From Machine Learning to Digital Twins: A Sea of Opportunity for Upward (and Downward) Technology Trends New and evolving technologies are rapidly changing the way we work, offering creative opportunities for those developers willing to refocus and learn new things. In this article, we'll take a look at 11 new technology trends that experts say could disrupt existing approaches in IT and create demand for future-oriented developers.
It's not just about the next mega-breakthrough (aka Next Big Thing). The source of future opportunities for developers lies in the confluence of several advanced technologies - artificial intelligence (eng. artificial intelligence, AI), virtual reality (eng. virtual reality, VR), augmented reality (eng. augmented reality, AR), internet of things (eng. Internet of Things, IoT) and cloud computing... and of course the security issues that flow from this merger. If you'd like to expand your personal toolbox, we recommend taking a look at these popular technologies, as well as listening to our tips on how to succeed with them.
Internet of Things Security
After tens of millions of connected devices were hacked in 2016, even outsiders are noticing that unsecured IoT devices (IoT = internet of things, Internet of things) lead to terrible security problems. In a recent report, analyst firm Gartner published recommendations for developers and expert teams. It suggested that these professionals should work with each other from the very beginning of the design process. In this case, you can eliminate threats as they arise. For example, by downloading security updates to IoT devices. The demand for IoT security specialists is very high, especially for those who understand the vulnerabilities of the hardware and software used by devices connected to the network. “The attack vectors in the IoT are almost identical to those for any distributed network, such as computers or cell phones. So security knowledge will be relevant and important in this area as well,” says Richard Whitney, vice president of product launches at startup Particle. “Learn the basics of cryptography and authentication, and you will achieve a lot.” Tom Gonser, founder of DocuSign and partner at Seven Peaks Ventures, says firms now need low-level microprocessor programming skills. “They also need experience with Bluetooth technology, [the Windows Identity Foundation] and fuzzy spectrum technologies. Knowledge of the latest security features of the Linux operating system is also valued, especially options optimized for the mini-kernel, for example, Qubes OS. Matt Abrams, a partner at Seven Peaks Ventures, suggests focusing on understanding processes and breaking down techniques. In his opinion, the era of post-quantum cryptography is approaching faster than expected. Specialists need to understand what differential privacy and adversarial networks are."
Artificial intelligence
The demand for AI-savvy engineers, in anticipation of a new wave of driverless cars, robots and "smart" electronics, is growing by leaps and bounds. “We are now at a tipping point, largely driven by advances in ubiquitous computing, affordable cloud services, and virtually limitless data storage,” said Nicola Morini-Bianzino, senior executive director and head of the AI group at Accenture. “Artificial intelligence is now built into literally everything.” Morini-Biancino predicts a demand for software developers, technologists and researchers with experience in the fields [automated - approx. transl.] translation from one language to another, speech recognition, machine vision, robotics, processing of texts in natural languages, representation of knowledge and construction of logical reasoning. The food for AI is data, so the need for data and content managers, data scientists and analysts is also extremely high. Kiyoto Tamura, vice president of marketing at Treasure Data, predicts that artificial intelligence will soon move from highly specialized, mundane applications to much broader - and exciting - applications. Previously, tasks for artificial intelligence looked something like this: "Find the best route for delivering a package ... or the most suitable sites for a search query." Now their wording is closer to the following: “play go at a decent level”, “driving safely”, etc. "This is cool, there is an asymmetry between supply and demand in this area. As a result, subject matter experts will remain a valuable resource until academia and industry strike a balance.
Machine learning
One form of artificial intelligence, machine learning, can process vast amounts of data to quickly find patterns, such as face recognition, and solve problems like recommending movies for streaming, without the need for explicit programming. Patrick Spedding, senior director of business intelligence research at Rocket Software, believes that cognitive technologies, together with bots and machine learning, can improve the efficiency of organizations engaged in the search for "a useful signal among the noise." “Machine learning, after all, is based on the capabilities of advanced analytics, formerly known as “data mining” (data mining), which for greater “popularity” lacked only the right platform,” Spadding comments. The question arises: how to gain experience in this area for those developers who want to devote themselves to machine learning? Abrams of Seven Peaks Ventures recommendsonline courseAndrew Ng on Machine Learning at Coursera. Those who complete this course perform better in Kaggle contests than some practitioners with years of experience. Not every developer involved in machine learning has a background in computer science. “Of course, a degree in computer science or a basic engineering degree usually helps technicians succeed in their jobs, and they are able to run experiments and refine machine learning models for a long time,” said Solvvy CTO and co-founder Mehdi Samadi ). However, I have often observed
data science
Data science is another trendy field that requires a variety of interdisciplinary skills, and each industry has its own. Transforming large amounts of data into a form suitable for making business decisions may require experience with machine learning and artificial intelligence. “Experienced data scientists are a scarce commodity,” Spadding says. “It seems to me that the areas where you can create technologies that help in decision making, for example, intelligent bots (eng. cognitive bots) and guided analytics (eng. guided analytics) are areas of extremely profitable opportunities.” For those who wanted to to work in these areas, a thorough knowledge of probability theory and mathematical statistics is a key requirement, says Gary Kazantsev, head of the machine learning group at the Bloomberg agency. “Engineering skills, such as the ability to write the code needed to create a system, will not be superfluous. However, with the advent of tools such as the TensorFlow machine learning library or Jupyter notebooks, this task has become much easier. Good research skills come in handy in data science, that is, the ability to formulate hypotheses, test them, study the current literature, and constantly follow the news in your field.” Gunter Ollmann, head of security at Vectra, says that in At present, many companies treat data scientists separately from designers, research teams, and developers.As deep learning and machine learning tools improve, and as training courses increasingly educate lead engineers about what's new in data science, the distinction between data science and software engineering will gradually disappear. In the future, the fusion of the skill set and the ability to handle both tools will become a must."
