AI World: Eight AI trends to watch in 2018

Computationally speaking, the big data analytics trend is not going away like a shooting star. As the amount of data continues to increase, the improvement of big data analytics will not stop. When it comes to predictive analytics applications, we have only seen the tip of the iceberg. Some organizations are using data mining, machine learning, and artificial intelligence technologies to analyze current data to better conduct business (e.g., predict sales, optimize marketing campaigns, etc.). All of these different types of artificial intelligence technologies have been closely combined to change our daily lives, and this change will continue.

Here are some of the key statistics on AI, big data, predictive analytics, and machine learning:

• By 2018, 75% of developers will use AI in one or more business applications or services - IDC
• By 2019, AI technology will be applied to 100% of IoT - IDC
• By 2020, 30% of companies will introduce AI to augment at least one major sales process - Gartner
• By 2020, algorithms will actively change the behavior of billions of workers worldwide - Gartner
• The artificial intelligence market will exceed $40 billion by 2020. - Constellation Research
• By 2025, AI will drive 95% of customer interactions - Servion

Eight AI trends to watch in 2018

​​

Trend 1 - Bigger companies will win in the future

Amazon, Google, Facebook, and IBM will lead the development of AI technology. As large companies, they have more resources to collect data and thus have more data to work with.

Here’s what the industry leaders are doing with AI:

Amazon:

• Investment in artificial intelligence has been going on for more than 20 years
• The crawled web data comes from more than 5 billion web pages
• In an operating Amazon distribution center, there are more than 500,000 JPEG images and corresponding JSON metadata files used to describe product information.
• Monitors over four billion records of broadcast, print and online news around the world every day
• Nearly 1 million images and videos, including audio and visual features and annotations
• Amazon's Echo dominates the voice assistant market, with a market share of over 70%.

Google:

• One of the largest data warehouses with 10-15EB of data - Cirrus Insight
• Focus on application and product development rather than long-term AI research
• A team of over 1,300 researchers - Google Brain
• Voicebot: 23.8% of the voice assistant market
• Using the open source platform TensorFlow to give anyone the opportunity to use the machine learning platform
• The size of the Google Earth database is approximately 3,017 TB or 3 PB - Google Earth Blog
• Google Street View has about 20 PB of street view photos - Peta Pixel

Facebook:

• Processing 2.5 billion pieces of content and over 500TB of data every day - Tech Crunch
• Facebook AI Researchers (FAIR) includes approximately 80 researchers and engineers - FAIR
• 2 billion "likes" and 300 million photos generated every day - Tech Crunch
• Scanning approximately 105TB of data every 30 minutes - Tech Crunch
• A 62,000-square-foot data center was built to accommodate 500 storage racks with 1EB of storage capacity.
• Translate 2 billion user posts in more than 40 languages ​​every day, and 800 million users can see the translated content every day - Fortune

IBM:

• Plans to invest $240 million over 10 years to create MIT Watson AI Lab - IBM
• More than 2,000 employees worldwide, more than 600 employees at the New York headquarters - IBM
• Watson client engagements span six continents and more than 25 countries - IBM
• IBM is investing $1 billion in the Watson Group, including $100 million in venture capital to support IBM startups and build cognitive applications with Watson - IBM
• More than 7,000 applications have been created through the Watson ecosystem - Fortune

Google is probably at the forefront of deploying machine learning for application and product development services. Not only is it the first company to conduct AI research, but it also has over 70,000 employees. Google is a very large company. In addition, Google Brain is a deep learning AI research project for which Google has an entire team. Google Brain research covers areas such as machine learning, natural language understanding, machine learning algorithms and techniques, and robotics.

​​

List of the 100 most promising artificial intelligence companies in the world

Trend 2 - Algorithms and technologies will be integrated

All the second-tier companies that have invested in AI (such as Intel, Salesforce, and Twitter) are following closely behind the companies with big data and starting to use their data algorithms and AI technologies. Data transactions will exist between industry users, and algorithms and technologies are likely to be integrated. Data transactions and the integration of algorithms and technologies will make AI more powerful.

As big companies like Google and Facebook continue to acquire smaller companies, the algorithms in the hands of smaller companies will be integrated into the core platforms or solutions of the big companies. Google acquired DeepMind, a London-based artificial intelligence company that built a general learning algorithm, in order to gain a greater commercial advantage than other technology companies. On the other hand, Facebook acquired Wit.ai to help its own speech recognition and voice interface. It also acquired artificial intelligence startup Ozlo to improve the technology of its M virtual assistant.

