We are living in interesting times. Multiple technologies, improving exponentially, are converging. I have been chronicling this convergence for several years in my newsletter, Exponential View. As Bill Gates said, “Most people overestimate what they can do in one year and underestimate what they can do in ten years.” Likewise, most annual predictions overestimate what can occur in a year, and underestimate the power of the trend over time. Here are 18 areas, excluding climate change risks, which I think will be interesting to watch in the new year and why:
1. International relations, the political economy and governance, will desperately need new design patterns as we enter a new phase of the digital revolution. These should be developed in the public sphere with a wide range of participants.
There are four major themes to explore:
The massive global platforms are defining a new political economy. Their corporate sovereignty will chafe with states’ own sovereignty. Those same nations will curry favour with the platforms to win the putative economic benefits provided by them. The large platforms know that governments will be seeking to reign in their power, through regulation or legislation. These firms will accelerate their efforts to secure platform advantage and raise the baseline from which their settlement will be judged in the years to come.
National Artificial Intelligence (AI) strategies will emerge from more countries. The result? More grounds for co-operation and more reason to argue about intellectual property, privacy, data and license to operate.
Silicon Valley’s political culture — and how that has been codified into software, corporate culture and strategy — will continue to smell. The Valley will hire outsiders to fix these problems or, more likely, just for the optics. This will take years.
And before we’ve tackled that smell, crypto whizzes will establish governance mechanisms on emerging blockchain networks. They will do so with a narrow, ideological framing which will threaten to hurt us in the coming decades by which time these networks will mediate many of the resources we need. This matters because information technology systems affect how we build our understanding of the world; they affect how we perceive our set of choices; they affect how we act in that world.
In short: they affect our understanding both of the “is” and the “ought”.
2. While Silicon Valley leads, both innovation and scaling increasingly occur across the globe.
Europe and central America lead the way in decarbonising their energy chains. China is making huge strides in the large-scale electrification of its urban transport systems. Its focus on AI, supported by the state and the Chinese tech giants, will show up as novel methods and large-scale implementations. And not just in personal surveillance.
The US, with its declining health & social outcomes and turn inwards, will become less appealing to some entrepreneurs. And its business culture, focusing solely on corporate profits, will lack the motives to innovate in areas which impact the social fabric (for the collective good) or areas where a state can directly offer a comfortable regulatory environment (such as Dubai and Singapore).
Curiously, the EU will provide room for innovation because of its ability to bring broader groups of stakeholders together than competition alone can foster. Watch the innovation around open banking and privacy in Europe this year. Leapfrogging in other innovation hubs will continue. We may not see an African firm to rival web giants like Google, Apple, Facebook and Amazon anytime soon, but we will see meaningful innovation in fields like agtech, fintech & distributed power generation.
However, the largest firms in the world will hail predominantly from Silicon Valley and one, most likely Apple, will exceed $1 trillion in market cap this year.
3. More money will flow into technology but it will be concentrated at later stages.
Following Softbank’s lead, funds bigger than $5bn will abound now that the investment case of platform monopolies is well understood. These will seek to back emerging winners at a regional and global level (look at Careem and Didi, in ride sharing, for example).
This may create funding gaps at earlier stages in the market, as already evidenced by the seed capital slowdown in Europe and the US.
4. The AI software stack will continue to diverge from traditional software.
Novel interface mechanisms. One will be voice, both as an input and as the output. The second will be images. Embedded cameras are providing large-scale inputs to machine learning systems: computers can now see. (One example will be the growth of affective computing applications.)
Specialist hardware (think TPUs and other ASICs being made available from a wider array of cloud services) will combine with novel frameworks (TensorFlow and its competitors).
Cloud-to-edge computing as we deliver an increasingly large proportion of intelligence at the locality where it is needed.
A new paradigm of software development (where the best developers nurture highly parameterised models and cajole the training data to feed them).
New development, deployment and testing tools to support all of the above.
5. Artificial intelligence will be the technology investment priority for large firms.
After years of prototypes, automation technologies and AI software now dominate the CIOs agenda. They will invest and invest big. One group of winners will be the crop of 2013/2014 vintage AI startups now maturing into serious businesses with meaningful revenues & growing fast. The best firms, incumbent and startup, will combine AI investment with strategic and organisational change. Those same firms will move from simple notions of data supply chains to rethink their business models around data network effects and AI lock-in loops.
Firms which view AI not as a tool with which to expand their offerings but merely to cut costs will become lords of an ever-diminishing manor.
6. We will increasingly demonstrate how AI is augmenting human capabilities and move the narrative beyond that of “AI exceeding human capabilities in speech/image recognition/Go”.
We will see more evidence for the tangible benefits the superpowers AI tools can give us individually, and we’ll increasingly witness the power of the AI-augmented human.
The collective efforts of the research community continue to impress us, especially as we see low-hanging breakthroughs in areas outside of vanilla deep learning, such as reinforcement learning, adversarial networks, one-shot learning and unsupervised methods. (By the way, we’ll be barely any closer to human-like intelligence, and no closer to artificial consciousness.)
7. The discussion on how AI will impact employment will shift from solely focusing on the elimination of jobs to how best to help workers accommodate the inevitable change: in their responsibilities, the skills they need or requirement to find a new role.
Different countries will take different approaches. Those which combine an investment in social goods (like education and a safety net) and maintain a healthy approach to entrepreneurship and innovation will do best.
