Data Driven Investor: Tech Trends 2020

The trends I am describing will have a spill-over effect and feed into other trends, including those left out from this year’s selection, such as the Internet of Things, autonomous vehicles and blockchain.

TechTrends 2020 blog Norbert Biedrzycki

My article in Data Driven Investor published 4th of December 2019 on Tech Trends 2020. 

I have a growing sense – which I think I share with many readers – that people are getting used to hearing about artificial intelligence. Neural networks are making themselves at home in our businesses and households as they either creep into them discreetly or splash in spectacularly. We no longer speculate about what its advent may mean. Instead, we focus on more specific applications and benefits. The changes seen over the past few years are extraordinary. Not only in the technology that surrounds us but also in our economies and ways of life. For instance, when I started writing this blog in 2016, the sharing economy was an idea that may have been appealing to many but that remained highly hypothetical. Today, as I go out, it takes me seconds to use an application to locate a shared car that I can pay for by the hour, sitting around the corner, and to start driving. This is just one example from daily life. I could name many more from the world of business.

What follows is a list of subjectively selected trends that I think will take off in 2020. It is difficult to say which of them will be the most successful and what that success might be. The trends I am describing are not isolated, meaning that growth in any given area, such as 5G technology, will have a spill-over effect and feed into other trends, including those left out from this year’s selection, such as the Internet of Things, autonomous vehicles and blockchain. The reason they were omitted is not because these technologies are any less significant or stagnant but rather because I have discussed them many times before.

Artificial intelligence taking to the cloud

Machine learningnatural language processing, and image recognition are becoming increasingly familiar to sales, marketing and IT. Debates on whether to take advantage of artificial intelligence have grown more substantive, which naturally entails questions about money. Only the largest players can afford to develop their own systems. Hence, a niche has emerged of machine learning platforms made available on a shoestring. Virtual tools allow companies to choose what data and resources they want to process with algorithms and when to do it. Graphical user interfaces support access to many functions on their platforms. Even without high technical skills, workers can test AI performance in their fields. The cloud AI segment is growing by leaps and bounds. The heyday of this service is still ahead. Less affluent businesses can test AI in relative comfort. The cloud is again changing the face of the market.

The world belongs to Big Brother

Computer image analysis and facial recognition have become all the rage. Every day, machines learn to see things better, which we can notice by for instance logging into a smartphone that runs a facial scan. By all indications, such scanning will soon be the norm for shoppers, rendering card readers obsolete. In China, such scanning is already commonplace. As machines improve their vision, robotics is inching closer to a major breakthrough. Robots capable of making sense of their immediate surroundings will better integrate with human environments. Better image analysis can also accelerate advances in autonomous vehicles and streamline the detection of defects in devices produced on assembly lines.

Needless to say, we are aware of the risks involved in such development. Image-recognition algorithms are imperfect; the presence of cameras on city streets and in schools and office buildings is sparking protests. Some US and French cities have already begun taking them down. The recognition of people’s images in social media is also an issue. Hence, proper regulation is necessary. Any technology that is either misused or used recklessly can generate social tensions and lead to abuse. This said, I firmly believe that machines that can see better are the future of our civilization.

Awaiting the arrival of a great network

5G technology can alter communication standards and create a new quality in data transfers. The consequences of this for the growth of industry, research and regular consumers of electronic content are indeed profound. Equipped with ultra-fast transmission networks, such consumers will be free to stream top quality movies and music. If everything goes to plan, for the first time in the history of mobile networks, location will no longer constrain people. In practical terms, this may mean the rise of an unlimited global transmission network. The consequences of this for business are enormous. Such robust data transfer technology may well fuel the growth of many other technologies and projects that now remain dormant, waiting for transmission channels to be unlocked. One example is the Internet of Things, which hooks up a wide range of digital devices to a global network, including phones, TV, cars, household appliances, cameras, robots and all kinds of electronic industrial equipment. If 5G develops as predicted, we will enter a new technological age characterized by unconstrained communication among various devices.

Seduced by machine eloquence

I think that the next few years may see a breakthrough in devices’ ability to understand human voice. Our interactions with bots are poised to change. Simple commands and questions are gradually replaced with more natural conversation. For this to happen, the voice-operated device must be more than a mere programmed machine that performs a few isolated tasks. Semantic Machines, a Microsoft-owned company, is doing research on conversational intelligence. The intention is for devices that respond to the human voice to be able to comprehend complex contexts, i.e. process information streams from multiple sources.

