According to a new study of global consumers, as many as 95% of drivers could opt to use voice assistants in their cars within the next three years. However, providers of the technology have some way to go before realising that level of market penetration, with half of the same potential customers currently saying they do not trust voice assistants with the security of their personal data.

With the automotive sector facing massive disruption in coming years, companies in the market are increasingly hard pressed to find new ways to improve their bottom lines. In the UK in particular, the pressure is on for firms looking to find ways to improve their value creation. The UK endured a 7% fall in car sales throughout 2018, and while Brexit was predictably cited as a cause of this, the nation’s automotive sector is unlikely to have seen the worst of that particular storm yet.

Amid this turmoil, innovation is even more essential. Automation and AI offer vast implications for engineering, production, supply chain, customer experience, and mobility services which could well help the industry combat its flagging performance. For example, half of all consumers have already indicated that they would be willing to pay a premium to benefit from owning a self-driving car in the next five years. The problem is that a number of key hurdles regarding safety and privacy need to be cleared before such technology can truly have mainstream appeal.

Further illustrating this, a new survey from Capgemini has found that three-quarters of global consumers would be willing to pay more to have Alexa-style voice assistants installed in their cars, either now or in the future. The consultancy surveyed over 7,000 consumers who had used a voice assistant inside the car as well as 300 auto executives, revealing the 95% of Capgemini’s sample expected to adopt the use of the technology permanently in the coming three years. On top of this, 37% would already be willing to pay a premium for it, while 48% might be willing to in the future.

Immediate enthusiasm for voice assisted driving was highest in India, where 68% of respondents would already be happy to shell out extra for its addition to their cars. This was more muted throughout North America and Europe, however, with over a fifth of Germans stating they would never be willing to pay a premium for in-car voice assistants. The UK was similarly reserved about the technology’s potential, but a higher 32% still admitted they would be willing to pay up for voice assisted driving in the here-and-now – representing a major opportunity to the automotive sector as it looks to buck its downward trend.

However, there is no guarantee that the apparent popularity of voice assistants will manifest as sales. Consumers are concerned about a lack of cybersecurity as the number of breaches balloon. Businesses that fail to protect their customers’ information are also falling prey to hackers, which can steal business secrets and disrupt operations. This has already hamstrung a number of innovations which ailing market sectors have sought to deploy to improve customer experience, as seen with consumer concerns about supermarkets using their data to tailor their shopping trip.

Natural Language Processing

Now that natural language processing, the ability for a machine to “read” language like a human and then take action from that information, has advanced to the point that it’s helping with everyday tasks, every business should be figuring out ways to apply this technology to their own processes. Conversational interfaces can break down communication barriers between companies and their customers. This will, for example, make it easier to overcome language barriers and do business with others from around the globe while at the same time creating efficiencies and enabling a better customer experience.

Computer Vision

The segment of artificial intelligence that allows machines to “see,” analyze visual data and then make decisions about the visual input is called computer vision. With this skill, machines can identify, verify, classify, and detect objects as well as, or in some cases better than, humans. Computer vision is already relied on for many healthcare applications, used in autonomous vehicles and many other industries, from agriculture to finance to manufacturing. As more companies adopt this technology, it will alter and create new tasks for humans in the workplace.

Robotics

Are you ready to have a robotic co-worker? While this reality has been commonplace in the industrial world for some time, robots will continue to expand influence in the business world and in our homes. Many new business ventures are developing robots to support humans in several ways, from being companions for senior citizens to delivering products. Cobots (collaborative robots) are meant to assist humans in the workplace with either advice (based on crunching data), mechanical muscle, or other support. Robots can drive efficiencies and safety in the workplace. And with open standards for robotics, many individuals and organizations have the opportunity to develop and deploy robots by using open source tools such as Amazon’s AWS RoboMaker platform. 

Autonomous Vehicles

Now that every major automobile manufacturer is investing in self-driving technology, it won’t be long before level-five autonomous vehicles (fully self-driving without any human intervention) will be on our roads. One of the first applications will be in the haulage and logistics industry where vehicles such as Mercedes Benz's Arocs semi-autonomous trucks will transport goods on highways. We can expect that autonomous technology will also alter our current vehicle models, as well. Many experts believe that autonomous vehicles will make it no longer necessary for every family to own their own car. Ridesharing and self-driving taxi services will alter our daily commutes.