We will also make more progress in understanding the commons questions of trust, fairness and justice in algorithmic systems. Sensible boards, prompted by legislators, regulators and activists, will make ethical AI a top-table issue.
8. Crypto technologies will become more important & start to demonstrate their utility.
In 2018, the activity in decentralised applications and protocols based on tokenisation will increase, and important networks like FileCoin, Orchid and Ocean will come live. Below the speculative froth, in calmer waters, sober-minded teams are coming together to tackle real problems using the unique attributes of blockchain technologies (solving the incentivisation problem across a market network of random participants, allowing the emergence of trust in such a system).
We’ll see AI developers increasingly experiment with the fruitful combination of AI and blockchain. These areas include how to build a data commons to incentivise the sharing of data, allow the sharing of models, using blockchains and smart contracts for individual AIs to mediate their machine-to-machine interactions.
9. Sordid revelations in crypto speculation will be outweighed by the wall of money entering the assets class.
Asinine press releases, speculative investors and shady enablers get out into the market much faster than the technology can become useful. (Look to Canal Mania and the British Railway boom for antecedents.) This lurid funk will penumbrate technology progress as more out-and-out frauds are met with regulatory intervention and the cacophony of commentators watching speculators from the safety of their schadenfreude pulpits. It may not matter because every discussion about crypto assets increases their actuality in the market. This will see an increasing range of products for institutional investors and, crucially, high net worth investors wanting to get exposure to this asset class. This speculative bubble might pop, it might not.
10. The Knight Industries Two Thousand will remain the gold standard for autonomous vehicles.
Autonomous vehicle pilots will become increasingly ambitious, but the real world hurdles will still take time to navigate, even with friendly city regulators. None will ship to the public in 2018.
11. Healthcare becomes increasingly interesting for entrepreneurs.
Why? The first FDA-approved gene therapy is in the market.
More approvals as likely to follow. CRISPR is likely to appear in human trials this year too. Separately, the success of applying deep learning techniques to electronic health records, low-quality consumer tracking data and medical images will create confidence in producing breakthrough applications. Non-digital healthcare has an ineluctable appetite for cash. Ageing populations, coupled with an affordability crunch at both the state and private level, will increase the need for novel solutions — which AI-powered digital health might just provide.
12. A novel cyberattack, in terms of scale or quality, will emerge.
This may involve attacks which leverage some type of machine learning technology: either using chatbots or natural language generation, smarter password attacks, taking connected devices hostage or adaptive systems which avoid detection.
13. The US mid-term elections will be a focus of systematic information warfare, with the advantage with the perpetrators.
The main political parties standing for election (and many groups who are not) will use a wide spectrum of tools to target, persuade and mislead voters, such as never before.
14. Despite Magic Leap’s recent impressive demo video, augmented reality will continue to simmer rather than boil.
True believers in AR & MR will persevere but the opportunity for large-scale change afforded by AI and blockchain (especially in fintech, healthcare and energy) will attract the majority of driven entrepreneurs. Edge cases that extend reality, especially in industry, will be the most interesting. Smart firms will start to build up their capabilities in this domain today, to reap rewards in the future.
15. Digital advertising has been invasive for far too long, and this year ad tech will suffer.
The enhanced privacy features in iOS and Google Chrome, and the requirements of data obligations of the EU’s GDPR will hurt ad tech and programmatic advertising. Facebook and Google will barely notice and will continue to dominate the market.
16. Crypto mining’s hunger for energy will overshadow the growth of renewables.
The price of renewables will continue to decline and new solar and wind contracts will be substantially below the best fossil fuels can offer. On the downside, the energy consumption mining bitcoin and other tokens will continue to grow at more than 20% per month, unless there is a huge price correction. So by this time next year crypto-mining using about 10 times as much energy as it does today, rivalling Italy’s consumption. Our one hope is the more energy efficient verification methods, like proof-of-stake, become commonplace.
17. Ethics will increasingly drive consumer choice and investing strategy.
Consumers will increasingly make purchase & investment decisions based on their resonance with the ethical positioning of a firm. These will get amplified by industry, particularly the insurance industry, which needs to price in risks related to climate change or regulatory malfeasance. University endowments may feel pressure to adjust their investing stance and divest from certain types of assets (or companies.) I’ll be intrigued to see if the #metoo movement gets reflected as an investment risk.
18. Buddha, Aristotle, Hayek and Marx make a comeback.
Marx because the last fifty-year consensus between workers and employers and financial capital is strained, so some will look for the pendulum to swing back.
Others will look at the combination of increasingly cheap energy (reducing the cost of production towards nil) and increasingly capable machines (reducing the average human’s ability to be paid for their outputs) and argue only a state of radical abundance — or “each according to their needs” can work. Critics of greater central intervention in our collective affairs will raise the spectre of Marx, and often through Friedrich Hayek’s critique of it. Hayek’s notions of the market as the most effective information discovery and transmission mechanism will attract more interest as blockchain-style networks show their utility as resource coordination systems.
Aristotle re-asserts himself because while we are wealthier than ever before, his call for eudaemonia (human flourishing) will seem to stand above the noise of “recommended for you” consumerism. Buddha’s relevance will be driven by a greater awareness of mindfulness and contemplation in our dopamine economy. Equanimity will be a helpful characteristic during turbulent times. And as machines appear to be more and more lifelike, and neuroscience unravels more mysteries of our consciousness, the quiet contemplation of our subjective personal experience may become a sanctuary for our humanity.
Originally published at www.weforum.org.
No comments:
Post a Comment