The research gives hope of producing voice assistants that link situational contexts and meanings with emotions and anticipate interlocutor intentions. If you tell your assistant to book a theater ticket, it should revert by asking you what time would suit you, whether you want to get there by metro, and whether you are planning to eat out after the performance. The assistant becomes proactive and capable of expressing complex meanings. The reports on the trend of “Alexization” of our lives, named after the popular voice assistant Alexa, may not be greatly exaggerated. Today, Alexa is already connected to over 80,000 digital devices of various types ranging from home appliances to all kinds of consumer electronics. Other assistants include Google Home and Apple HomePod. If these devices master the art of conversation, their presence in our lives will become as significant as Google searches are today.

Related Article:   Tech Attacks Bias, But Not Totally

Robots learning to be versatile

A machine that could move objects over distances would significantly reduce delivery time. But to be accepted in a business environment, it will have to woo people with additional skills. And it soon will. New-generation robots have technologies for processing visual data and understanding what happens around them. The effect can be impressive: a robot capable of navigating its way neatly around employees, avoiding obstacles, recognizing floor surface anomalies and, importantly, taking in a growing volume of data to bolster its “brain”. This portends a breakthrough in robotization, promising to overcome the two barriers that have had significantly slowed their conquest of the business world. It will no longer be necessary to keep people and machines in strictly segregated zones because robots will be able to yield to humans when needed. 

The skill most in demand is for robots to learn how to perform many tasks simultaneously. While today’s machines are greatly limited in their ability to multitask, their counterparts that excel in that skill are looming. As robots continue to benefit from advances in voice processing, their communication skills will explode allowing them to blend seamlessly into the human environment. And one more thing: agility. The experimental robots manufactured by Boston Dynamics are outstandingly good gymnasts, and yet certain human activities remain beyond their grasp. Even here, however, there has been some progress. Robots are currently in intensive training to learn to accurately recognize the shape of the objects they hold.

Will the robots of the near future be able to take an order, find goods on a shelf, place them in a container and move them to a postal machine? Affirmative on all counts.

Sharing: Connect rather than own.

The business model that has companies earning money by sharing their unutilized resources has been around for over a decade. In the early days, initiatives such as BlablaCar, Couchsurfing, and Airbnb were viewed as startup experiments tapping into the huge potential of communities. These initiatives have by now showed their inherent business value. Interestingly, the trend has evolved over the years. As it lost its luster of alternative novelty, it became a serious innovative way to turn a profit, invest money and manufacture goods. Fund financing fueled the trend, resulting in today’s popularity of scooters, cars, bicycles, and apartments rented for short durations.

However, it is impossible to ignore the disruptive upshot of the phenomenon. “Sharing”, which started out as a glorification of social bonds, turned into “uberization”, with all its downsides. This is not only about Uber’s effects on the taxi business and controversies over drivers’ pay. It also concerns the disruption of property rental rates in big city centers caused by the mass purchasing of apartments for community letting. We have come to a point where business on technological steroids requires regulation. If laws can gradually be passed, the trend will grow on. We will find we can share anything: luxury clothes, computers, photographic equipment, cars, airplanes and virtually any services. As it turns out, connecting instead of owning doesn’t only appeal to the young. It is a reasonable strategy for cost reduction in almost every business. Building your own company in which all equipment is paid by subscription? Why not?

Algorithms scan our bodies

Whether AI can prolong our lives and improve its quality is no longer philosophical speculation. The benefits of machine learning with its predictive functions that support the analysis of infinite data combinations can be seen in hospitals and medical laboratories around the world. A real, profound change is unfolding before our very eyes, not only in the way we treat and prevent diseases but also in the way we tap into the unutilized potential of our bodies. 

In the near future, an entire industry will provide the average person with a range of tools to observe the performance of their bodies. Applications will emerge that will take our EKG and detect early symptoms of serious diseases. Pharmaceutical companies will use artificial intelligence to dramatically accelerate processing the kinds of data sets that have so far required years of painstaking analysis. Algorithms capable of combining an infinite amount of information about age, medical history, current results of laboratory tests and physical examinations, image analysis, and our DNA, will be available not only to specialized laboratories but also to individual doctors. Swallowed capsules containing miniature microscopes will offer insights into our bodies. The data sourced in this way will go to digital files along with data from our personal medical wristbands, and DNA test results that will become ever easier to interpret. As such files fill up with continuous data streams, self-learning machines – which work better on large data collections – will deliver a more precise picture of our bodies and therefore ourselves.