Extended Reality

Extended reality is the umbrella term to describe immersive technologies (virtual reality, augmented reality, and mixed reality) that merge the physical and virtual worlds. It's a growing market and is expected to reach $209 billion by 2022. There are many practical applications of extended reality for businesses from giving customers the ability to try before they buy in retail to training and marketing for any industry. Once the challenges of full-scale adoption for extended reality are overcome—privacy, cost of implementation, and technical and hardware issues—extended reality will be transformative in the business world, and we will see many more mainstream applications.

Smart Assistants and Chatbots

Most of us are quite familiar with smart assistants and online chatbots already. Virtual smart assistants can understand spoken language commands and can help us complete tasks through user input, location awareness, and by accessing online information. Often referred to as conversational agents, chatbots are similar to virtual digital assistants because they are meant to interact and communicate with humans. These virtual tools help companies deliver a positive customer experience, free up human personnel for other tasks, and streamlines service.

While there is no crystal ball to determine the timeline for when each individual will definitely be impacted or have their positions augmented by these technologies, the evolution has begun.

Source : https://www.forbes.com/sites/bernardmarr/2019/11/18/the-6-careerdefining-tech-trends-in-2020-everyone-must-prepare-for-now/?ss=ai#5046d2a9fbdd

1. Registration, administration and support of IoT devices.
2. Collection and contextualization of events generated by IoT devices.
3. Processing IoT events to record them, convert them to business transactions, turn them around as commands to IoT controllers or any combination of these functions.
4. Any special services relating to the hosting of IoT application components.

Register and support devices

Both Amazon's and Google's IoT core service portfolios and Azure IoT Hub register, manage and support IoT devices. This area deals with the IoT device community itself. By providing secure device communications, registration of new devices, decommissioning of ones taken out of service and other routine community tasks, these services unload the burden of the IoT device community from IoT applications.

The tasks can be implicit in the way an IT pro develops an IoT application, but admins and operations personnel can also add tasks with minimal effect on development.

Cloud apps offer many advantages over traditional applications.

Contextualize IoT application events

Contextualization means ensuring that the event associates with the state of the outside system it's originating from. Both developers and operations or administration personnel should regard IoT as an event source. Devices send signals in response to real-world conditions, and these signals then activate application processes. In some cases, an event is a self-contained signal or request, and, in others, the context of the event is important. Open gate signal, for example, is processed one way if the gate is closed but in another if it's open already.

Simple event flows might not need sophisticated context. They can be processed through a data pipeline, provided by Amazon's Kinesis; Google Cloud Pub/Sub or, for more flexibility, Cloud Dataflow; or Microsoft's Event Hub. These pipelines can connect to analytics if the primary goal is to understand event counts and can also be input into complex event processing applications. Some of this can be done with little or no custom development, but obviously more is possible if software can be adapted to the specific events in the flow. The big advantage of this approach is that it lets IT pros deal with massive amounts of data.

IT pros should review IoT cloud platforms before they start developing an application to maximize the advantages these tools can bring.

Contextualizing tools offered by cloud providers include AWS Step Functions, Google's Cloud Dataflow -- which also handles pipeline applications -- and Azure Logic Apps. These can be used externally to an application to create composite events from the raw events generated by IoT devices. However, the applications must be programmed for the composite events. If the applications already exist, it will be necessary to either use contextualizing tools to generate the expected composite events or change the software to accommodate new composite events.

Events are sourced from a cohesive real-world system and linked to the system's state, including how people or vehicles move within it. IT pros can diagram a complex physical system and then place events within the structure or query to obtain information on the structure based on events using tools such as AWS IoT Things Graph, Azure's Digital Twin and, with some more effort, Google's Cloud Dataflow. Diagramming is a somewhat complex task, but it doesn't require software development, and for some, it represents a transition into the processing task.

How IoT cloud platforms process application events

IoT cloud platforms support IoT event processing in a variety of ways. Some are specific to IoT and event processing, and some are designed for more traditional business applications. The difference between IoT event processing as a specialized cloud service and as a general business service is in the distribution and variability of the event flow. Events occur at a relatively steady pace -- not unlike the pace of transaction generation in an online system. Organizations can best handle events with container services managed in IoT cloud platforms or managed Kubernetes services.

Most organizations reported using AWS Lambda in a Cloud Native Computing Foundation 2018 community survey.