Trend 3 - The data crowdsourcing market will be huge

All AI companies are eager to obtain large data sets in order to realize their AI ambitions. These companies will use crowdsourcing to obtain large amounts of data. There are already many different ways to evaluate the quality and reliability of crowdsourced data, which not only allows companies to benefit from this data, but also gives consumers a guarantee.

"We live in a crowdsourcing culture where more and more people are willing and happy to share their knowledge through social media," said Joel Gurin, founder and editor of OpenDataNow.com.

Google is crowdsourcing a vast amount of images to build imaging algorithms. It is also using crowdsourcing to help improve services such as translation, transcription, handwriting recognition, and maps. Amazon is also using crowdsourced AI to improve Alexa’s more than 15,000 existing features.

Trend 4 - M&A, and more M&A

According to statistics from CBInsights, the race to acquire artificial intelligence companies has begun. In 2018, we will see more mergers and acquisitions for intellectual capital and talent. All small companies in the field of machine learning and artificial intelligence will likely be acquired by large companies for two main reasons:

AI cannot work independently without data sets. Since large companies have large data sets, small companies do not have much competitive advantage.

Algorithms without data are of no use. Without algorithms, data is of little use. Data is the core of algorithms, and it is very important to obtain a large amount of data.

“If data is the fuel, then algorithms are the engine,” said Hod Lipson, a robotics engineer and director of Columbia University’s Creative Machines Lab.

​​

Trend 5 - Democratizing tools to gain greater market share

Large companies will open source their algorithms and tool sets to gain a larger market share. Market-based barriers to data and algorithm acquisition will be greatly reduced, and new applications of AI will increase. By democratizing tools, small companies that previously had limited or no access to AI tools will have access to large amounts of data to train and launch complex AI algorithms.

Google CEO Sundar Pichai talked about democratizing AI: “One of the most exciting things we can do is to demystify machine learning and AI and make them accessible to everyone.”

In addition, frameworks, SDKs, and APIs will become the standard for all major companies to guide consumer usage habits. SaaS and PaaS-based models will become the business models followed by all these companies.

Trend 6 - Human-computer interaction technology will improve

Siri and Alexa are probably the two most popular human-computer interaction tools. More robot-based solutions like them will become entry-level products for AI companies. For example, computers can currently be used for voice analysis and facial recognition, but in the future, computers will be able to identify a user's mood based on his tone of voice, which is called sentiment analysis.

Solutions in manufacturing automation and non-consumer focus areas will be the first to improve. Improvements in manufacturing automation are mainly attributed to labor cost savings from the use of complex technologies including automation, robotics, and advanced manufacturing. In 2018, improvements in non-consumer solutions will be prevalent, such as human-machine interaction technology in agriculture and medicine.

Trend 7 - AI will definitely impact all vertical industries

Manufacturing, customer service, finance, healthcare, and transportation have already been impacted by AI. Self-driving vehicles are expected to be available in 2018. Next year, AI will impact even more vertical industries, such as:

• Insurance - AI will improve the claims process through automation
• Legal - Natural language processing can summarize thousands of pages of legal documents in minutes, reducing time and increasing efficiency
• PR & Media - AI can increase the speed of data processing
• Education – Development of virtual tutors; AI-assisted essay grading; adaptive learning programs, games, and software; personalized education courses powered by AI will change the way students and teachers interact
• Health - Machine learning can be used to create more sophisticated and accurate methods for predicting how long a patient will be ill before they develop symptoms

The Industrial Revolution changed almost everything 100 years ago, and artificial intelligence will change the entire world in the next few years.

Trend 8 - Security, Privacy, Ethics and Moral Issues

Everything under the umbrella of artificial intelligence, including machine learning and big data, is vulnerable to new security and privacy issues. Sometimes, it is critical infrastructure that plays a role. Security needs related to privacy issues, such as keeping bank accounts and health information private, will rely more on security research. 2018 will be the year when security and privacy issues are resolved and there will be new developments.

The ethics of AI will also be a major focus in 2018. Ethical and moral issues that need to be addressed include whether AI can harm or benefit humans. There are concerns that robots may replace humans, especially in areas that require empathy, such as nurses, physiotherapists, and police officers. Another issue to deal with is autonomous weapons. Consider a certain level of autonomous functionality, where AI should take over certain functions of a weapon, rather than humans fully controlling the weapon.

Our recommendations

Although AI has been around for many years, AI as we know it today is still in its infancy. There is a lot of hype around AI and its applications, from autonomous vehicles to virtual personal assistants and many other technologies that require human intelligence to complete tasks. While there are a large number of AI use cases, most of which are improvements to specific processes, it will take time to successfully deploy them. In addition, there are not many companies in the AI ​​industry, so fragmentation will not appear for the time being, but unstructured data and algorithms to process this data will appear. The road to AI is long and arduous.