Logo

The Data Daily

9 Indicators Of The State Of Artificial Intelligence (AI), May 2019

Last updated: 06-12-2019

Read original article here

9 Indicators Of The State Of Artificial Intelligence (AI), May 2019

US federal government contract obligations and AI-related investments grew almost 75% to nearly $700 million between fiscal 2016 and 2018 [Federal News Network].

85% of US CEOs and business leaders are AI optimists; 87% are investing in AI initiatives this year; 82% expect their businesses will be disrupted by AI to some extent within the next three years; 29% said AI will disrupt more than half of their business; 47% see China as the biggest obstacle to the advancement of AI in the US; 33% say employee trust is one of the greatest barriers to AI adoption [EY].

57% of IT and business executives (from Australia, Canada, China, Germany, France, UK and US) believe AI will transform their company within the next three years; nearly half have major concerns about potential AI implementation risks, but only 40% believe they are fully prepared for them; 85% of respondents in China expect that in two years AI will be very or critically important to business success—the highest level globally; cybersecurity is a top concern for US executives, second only to China [Deloitte].

29% of companies are making regular use of artificial intelligence, up from 24% in 2017; 19% say they have expert knowledge around AI; 29% classify their knowledge of AI as moderately high; potential uses for AI: improving workflows (52%), analyzing large datasets (51%), enhancing the customer experience (48%) and in security monitoring and detection (47%) [CompTIA].

29% of Line-of-Business owners spend $250,000 or more on AI [Figure Eight].

41% of organizations are planning to invest at least $500,000 to support AI initiatives over the next 12-18 months [ESG].

89% of customer service decision-makers in Canada, the UK and the US believe chatbots and virtual agents are useful technologies for personalizing customer interactions; 90% think it is important or very important to lead users through an automated dialog to clarify intent but only 47% of companies’ chatbots can do this today [Ada and Forrester].

Of those that have adopted an AI-driven marketing solution, 74% reported using AI in an “assistive” fashion, which surfaces insights for marketers to consider during manual decision making. Only 26% of marketers reported using autonomous AI, which can act on its own insights and work collaboratively with marketers (without adding manual work) [Albert and Forrester].

35,880 robots were shipped to North American companies in 2018, up 7% from 2017, with 16,702 shipments to non-automotive companies, up 41%. Notable growth came in areas like food and consumer goods (48%), plastics and rubber (37%), life sciences (31%), and electronics (22%) [Robotic Industry Association].

Nearly eight out of 10 enterprise organizations currently engaged in AI and ML report that projects have stalled, and 96% of these companies have run into problems with data quality, data labeling required to train AI, and building model confidence; only half of enterprises have released AI/ML projects into production; 78% of their AI/ML projects stall at some stage before deployment; 81% admit the process of training AI with data is more difficult than they expected; 76% combat this challenge by attempting to label and annotate training data on their own; 63% go so far as to try to build their own labeling and annotation automation technology; 71% report that they ultimately outsource training data and other ML project activities [Alegion and Dimensional Research].

Nearly two-thirds of New Zealand managers have no confidence or trust in big data, preferring to rely instead on their intuition and experience to make decisions; one-quarter also confessed they had only a modest knowledge of what big data is, or what it can do [Massey University].

38% of organizations investing in AI initiatives are doing so without a data scientist [ESG].

56.7% of technical practitioners believe they have enough data to support their AI initiative; 73.5% spend 25% or more of their time managing, cleaning, and/or labeling data [Figure Eight].

18% of companies say they are exactly where they want to be with their data practices [CompTIA].

Data workers waste 44% of their time each week because they are unsuccessful in their activities; they spend more than 40% of their time searching for and preparing data and, on average, use four to seven different tools to perform data activities; 88% of data workers, approximately 47 million people worldwide, use spreadsheets in their data activities [Alteryx and IDC].

73% of consumers in the United States, Canada, Japan, Australia, France and the United Kingdom think people using connected devices should worry about eavesdropping, and 63% think connected devices are “creepy” in the way they collect data about people and their behaviors. 88% agreed that privacy and security should be assured by regulators [The Internet Society and Consumers International].

C-level executives are increasingly and proactively targeted by social breaches – correlating with a rise of social-engineering attacks with financial motivation; one quarter of all breaches still associated with espionage; ransomware attacks still strong, accounting for 24% of the malware incidents analyzed and ranking #2 in most-used malware varieties [Verizon 2019 Data Breach Investigations Report].