Using traditional cloud provider containers or IaaS for highly variable event processing will risk either under-resourcing peak periods or wasting capacity and money during slack periods. For periods of intense demand, microservice, function or lambda hosting is the best approach. For example, Amazon's Lambda, Google Cloud Function and Azure Functions are the appropriate cloud provider tools. These serverless platforms mean developers do not need to have the hardware and software themselves. Organizations don't pay for the services unless they're actually used. Pay-as-you-go makes event-processing services ideal for IoT applications that experience significant event-generation variability. Don't use them for regular events; it will end up increasing costs.

What special services do cloud providers host?

Admins might think that cloud provider's hosting processes are located in the cloud or not in one of the prior points. IoT cloud platforms host processing options like function and microservice or serverless computing. However, it's possible to host some of the cloud provider IoT features and tools on premises using another cloud provider tool for edge computing.

Amazon's Greengrass local hosting is extended for IoT in IoT Greengrass. This makes Amazon IoT able to support local handling of events even when WAN and cloud connections are irregular. Google's Cloud IoT Edge and Azure IoT Edge offers comparable capabilities for Microsoft cloud users. All these tools are likely to require some custom development.

The cloud is an important part of any IoT application that has to absorb events from widely separated sources or support a highly elastic volume of events. IT pros should review IoT cloud platforms before they start developing an application to maximize the advantages these tools can bring.

Source : https://internetofthingsagenda.techtarget.com/tip/4-ways-IoT-cloud-platforms-improve-application-development

Now Google is offering audio news through Google Assistant. It’s called Your News Update. By saying, “Hey Google, play me the news,” on any phone or speaker with Assistant enabled, you can hear the top news stories of the day, but also stories tailored to your interests.

It’s available today by updating settings on Assistant.

Google’s hypothetical user is a Pittsburgh Steelers fan who now lives in Chicago and follows the stock market. Their audio news feed might include an analysis of the Steeler’s game, stock market update, stories on the upgrades to the Chicago transit system along with the latest national headlines. The longer you listen you more long-form content and in-depth coverage you hear.

For now, it’s available in the United States and in English only and will be launched in other countries next year. News partners include The Associated Press, major newspapers and broadcast networks, and news websites.

PICTURE ALLIANCE VIA GETTY IMAGES

Source : https://www.forbes.com/sites/beasleydavid/2019/11/19/google-unveils-a-new-way-to-get-the-news/?ss=big-data#33d43d902a46

Properly integrating AI technology with the complicated and nuanced world of human behavior is a huge barrier to overcome, especially in applications that can make a difference between life or death. Apart from making self-driving cars safe enough for our streets, teaching AI how to comprehend the less quantifiable parts of life could give AI the ability to help humans in roles it previously could not handle, and could advance AI applications in general.

Driving amongst us

The new study, headed up by researchers from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL), aims to teach AI how to classify personalities based on tools used in social psychology. The team used a metric called Social Value Orientation (SVO), which represents the degree to which someone is selfish (“egoistic”) versus altruistic or cooperative (“prosocial”), and then taught the AI system to estimate the SVOs of different drivers. Classifying a driver as egoistic or prosocial tells the system how likely they are to drive aggressively (for example, running a red light) or to be more passive on the road (such as slowing down to let someone turn into the road). This allows an autonomous driving system to comprehend signals that humans understand from experience—such as what it means when someone is speeding close behind you in the outside lane, or leaving space in a queue of traffic.

In tests that simulated merging lanes and making unprotected left turns across traffic, the system could predict the behavior of other cars 25% more effectively than current self-driving systems. In the left-turn simulations, for example, the system knew when to wait for a selfish driver and when to turn ahead of an altruistic one. This is a promising glimpse at what the AI could do in real driving conditions, but it is nowhere near ready to be rolled out. Instead, the team suggest that a partial version could be used in the near term to complement existing in-car AI, for example to highlight an aggressive driver entering someone’s blind spot. Training AI with this rudimentary social awareness helps to overcome a major obstacle for autonomous vehicles (AVs) and other human-facing AI applications. Removing uncertainty as to what humans might do will make these AI systems more ‘confident’, and opens up a range of possibilities for AI that can interpret and predict human behavior.

