Where will artificial intelligence go in 2019? Experts open their minds to make predictions
Where will artificial intelligence go in 2019? Experts open their minds to make predictions
It is safe to say that artificial intelligence (AI) will continue to be at the peak of the hype cycle in 2018. But the next predictions also show that AI is becoming more practical and useful. AI will help automate certain tasks and enhance many other tasks. It will combine machine learning and big data to gain new insights, and more chat bots will be added to the enterprise.
1. As the automotive industry is experiencing large-scale disruption, existing OEMs (original equipment manufacturers) and Tier 1 (auto parts suppliers) are increasingly aware that they need to immediately adopt AI to help deal with the outside of the vehicle. The environment, but also the experience inside the vehicle. Semi-automatic and fully automated vehicles will require AI-based computer vision solutions to ensure safe driving, seamless switching with human drivers, and a rich travel experience based on passengers, drivers’ emotions, cognition and health. – Rana el Kaliouby, co-founder and CEO of Affectiva, a cloud-based facial emotion recognition analysis service
2,I think we will see a lot of “firsts”, including the AI system can directly explain their behavior, and no longer need external evaluation. Network neutrality is eroded by the emergence of increasingly personalized and optimized AI-driven content. The deep learning bubble will burst. Applying AI only to specific areas AI startups will no longer be affected by overvalued valuations, and those surviving companies will provide basic, provable progress for AI capabilities. We will also see at least one fatal car accident involving driver less cars, and realize that unmanned drivers comparable to humans require longer tests to mature. — Monty Barlow, Head of Machine Learning at Cambridge Consultants
3. AI will begin to answer the question “Why?” In the past few years, we have learned two things that early adopters interact with the AI system: 1) Humans are not good at knowing what AI can do. 2) AI is also not good at telling humans what it is doing. This frustrates the user. In the face of AI’s doubts, our only explanation is: “Because I said this.” In 2018, AI developers will no longer be annoyed by users’ desire to increase transparency. In order to gain the trust of users, the AI system is working hard to achieve this common goal. AI developers will begin to prioritize advanced forms of accountability, reporting, and system queries so that users can ask “why?” Response to specific actions.
4. Personalized dynamic pricing. We predict that major e-commerce sites (mainly fashion, electronics, food and pharmacies) will offer real-time pricing personalized services. Pricing for online and physical stores will be based on behavior, supply and demand, and competitiveness. Today’s dynamic pricing is priced based on different variables, especially those that are not related to the customer, and personalized pricing will reflect the unique quotes each shopper receives. Prices will change frequently to reflect individual bids. The online experience will be simulated offline, and all stores will be electronically priced. There is a price on the page or on the shelf, and then the “Your price” will be displayed, only the only price you can receive. – Real-time adjustment of online service service Personalized CEO Dan Baruchi
5. In the near future, we will see the battle between AI and AI for the first time. In general, war will turn from national actors, hackers, and humans involved in the process to AI. AI will be directed to attack other countries and companies, and humans may not be able to defend, so it is time to discuss ways to prevent AI from doing evil and to supervise them. — Chad Steelberg, CEO and Chairman of AI Startup Veritone
6. If you are in a software company and don’t consider adding AI to your product or service, you may be behind others. AI is like water or air around us, it will become ubiquitous and will be embedded in most of the software we use, whether we know it or not. ——Ed Sim, founder of Boldstart Ventures, an AI venture capital firm
7. By 2018, the AI will be packaged and offered to others in a way that does not require a degree in computer science. Our goal is not to create “singularities,” but to make sound judgments scalable. Observation mode, learning mode, application and test guessing, and finally inference. AI will evolve at a faster rate, but it will not be mainstream if there are no formats available and the results are available to other software and users. We will eventually see the emergence of APIs and client applications, which shows that we have reached this milestone. ——Mike Fitzmaurice, Vice President of Workflow Engine Nintex
8. Silicon Valley will not be the only innovation in this field. Several countries in the world are making heavy bets on AI, and it will become the real battlefield for future technologies. If a company intends to use AI as part of its future business plan, it is best to develop a long-term development plan that may include iterative reconstruction and subversion. AI will experience several periods of slow and rapid change. — Todd Thibodeaux, CEO and President of CompTIA
9. AI will not disappear in this hype cycle in a short period of time, and current advanced analytics solutions will continue to transform into AI solutions using machine learning and deep learning. Looking forward to 2018, the company is expected to invest in driverless car research and add assistive driving capabilities to new models. This is an example of how computer vision can take over a car when the driver shows signs of fatigue. We predict that companies that traditionally use statistical models as advanced analytics solutions (for example, improved forecasting) will invest in machine learning technologies based on adaptive solutions that acquire data internally and externally to improve their models. – Naresh Koka, Vice President of SPR
10. It is now possible to mix AI with real-time transaction data through a single platform. This opens up a new world of possibilities. For example, AI can help companies take advantage of fleeting opportunities on a platform, such as real-time optimization of the cost per unit when purchasing a variety of commodities, such as energy. On energy information platforms based on wind, solar and grid energy, AI enables companies to adjust their actions in the face of real-time cost volatility and use price changes to reduce energy consumption. This is just a use case, and AI and real-time transaction data can enable companies to take advantage of many of these short-term opportunities. Bob Renner, CEO of Liaison Technologies
11, AI technology deployed in enterprises will be people-centric and produce measurable business results. These technologies will enhance human wisdom and make us better human beings. Those who enhance human AI will be widely accepted, and they will have a positive impact on society, rather than trigger human fear of the machine. — Joshua Feast, co-founder and CEO of Cogito Corp