46% of organizations that store customer personally identifiable information (PII) in the cloud are considering moving it back on premises due to data security concerns [Netwrix].

70% of employees in Europe said that they have noticed stricter policies at work regarding technology or customer data due to GDPR, compared to 41% in the US and 61% in APAC; only 39% of respondents worldwide feel their personal data is better protected since GDPR enforcement began and 6% believe their personal data is less protected than it was prior to enforcement; a majority of global respondents (74%) noted an increase in pop-ups or opt-ins requesting consent regarding the use of personal information; of those that saw an increase in pop-ups or opt-ins, 19% believe these consent requests negatively impact their productivity and 32% are increasingly annoyed by them; 74% of respondents also believe that the technology industry still needs more regulations [Snow Software].

Personally identifiable information (PII) and intellectual property (IP) are now tied as the data categories with the highest potential impact to 43% of respondents; PII is of greater concern in Europe (49%), most likely due to the recent enforcement date of the General Data Protection Regulation (GDPR); in Asia-Pacific countries, intellectual property theft is of greater concern (51%) than PII; 52% of respondents claim IT is at fault for creating the most data leakage events; 61% of incidents are being discovered by the security team, up 14% from 2015 [McAfee].

24% of government agencies failed a compliance audit in the last year; 98% of U.S. agencies are using sensitive data within digitally transformative technologies and 70% are not protecting the data with encryption tools; 60% of U.S. agencies have encountered a data breach, 35% of the breaches occurred in the last year [Thales].

36% of UK adults trust companies and organizations with their personal data more since GDPR came into effect one year ago; 47% have exercised some of their GDPR privacy rights and 57% are also more likely to use websites that have a certification mark or seal to demonstrate GDPR compliance [TrustArc].

New jobs created, especially for AI experts

43 million jobs have been created in the past 5 years across the OECD – “the lesson of the past half-millennium is that technological change complements jobs rather than destroys them” [The Economist].

Job postings for AI positions in the U.S. increased 159% over the past year [CompTIA].

56% of the top AI talent pool in America (38 out of 68 of the 113 authors of the 30 papers that made it to the oral presentation stage of NIPS 2018) is composed of foreign nationals who chose to work in the United States; when examining the country of origin of these immigrant scientists, the largest supply (10 or 26%) comes China [MarcoPolo].

About 14% of digital experts have the high-level AI skills—people with specialized knowledge of AI skills and the ability to teach what they know; 70% are willing to relocate for work, slightly more than other digital experts; some of the job factors that AI experts value the most differ on the basis of where they live: AI experts in North America, for example, place the greatest value on opportunities for learning and skills training and good relationships with their manager and co-workers; AI experts in Europe prioritize opportunities for learning and skills training, a good work-life balance, and good relationships with colleagues; and AI experts in Latin America place the highest value on job factors that help them get ahead, including learning and skills training, career development, and opportunities to lead and take responsibility [BCG].

AI funding more than doubled from 2016 to 2017 and more than tripled from 2016 to 2018 with a total of 3,434 investors and $62 billion total all time funding; top 5 AI domains in terms of cumulative funding worldwide through March 2019: Machine learning applications ($28.5B), machine learning platforms ($14.4B), smart robots ($7.5B), computer vision platforms ($7.4B), natural language processing ($6.7B) [Venture Scanner and Statista].

Top 5 most valuable AI Startups: SenseTime ($4.5B, China, security), UiPath ($3B, US, RPA), Automation Anywhere ($2.6B, US, RPA), YITU Technology ($2.365, China, security), Graphcore ($1.7B, UK, semiconductors) [CB Insights].

Deep learning techniques that use artificial neural networks have the potential to create between $3.5 trillion and $5.8 trillion in value annually across nine business functions in 19 industries [MGI].

The AI software market will reach $118.6 billion by 2025, up from $9.5 billion in 2018 [Tractica].

Asia/Pacific spending on AI will reach nearly $5.5 billion in 2019, up almost 80% from 2018, risg to $15.06 billion in 2022 [IDC].

Middle East and Africa spending on AI will reach $290 million in 2019, up 42.5% from 2018, rising to $530 million in 2022 [IDC].

The wearable AI market will surpass $185 billion by 2026 [Acumen Research and Consulting].