Transaction block chain
The advantages of this way of creating a distributed financial accounting tool for transactions are both transparency and security, although the lack of standardization is slowing its adoption in a wide range of industries. Peter Loop, assistant vice president and principal technology architect at Infosys, is optimistic about the technology: We will see its full-fledged deployments in financial services, insurance and healthcare in 2019. This will completely undermine our payment system on an international scale." Other emerging technologies have steeper learning curves, says Robert Bardunias, co-founder and chief tax officer of IRIS.TV, fascinated by the technology's inherent emphasis on entrepreneurial activity. “These technologies are growing from day one, driven by operational commercial use cases, so developers don’t have to imagine use cases – they emerge and evolve in real time,” comments Bardounias. “Keeping up with new developments and changes will be challenging. a task for those who want to become a professional in this field.I remember how I once developed secondary professional skills, such as reading websites - and trade magazines.This is the last thing I wanted to do, but today it is mandatory part of developer training,
Application and Service Mesh Architecture (MASA)
There is also an ever-increasing demand for apps that can stay connected, switch, and work as we move around the house. "The point of a meshed mesh is high availability: all the elements are connected," says Joseph Carson of Thycotic. , as well as for cryptocurrencies such as bitcoin as a distributed financial accounting tool." Other experts point to a potential bottleneck in the lack of sufficient device compatibility. "Manufacturers are all trying in their own way to gain consumer confidence by making their own ecosystems, if they even exist, closed," says Derek Collison, CEO of Apcera (previously an employee of Cloud Foundry). “It seems to me that artificial intelligence will be trained in the clouds, on huge amounts of data from all users,” Collison says. “These algorithms will continuously update their execution models, which will be transmitted to end devices using wireless communication and used to update firmware on our phones, cars and home devices. Data processing will take place on the hardware of local devices, and training will be done using software in the cloud." “These algorithms will continuously update their execution models, which will be transmitted wirelessly to end devices and used to update the firmware on our phones, cars and home devices. Data processing will take place on the hardware of local devices, and training will take place using software in the cloud." “These algorithms will continuously update their execution models, which will be transmitted wirelessly to end devices and used to update the firmware on our phones, cars and home devices. Data processing will take place on the hardware of local devices, and training will take place using software in the cloud."
Digital twins
Connected to physical and virtual sensors, software models can be used to predict product and service failures, allowing companies to plan and allocate resources for repairs before failure occurs. Advances in machine learning and the introduction of artificial intelligence technologies are making it possible to reduce the cost of such predictive modeling, referred to as "digital twins", which can significantly increase efficiency and reduce operating costs over the life of, say, a jet engine or power plant. According to Mattias Woloski, CTO and co-founder of Auth0, businesses will be able to use digital twins also at the concept and design stage, conducting simulation modeling of new software products with phased changes until a satisfactory result is obtained. Information obtained on the basis of digital twins will be taken into account when creating a product. "Several organizations are already using digital twins. This technology is mainly in demand in those projects where the upfront costs are too high, and, consequently, the cost of failure," Voloshsky shares. SpaceTime Insight CTO Paul Hofmann says that digital twins use machine learning, which makes them more effective at predicting failures than condition-based maintenance models. "Thanks to IoT systems and machine learning, companies can be sure that their resources will not start to fail randomly,
Driverless cars, robots and home appliances
Household appliances, industrial equipment, cars and drones are getting smarter thanks to artificial intelligence and machine learning. Research firm Gartner estimates that by 2020, 61 million connected vehicles per year will roll off the production lines of automotive companies. “Entire economies grow here,” says Vince Jeffs, director of product strategy and marketing at Pegasystems. “For example, there are startups and already formed companies that are engaged in artificial intelligence, which are quite firmly established in the field of autonomous vehicles. MobileEye, for example, is a $500 million venture capital company that specializes in small cameras that fit all over the car. Similarly, there are companies that sell physical robots - for example, SoftBank Robotics specializes in hotel concierge robots. Their venture capital is $250 million.” Advances in deep learning have led to improvements in machine vision, natural language and speech processing, and the ability of machines and software to “seek rewards” and maximize performance,” says Wayne Thompson, Chief SAS data scientist: “The result is a new generation of machines that can see the world, hear and read natural languages, communicate with people, and self-regulate both mechanically and behaviorally, which is completely unprecedented.”Although for many, automation is a nightmare , depriving people of work, others, on the contrary, argue that that these technologies lead to a brighter and more humane future. "I'm often asked about the implications of automation," says Michael Hubbard, Global Director of ServiceNow. "Intelligent automation isn't a threat, it's an opportunity. stronger and more productive working relationships."