Printers to replace farms

One of the saddest paradoxes of our civilization is that the way the fundamental commodity, which is food, is produced violates ethical principles, causing environmental degradation and reinforcing bad eating habits. This applies primarily to the mass production of meat. And although estimates show that meat consumption will increase by up to 70 percent in the next two decades, our awareness of the woes is growing. Tech companies and startups follow the trend of ethical and healthy food, join the debate and propose innovative solutions. Over the past two years, there has been much talk of an initiative financed by Bill Gates of providing an alternative to the mainstream approach to meat production and consumption. The initiative involves Beyond Meat and Impossible Foods, two companies that offer healthier food and show that mass production of food may well be in line with sustainable development. Technology provides three options. We can grow meat in a laboratory using animal stem cells. We can also produce and share, on a massive scale, an entirely plant-based product that tastes just like meat. Vegan burgers have a strong foothold in the menus of McDonalds and Burger King. There is also another way that is being tested by startups. An example is the Israeli company Redefine Meat, which has recently unveiled a burger produced in a 3D printer. While this last proposal may strike you as a bit odd, technological advances can surprise you and change both views and habits. Only one thing is certain: it is absolutely imperative to invest in innovative food production. As funding streams in, one can only expect the trend to continue for years to come.

Link to the article.

Related articles:

– Artificial intelligence is a new electricity

– Will a basic income guarantee be necessary when machines take our jobs? 

– Machine Learning. Computers coming of age

– The brain – the device that becomes obsolete

Leave a Reply


  1. Adam

    interesting thoughts and a good, succinct article.

  2. Mac McFisher

    This is is gonna be more less speculative because i havnt studied quantum computing in any great depth – but its possible another avenue is using QC for reinforcement learning where an agent learns through repeated trials and optimization of policies. Still dont think we’d be able to achieve a general intelligence but we would I suppose in theory be able to very rapidly simulate extremely large numbers of trials to determine optimal policies (then glue enough trained models together to get something approximating a gen intel).
    In general though, while quantum has alot of potential im pretty weary of putting all our eggs in that basket. We are very very far from being able to even test those sorts of things. Generally notice whenever I comment on the unfeasability of something compsci related I invariably get something like “its possible we may be able to with quantum computing”. From my current understanding quantum will have a limited but very powerful set of use cases. It doesnt exactly work as a “speed boost” of classical computation, rather a completely different paradigm.

    • Jack666

      Right, but the so-called neural processor is mostly being used to do IR depth mapping quickly enough to enable FaceID. It just doesn’t really make sense that it would be wasting power updating neural network models constantly. In which case, the AX GPUs are more than capable of handling that. Apple is naming the chip to give the impression that FaceID is magic in ways that it is not.

    • tom lee

      AGI would need “Einstein theory of relativity level” breakthrough at unsupervised and semi-unsupervised learning currently unsupervised methods are very lacking and still many ways best methods are from 1900-1960.

    • SimonMcD

      I understand that a virtual machine is a self contained operating system running on another piece of hardware(ex – running a windows virtual machine on linux) – emulation
      I know all virtual machines on a machine will share that machine’s resources(RAM, CPU, Storage)
      From reading about containers, I can’t differentiate them and virtual machines and identify what problems containers are supposed to address that virtual machines can’t.
      One thing I did learn was that containers require less overhead than virtual machines because you’re emulating an entire operating system with virtual machines

  3. AKieszko

    It depends on how you define AI. I use a broader definition, because I believe a wider range of things might pose risks if we aren’t very careful. Like corporations easily could be considered AI if you remember that intelligence is platform neutral. I’m more an AI behaviorist as in if it behaves with a certain degree of complexity I’m comfortable saying it has a certain amount of intelligence.

  4. Simon GEE

    AI should be regulated, but that doesn’t mean it will be….
    As much as I love tech and the bare bones of it, I think AI is not a good idea…..Now the kid in me wanted it for years! The adult in me is scared of that shit!

    • John Accural

      Your brain is estimating how much you will enjoy something (making you motivated to do it) not as an average or single data point, but is predicting a range of possible outcomes for how awesome it might be.

      I think.

      • Mac McFisher

        Personally, I think we’re already there, it just isn’t cost effective yet. And we’d have to do it the hard way. Based on experiments with segments of rodent brains, we could simulate a full human brain now if we had dump trucks of money to throw at it. The procedural/programmatic (not simulating a full brain) approach probably won’t happen until after someone does that, and that’s not going to happen until the hardware to do it is cheaper. Which, IMO, makes 90 years an optimistic estimate. As much as I make fun of the 10-15 year thing, I think it’s true if you assume unlimited funding.