Retail sales from chatbot-based interactions are forecast to almost double every year to $112 billion by 2023 from $7.3 billion in 2019; retailers can expect to cut costs by $439 billion a year in 2023, up from $7 million this year, as AI-powered chatbots get more sophisticated at responding to customers [Juniper Research].

The AI in cybersecurity market will reach $38.2 billion by 2026, up from $8.8 billion in 2019 [ResearchAndMarkets].

The AI in manufacturing market will reach $16 billion by 2025 [Global Market Insights]

The AI in manufacturing market  will reach $18.5 Billion by 2025 [Zion Market Research].

The AI in Internet of Things (IoT) devices market will reach $9.5 billion in North America by 2024 [ResearchAndMarkets].

The mobile AI market worldwide will reach $22.4 billion by 2024 [Zion Market Research].

Bots defeat humans at yet another game—this time by working together: Engineers trained a total of 30 virtual gamers on a capture-the-flag game in the shooter Quake III Arena. During training, the digital players learned the rules of the game themselves. In later matches with professional, human game testers, the machines won roughly three quarters of the time [WSJ and Science Magazine].

A deep neural network that was trained for simple visual object detection has spontaneously developed  a human-like number sense [phys.org,ScienceAdvances].

Can a robot create art? Developed for the WSJ Future of Everything Festival, an autonomous robot spent two days painting advertising posters featuring the various event speakers including Martha Stewart, Trevor Noah and Jonathan Van Ness [The&Partnership and Traction3D].

In a pun contest pitting the AI against (human) humorists, AI beat humans only 10% of the time [Wired].

The country song “You Can’t Take My Door” was created by training a neural network to learn country music hits and then produce one of its own. The song was then arranged and performed by humans. The video below reflects all of the colorful imagery in the song [languidsquid.com].

“Duplex, which Google first showed off last year as a technological marvel using A.I., is still largely operated by humans. While A.I. services like Google’s are meant to help us, their part-machine, part-human approach could contribute to a mounting problem: the struggle to decipher the real from the fake, from bogus reviews and online disinformation to bots posing as people” [New York Times].

“The first challenge [for autonomous cars], no human safety driver, has not been met by a single experimental deployment of autonomous vehicles on public roads anywhere in the world” [Rodney Brooks].

The growing promise of AI in healthcare

The term “artificial intelligence” appears in nearly 90,000 publications from biomedical literature and research on the PubMed database [singularityhub].

Using reinforcement learning to determine appropriate amounts of morphine to give patients in intensive care [Arxiv].

The FDA has approved for the first time an AI-based chest X-ray solution, HealthPNX, an AI alert for pneumothoraces based on chest X-rays, from Zebra Medical Vision[Zebra].

A deep neural network was used to predict and identify the structural features that are associated with knee pain. The radiologist review revealed that about 86% of the cases that were predicted correctly had effusion-synovitis within the regions that were most associated with pain [Biorxiv.org].

A deep learning algorithm detects malignant lung nodules on low-dose chest computed tomography (LDCT) scans with a performance meeting or exceeding that of expert radiologists [Nature Medicine].

An explainable deep-learning algorithm for the detection of acute intracranial haemorrhage from small datasets [NIH].

Deep learning algorithms can identify with a great degree of accuracy whether a 5-day-old, in vitro fertilized human embryo has a high potential to progress to a successful pregnancy [Weill Cornell Medicine].

85% of people in the UK support the use of AI in diagnostics and treatment, and 86% say they were happy for their anonymized health data to be shared to better diagnose medical conditions [Microsoft].

The growing but limited practice of AI in healthcare

AI-assisted breast density measurements are already in use for screening mammograms performed at Massachusetts General Hospital (MGH), helping predict more accurately a woman's future risk of breast cancer [RSNA].

Walklake, a health checking robot, takes just 3 seconds to diagnose a variety of ailments in children, including conjunctivitis, and hand, foot and mouth disease. Over 2000 preschools in China, with children aged between 2 and 6, are using Walklake every morning to check the health status of their students [NewScientist].

US federal government contract obligations and AI-related investments grew almost 75% to nearly $700 million between fiscal 2016 and 2018 [Federal News Network].

85% of US CEOs and business leaders are AI optimists; 87% are investing in AI initiatives this year; 82% expect their businesses will be disrupted by AI to some extent within the next three years; 29% said AI will disrupt more than half of their business; 47% see China as the biggest obstacle to the advancement of AI in the US; 33% say employee trust is one of the greatest barriers to AI adoption [EY].