Binary behavior

The ability to understand (or at least quantify) human behavior is still a major sticking point for AI, and the team hopes to extend their training set to include pedestrians, cyclists, and other road users to expand the system’s understanding. “Creating more human-like behavior in AVs is fundamental for the safety of passengers and surrounding vehicles,” says graduate student Wilko Schwarting, who was lead author on the new paper published in the Proceedings of the National Academy of Sciences (PNAS), “behaving in a predictable manner also enables humans to understand and appropriately respond to the AV’s actions.”

This aspect of the system, that AI behaving more like us will make it easier for humans to interpret its actions, implies a much wider application for this technology than just AVs. An algorithm that can not only understand human behavior, but can also react and display an appropriate “human” response would make AI much more versatile and far better suited to time-intensive care work. When caring for dementia patients, for example, an AI system similar to this could potentially detect when a person was behaving normally or when their mood was erratic, to ascertain when to call a human medical professional. Outside of service-based roles, industrial co-bots that could predict human behavior, could work more safely and effectively amongst humans, and even replace the most dangerous roles if social cues and signals were understood - the volume and urgency in a foreman’s command for example.

Human understanding

While the MIT CSAIL team may not have captured human behavior exactly—a feat which would be one step removed from artificial general intelligence (AGI)—their work toward quantifying whether someone is likely to act selfishly or altruistically will certainly help to advance AVs, and could allow AI to be used in far more sensitive applications than it can currently handle. Creating AI that can operate safely and autonomously amongst humans without misunderstanding social signals or acting too cautiously in an uncertain situation will make AI and particularly robots far safer and more versatile. This particular system may be in its infancy, but I look forward to seeing how far this level of understanding can be developed, and what applications it may serve in the near future.

Source : https://www.forbes.com/sites/charlestowersclark/2019/11/18/autonomous-cars-can-predict-how-selfish-your-driving-is/?ss=ai#59c3d91e265f

Miguel, you have spent a lot of time examining the AI & Ethics landscape. What did you find?

Konstanze Frischen: In the AI field, there is a lack of consideration for some of the core principles that went into constitutions around the world and inspired the Universal Declaration of Human Rights and the Sustainable Development Goals (SDGs). When you look at what principles most corporations, think tanks, or governments propose should underpin AI, you’ll see there is overwhelming emphasis on trustworthy AI — AI that is transparent, reliable, non-biased, accountable, safe, etc. This is indeed a necessary condition, but it basically means technology working as it should. And I am thinking: this is great, but it’s not enough.

Why not? 

Take genetic engineering: We all like to have genetic engineering that is trustworthy in that it works as it should; but that doesn’t imply it is okay to copy and paste pieces of genome to create chimeras. Same with AI. 

What then?

I argue that we need to move to a humanity centric AI. If AI is a real game changer, we must take into considerations the implications of AI for humanity as a whole - at present, and in the future. I call that solidarity. Yet, only 6 out of 84 “ethical AI” guidelines examined (by the Health Ethics & Policy Lab from the University of Zurich) mention solidarity as a principle.

Data is so highly valued today because of the essential role it plays in powering machine learning and artificial intelligence solutions. Training an AI system to function effectively—from Netflix's recommendation engine to Google's self-driving cars—requires massive troves of data.

The result has been an obsession with bigger and bigger data. He with the most data can build the best AI, according to the prevailing wisdom. Incumbents from IBM to General Electric are racing to re-brand themselves as "data companies." SoftBank's Vision Fund—the largest and most influential technology investor in the world—makes no secret of the fact that its focus when looking for startups to back is data assets. "Those who rule data will rule the world," in the words of SoftBank leader Masayoshi Son.

As the business and technology worlds increasingly orient themselves around data as the ultimate kingmaker, too little attention is being paid to an important reality: the future of AI is likely to be far less data-intensive.

At the frontiers of artificial intelligence, various efforts are underway to develop improved forms of AI that do not require massive labeled datasets. These technologies will reshape our understanding of AI and disrupt the business landscape in profound ways. Industry leaders would do well to pay attention.

Today, in order to train deep learning models, practitioners must collect thousands, millions or even billions of data points. They must then attach labels to each data point, an expensive and generally manual process. What if researchers didn't need to laboriously collect and label data from the real world, but instead could create the exact dataset they needed from scratch?

Leading technology companies—from established competitors like Nvidia to startups like Applied Intuition—are developing methods to fabricate high-fidelity data, completely digitally, at next to no cost. These artificially created datasets can be tailored to researchers' precise needs and can include billions of alternative scenarios.

