by Huawei and allegations of IP theft. Add to all this the nationalism overtones in daily news coverage from the likes of CCTV, Fox News and Trump and we have a recipe for deep divisions between the societies of the world’s largest economies.
China arguably already has a newer version of the Web through the mass adoption of WeChat by the Chinese population. WeChat has become the de facto operating system in China, with its integrations to most applications & payments (WeChat Quick Pay and QR Code Payments), digital identity and social networking, all available within the app.
The information is accessible by the Chinese government, and is added to the social graph that the government has on all Chinese citizens. By 2020, China plans to roll out nationally the Social Credit System, which has been piloted in cities such as Shenzhen (Guangdong), Jinan (Shandong), and Suzhou (Jiangsu).
China has tested opening up the social credit scores to social media platforms like WeChat, creating mini-apps like this heatmap of people around you who are in debt:
Your social credit score is even penalised if you are “friends” with people with lower scores.
Extrapolating this divisive trend in society with software, hardware and political ideology, imagine a future where the very means of communication were restricted altogether between members of the human race? In the coming decades, former Google CEO, Eric Schmidt suspects there will be two major internets out there — the one we know and love or hate, and a new, more heavily-censored one built by the Chinese government.
By 2029, it’s likely that the evolution of brain-machine interfaces (BMIs) will be contributed by smart devices and operating systems from various tech companies and government organisations, as it will become a “need-to-have” in order for citizens to stay competitive amongst themselves, and against other nation state populations.
But what if the other person or services you’re trying to communicate with using a BMI is on a different platform — one that’s backed by another government or mega-corporation that is incentivised to keep a walled garden for its netizens?
Privacy vs Transparency, Open-Source vs Proprietary
In the short story, Manna by Marshall Brain, the utopia of the “Australia Project” depicted how to potentially avoid the fate of the future dystopian society in the United States. In the US, automation through better and better narrow (or weak) AIs created an unemployable class trapped in a pacified, unfulfilling life on basic services provided by the government. It allowed an exorbitant expansion of inequality between those who owned the means of production via the Manna AI, and the majority of the population who became part of the “useless class”.
The Australia Project existed on the basis that a competing AI to Manna was developed by its founder, and was open-sourced to serve citizens of the Australia Project. In exchange, an initial investment was made by its new citizens for one “share” in the Australia Project. As the Fourth Industrial Revolution accelerated, a post-scarcity society emerged and the fruits of automation created abundance for its citizens. Every citizen was given 1,000 credits to spend on whatever they like, and this covered almost everything they wanted, let alone needed, with the exception of certain large projects which required pooling of resources.
In this “utopia” however, the only catch is that the AI of the Australia Project was granted full transparency over its citizens, with the following principles:
- Everyone is equal
- Everything is reused
- Nothing is anonymous
- Nothing is owned
- Tell the truth
- Do no harm
- Obey the rules
- Live your life
- Better and better
You see, although the citizens had access to BMIs which enabled them full access to the collective intelligence of humanity (along with an ability to perform brain-to-brain communication), nothing was done anonymously. This enabled the AI powering the BMI network to function with full effect. It enabled the AI to stop people murdering each other, and penalise those who abuse or exploit the abundant resources in the economy.
Sound familiar? Try comparing the Australia Project’s “principles” to the basic policies of “Xi Jinping thought”.
The Australia Project’s credit system is also reminiscent of the universal basic income (UBI) movement, which argues that such a safety net would remove wasteful bureaucracy in administering welfare payments (since everyone receives the same amount of UBI, there’s no need to prevent rorting), and promote economic activity in the form of risk taking, pursuit of projects aligned with the individual’s skill sets and passions, as well as quantifying the value for those who perform tasks that are currently not recognised by economic measures like Gross Domestic Product (GDP). This includes looking after children and the elderly at home.
There is a growing trend to recognise measures like the Gross Happiness Index, such as in Bhutan and the focus on wellbeing, as promoted by New Zealand’s Prime Minister, Jacinda Ardern in her government’s 2019 Budget.
One potential way to fund a UBI or enable citizens to live a UBI-like lifestyle is moving towards Web 3.0, where personal data can be monetised by individuals users. You see, services like Facebook and Google are “free” because of the data users enrich these platforms with. As the data becomes more and more valuable, we are essentially paying a higher and higher price for these services. In the age of automation, this may be the mechanism for enabling individuals to fairly benefit from these technologies.
AI Supremacy and Accelerating Artificial General Intelligence
On the journey to 2029, the geopolitical battle on artificial intelligence between China and the US is likely to accelerate. China has a 2030 strategy to become the global dominant player in AI, and with “Xi Jinping thought” weaved indefinitely into the DNA of Chinese governance, this path is unlikely to be altered or deterred.
Rather than an iOS vs Android integration problem, we could be racing towards a real Cold War division of internet communications, and future brain-to-brain communications — one developed with “Xi Jinping thought”, and the other powered by AI from the G-MAFIA US tech giants (Google, Microsoft, Amazon, Facebook, IBM and Apple). In the future, the G-MAFIA may need to merge (along with Elon Musk’s Neuralink) just to compete with China’s unified front from Baidu, Alibaba and Tencent (BAT), backed by the Chinese government.
This AI race will accelerate the development of Artificial General Intelligence (AGI). In what shape or form that AGI arises will be determined by the winner, and it’s a winner-takes-all situation. This why thought leaders in AGI such as Ben Goertzel at SingularityNet have expressed that an open-source approach to AGI development provides the best chance of it aligning with all of humanity’s values. Elon Musk, however takes a different approach with Neuralink, essentially enabling humans to merge with a potential AGI. When you can’t beat ’em, join ’em. Resistance is futile?
Or I could be wrong on the China vs US front or on AGI, and xenophobia becomes no longer about what country you’re from, but which operating system you’re running on. This could cultivate mistrust and animosity arising from miscommunication among those running different BMI operating systems. The tech companies would continue to inflame these divisions, pouring investment into Cold Wars between mega-corporations, one fought on better features, functionality, and more immersive brain-to-brain experiences & virtual realities. The virtual world could be the new islands of the iron curtain, just as the citizens of the Australia Project were divided based on those who preferred the virtual world, and those choosing to interact mostly in the physical world.
We’re at a crossroads in our journey towards 2029 and the future of society. That’s why creating an open-source version of the latest features and functionality offered by the likes of G-MAFIA, BAT and Government organisations is potentially more important now than ever.
If you’re interested in some of the latest developments on open-source DLT protocols and AGI, check out the developer community at SingularityNet. And to learn more about those predicaments faced by society today, check out Yuval Harari’s 21 Lessons for the 21st Century.
DDI Recommended Reading:
- Humans Need Not Apply: A Guide to Wealth and Work in the Age of Artificial Intelligence by Jerry Kaplan
- Life 3.0: Being Human in the Age of Artificial Intelligence by Max Tegmark
- Artificial Intelligence with Python: A Comprehensive Guide to Building Intelligent Apps for Python Beginners and Developers by Prateek Joshi