12. We expect AI investment in venture capital, technology and non-technical areas to continue to increase. This is the next step in the evolution of AI to unlock and leverage the full potential of data, whether it’s internal company data, connected to external industry sources, macroeconomic trends, or data from sensors and devices. We anticipate that insights derived from these data will be automated in 70-80% of the time through training and learning. But it will require the right human skills and a combination of feedback loops and technological advances. During this journey, we continue to need human expertise to continue to function, and we will see more focus on strategic decision making. — Subrata Chakrabarti, Vice President, Product Marketing and Strategy, A
13 the implementation of “intelligent automation” will bring the most direct results to the enterprise. As a result, many companies still rely on decades-old, law-oriented manual processes, although this creates bottlenecks in the digital world of commerce. Automation technology has evolved to the point where these manual tasks can be effectively taken away from humans. More importantly, we are now in a commercial user’s own stage of managing this process, rather than requiring full-time IT assistance. This means that we will see CIOs have more and more say in the business because they have built enterprise-level automation strategies. These strategies provide immediate value to the company and will lay the foundation for AI’s long-term success. — Dennis Walsh, President, Redwood Software Americas and Asia Pacific
14 we will see that AI technology drives business intelligence (BI) to exponential explode, no longer linear sustainable innovation, but truly disruptive innovation, usually this is what we see every few decades. To one innovation. In today’s BI and analytics world, it can take a lot of time and money to look up really large and complex data sets and find insights in large amounts of data. With the support of AI, BI will make greater progress. In 2018, companies can query very large amounts of data in milliseconds, enabling them to learn at a faster rate. Not only does this allow them to move toward greater business intelligence faster, but they can also move toward true “business awareness,” where AI will eventually “understand” business data rather than simply reporting.
15 the company will begin to hire people who can properly analyze algorithms. We will call these people “algorithm whisperers”. Next year, chat bots will help everyone – from being included in the phone to introducing a physical shopping experience. In the future, all products, services and business processes will be self-improving. — SAP Innovation Specialist Timo Elliott
16. Analysis and advances in AI will play an important role in the health care sector next year. Not only can patient population tools be optimized, but workflows that inpatients and outpatients need to go through can also be optimized. The era of EHR deployment is driving companies to modify, enhance and develop new care processes that fundamentally change the way they work, treat patients, and receive care in a healthcare environment. ——Citrix’s medical evangelist Christian Boucher
17. AI will accelerate the demise of simple order sales. By effectively clarifying business value, it can help sellers win more customers by improving sales methods. The AI-driven sales learning tool will provide sales representatives with action advice, micro-training, and instant content based on customer needs assessments, sales representative skills and experience assessments, and competitive dynamics in the sales process, just like Netflix recommends movies.
18. General AI will take decades to become a reality. However, as call centers, financial and IT managers begin to shift conversational AI, image recognition and autonomous applications from pilot to real-world use, narrow-form AI applications will cause a huge stir in supporting enterprise functions in 2018. These applications will complement existing robotic process automation deployments, increasing the productivity of turbocharged employees and exceeding the speed of traditional industry benchmarks. — Stanton Jones, ISG Director and Principal Analyst
19, will be the year of the AI outbreak and the year of failure in its healthcare sector. AI has been used in large-scale “de-identification” data sets and has been plagued with many useful insights in areas such as responsible care and drug discovery. But when it came to “solving” personal care plans, AI failed. The main reason is that for today’s computers and algorithms, using all the data to provide treatment “automation” for one person is too complicated and mysterious. ——Frank Ingari, CEO of Growth Ally
20. I expect 2018 to be a more exciting year for companies to move toward “smart businesses.” More and more companies will no longer be limited to proof of concept, but will effectively start applying AI throughout the business. As machine learning algorithms mature, disruptive business models will emerge.Reputation Management services business They will force the entire industry to realize that digital transformation is not just a future trend, but also the key to staying competitive. At the same time, deep learning will be built as a standard machine learning commodity, but efforts are now being made to increase efficiency and availability within the system. Finally, we can expect further breakthroughs in intensive learning and will see the academic community further adapt to industrial research to ensure their competitiveness. — Markus Noga, Director of SAP Machine Learning
21. AI will drive demand for data quality. More and more companies are taking humans out of the “loop” and allowing AI to make real decisions, including pricing airline tickets and replenishing shelves. At the same time, the researchers found that the “black box” deep learning algorithm (which can’t be adjusted once it is accepted, or even impossible to understand) is the most effective for many problems. Since these algorithms are “useless data input, useless data output”, and because the results of “useless data output” become more and more important, high-quality training data will become a coveted resource, just like oil in the information age. . In the field of technology, the most acute human thinking may even shift attention from creating algorithms to providing the best data for these algorithms. — Aaron Kalb, co-founder and product director of Alation
22. The advancement of AI will lead to specialized cloud tools. As companies seek innovation and advances in machine learning and AI technology, the cloud will have more specialized tools and infrastructure to support specific use cases, such as the introduction of multi modal sensory inputs (sound, touch, and vision) in human-robot interactions. Solution, or combine satellite imagery with financial data to increase
December 25, 2020
December 25, 2020
December 22, 2018