Virtual and augmented reality
After decades of hype, virtual and augmented reality is finally reaching a tipping point. And for those who want to develop products based on these technologies, new horizons are opening up: the expansion of sensations during games. "These technologies are not yet very widespread, but have become much more mature in recent years," Anup Nair, vice president and CTO of Mphasis Digital, shares his observations. "I believe that in the biomedical and healthcare industries, AR /VR will be of great benefit, both for education purposes and for sharing information about complex surgical procedures.We see plans for AR activities aimed at performing in-depth analysis in the social media control centers of large banks, as well as in the exchange rooms, where they will provide stockbrokers with an endless space for data analysis and collaboration.” Christian Sasso, Associate Professor of the VR/AR Qualification Program at San Jose-based Cogswell College, considers augmented reality to be the biggest technology trend of the year. “Very soon it will be possible to use AR to serve customers when a company-made device needs to be repaired,” Sasso says. “For example, in a project I’m working on, augmented reality glasses are used to contact a customer service consultant in the event of a TV or monitor failure. .When talking directly with the client through the augmented reality interface, a company representative can get all the necessary information by visually inspecting the broken screen, “AR and VR technologies will not be widely adopted until more affordable and high-quality hardware for them becomes available,” said Vishwa Ranjan, head of the augmented and virtual reality division at Infosys. “As early as 2017, we will see smartphone companies start developing augmented and virtual reality capabilities, such as face recognition, location, sensors and 360-degree cameras, which will play a significant role in promoting sales of AR and VR devices to first-time buyers". “AR and VR technologies will not be widely adopted until more affordable and high-quality hardware for them becomes available,” said Vishwa Ranjan, head of the augmented and virtual reality division at Infosys. “As early as 2017, we will see smartphone companies start developing augmented and virtual reality capabilities, such as face recognition, location, sensors and 360-degree cameras, which will play a significant role in promoting sales of AR and VR devices to first-time buyers".
Humanoid helpers
This is the next stage of artificial intelligence! We will end the clumsy tools now used to interact with the digital world. According to experts, soon we will soon use auxiliary tools not just for ordering goods and services online or searching the Internet. They will become extensions of our own brains. We will no longer need to memorize so much information: with the help of technological tools, we will free up resources for analytical and critical thinking. "What should we pay attention to if we are interested in the development of such assistive technologies? "The greatest demand now is for deep knowledge (deep knowledge)", says Günter Ohlmann, Head of Security at Vectra. that the key to this, even more than a degree in computer science, is a passion for mathematics and logic. "Creativity, a willingness to learn, a customer-centric mindset, resilience in the face of failure - the results of machine learning are by no means a finished product - and communication skills are critical social skills for engineers working in this field." that the key to this, even more than a degree in computer science, is a passion for mathematics and logic. "Creativity, a willingness to learn, a customer-centric mindset, resilience in the face of failure - the results of machine learning are by no means a finished product - and communication skills are critical social skills for engineers working in this field."
And the winner... is a combination of it all!
Although artificial intelligence is probably the most mentioned revolutionary technology of the current year, the most important trend is the confluence of rapidly developing emerging technologies. Maarten Ectors of Canonical lists more than a dozen different technologies that combine more than the sum of their parts: "cloud, mobile, IoT, artificial intelligence, transaction block chain, augmented reality, voice interfaces, software controlled radio communications, the 'fourth industrial revolution' [industrial automation and communication], robotics, edge computing and driverless vehicles." Patrick Spadding, of Rocket Software, says the individual technologies are coming together mainly because of the need for companies to get out of the rubble of their own data - for example, when analyzing website traffic. "When you add to that the growth of new data sources such as the Internet of Things," he says, "it's not easy to even keep up with the amount of information available to make business decisions." Spadding believes that the prospects for the merger of cognitive technologies, bots and machine languages will be greater, the more understandable they are. A new generation of digital natives will accelerate the adoption of these combined technologies, he says, as they expect usability, game-like interfaces, and the ubiquity of augmented and virtual reality.
GO TO FULL VERSION