"It's very expensive to go out and vary the lighting in the real world, and you can't vary the lighting in an outdoor scene," said Mike Skolones, director of simulation technology at Nvidia. But you can with synthetic data.

As synthetic data approaches real-world data in accuracy, it will democratize AI, undercutting the competitive advantage of proprietary data assets. If a company can quickly generate billions of miles of realistic driving data via simulation, how valuable are the few million miles of real-world driving data that Waymo has invested a decade to collect? In a world in which data can be inexpensively generated on demand, the competitive dynamics across industries will be upended.

Unlike today's AI, humans do not need to see thousands of examples in order to learn a new concept. As an influential Google research paper put it, "A child can generalize the concept of 'giraffe' from a single picture in a book, yet our best deep learning systems need hundreds or thousands of examples."

In order for machine intelligence to truly approach human intelligence in its capabilities, it should be able to learn and reason from a handful of examples the way that humans do. This is the goal of an important field within AI known as "few-shot learning."

Exciting recent progress has been made on few-shot learning, particularly in the field of computer vision. (The technique is called one-shot learning or zero-shot learning, respectively, when only one or zero data points are used.) Researchers have developed AI models that, under the right circumstances, can achieve state-of-the-art performance on tasks like facial recognition based on one or a few data points.

These advances remain mostly confined to the academic world for now. As small-data methods migrate from academia to commercial production in the coming years, however, they will fundamentally change the way AI is done—eroding the importance of big-data assets in the process.

"If you're doing visual inspection on smartphones, you don't have a million pictures of scratched smartphones," explained Andrew Ng, deep learning pioneer and former AI head of Google and Baidu. "If you can get something to work with just 100 or 10 images, it breaks open a lot of new applications."

A final AI method making important advances without the need for gobs of real-world data is reinforcement learning.

In reinforcement learning, an AI model learns not through brute-force data ingestion but through self-guided trial and error: it is let loose to experiment with different actions in a given environment, and it gradually optimizes its behavior as it receives feedback about which actions are advantageous and which are not.

One of the most widely-publicized AI breakthroughs in recent years was powered by reinforcement learning: DeepMind's defeat of the world's best human players in the ancient game of Go.

DeepMind's original model, AlphaGo, learned the game using a combination of historical data and reinforcement learning. But the truly remarkable achievement came with its more sophisticated successor, AlphaGo Zero. AlphaGo Zero was given absolutely no prior data other than the game's rules. With no other input, simply by playing against itself, AlphaGo Zero learned the game of Go better than any human or machine ever had: it defeated the original AlphaGo 100-0.

"Expert data sets are often expensive, unreliable or simply unavailable," explained the AlphaGo Zero team. "By contrast, reinforcement learning systems are trained from their own experience, in principle allowing them to exceed human capabilities and to operate in domains where human expertise is lacking."

Beyond board games, reinforcement learning is finding real-world applications in robotics, chemical engineering, advertising, and beyond. Reinforcement learning represents a novel approach within AI: rather than requiring massive preexisting datasets, it can generate its own data, learning as it goes. As it finds its way into commercial applications, reinforcement learning will represent yet another challenge to the orthodoxy of big data.

The world of artificial intelligence is in constant flux. As the field's frontiers advance at breakneck speed, methodologies that are cutting-edge today can become dated tomorrow.

The dominant AI paradigm at present is deep learning, which relies on up to billions of labeled data points to train neural networks to recognize patterns and make predictions. Because neural networks are so data-hungry, business and technology leaders have become obsessed with amassing the largest datasets that they can, hoping that data will be their ultimate competitive advantage in an AI-driven world.

But deep learning is a waypoint on the long road ahead for AI, not its final destination. To base one's long-term business strategy on present-day neural networks' massive data needs is to fail to appreciate the future paradigm shifts in AI that lie ahead. Recent advances in fields like synthetic data, few-shot learning and reinforcement learning make clear that as AI gets smarter in the years to come it is likely to require less data, not more.

These new paradigms will reset the AI landscape and redefine the terms on which companies will compete. For forward-thinking businesspeople and technologists, it will be a massive opportunity.

Source : https://www.forbes.com/sites/robtoews/2019/11/04/questioning-the-long-term-importance-of-big-data-in-ai/?ss=big-data#793383f82177