57% of IT and business executives (from Australia, Canada, China, Germany, France, UK and US) believe AI will transform their company within the next three years; nearly half have major concerns about potential AI implementation risks, but only 40% believe they are fully prepared for them; 85% of respondents in China expect that in two years AI will be very or critically important to business success—the highest level globally; cybersecurity is a top concern for US executives, second only to China [Deloitte].

29% of companies are making regular use of artificial intelligence, up from 24% in 2017; 19% say they have expert knowledge around AI; 29% classify their knowledge of AI as moderately high; potential uses for AI: improving workflows (52%), analyzing large datasets (51%), enhancing the customer experience (48%) and in security monitoring and detection (47%) [CompTIA].

29% of Line-of-Business owners spend $250,000 or more on AI [Figure Eight].

41% of organizations are planning to invest at least $500,000 to support AI initiatives over the next 12-18 months [ESG].

89% of customer service decision-makers in Canada, the UK and the US believe chatbots and virtual agents are useful technologies for personalizing customer interactions; 90% think it is important or very important to lead users through an automated dialog to clarify intent but only 47% of companies’ chatbots can do this today [Ada and Forrester].

Of those that have adopted an AI-driven marketing solution, 74% reported using AI in an “assistive” fashion, which surfaces insights for marketers to consider during manual decision making. Only 26% of marketers reported using autonomous AI, which can act on its own insights and work collaboratively with marketers (without adding manual work) [Albert and Forrester].

35,880 robots were shipped to North American companies in 2018, up 7% from 2017, with 16,702 shipments to non-automotive companies, up 41%. Notable growth came in areas like food and consumer goods (48%), plastics and rubber (37%), life sciences (31%), and electronics (22%) [Robotic Industry Association].

Nearly eight out of 10 enterprise organizations currently engaged in AI and ML report that projects have stalled, and 96% of these companies have run into problems with data quality, data labeling required to train AI, and building model confidence; only half of enterprises have released AI/ML projects into production; 78% of their AI/ML projects stall at some stage before deployment; 81% admit the process of training AI with data is more difficult than they expected; 76% combat this challenge by attempting to label and annotate training data on their own; 63% go so far as to try to build their own labeling and annotation automation technology; 71% report that they ultimately outsource training data and other ML project activities [Alegion and Dimensional Research].

Nearly two-thirds of New Zealand managers have no confidence or trust in big data, preferring to rely instead on their intuition and experience to make decisions; one-quarter also confessed they had only a modest knowledge of what big data is, or what it can do [Massey University].

38% of organizations investing in AI initiatives are doing so without a data scientist [ESG].

56.7% of technical practitioners believe they have enough data to support their AI initiative; 73.5% spend 25% or more of their time managing, cleaning, and/or labeling data [Figure Eight].

18% of companies say they are exactly where they want to be with their data practices [CompTIA].

Data workers waste 44% of their time each week because they are unsuccessful in their activities; they spend more than 40% of their time searching for and preparing data and, on average, use four to seven different tools to perform data activities; 88% of data workers, approximately 47 million people worldwide, use spreadsheets in their data activities [Alteryx and IDC].

73% of consumers in the United States, Canada, Japan, Australia, France and the United Kingdom think people using connected devices should worry about eavesdropping, and 63% think connected devices are “creepy” in the way they collect data about people and their behaviors. 88% agreed that privacy and security should be assured by regulators [The Internet Society and Consumers International].

C-level executives are increasingly and proactively targeted by social breaches – correlating with a rise of social-engineering attacks with financial motivation; one quarter of all breaches still associated with espionage; ransomware attacks still strong, accounting for 24% of the malware incidents analyzed and ranking #2 in most-used malware varieties [Verizon 2019 Data Breach Investigations Report].

46% of organizations that store customer personally identifiable information (PII) in the cloud are considering moving it back on premises due to data security concerns [Netwrix].

70% of employees in Europe said that they have noticed stricter policies at work regarding technology or customer data due to GDPR, compared to 41% in the US and 61% in APAC; only 39% of respondents worldwide feel their personal data is better protected since GDPR enforcement began and 6% believe their personal data is less protected than it was prior to enforcement; a majority of global respondents (74%) noted an increase in pop-ups or opt-ins requesting consent regarding the use of personal information; of those that saw an increase in pop-ups or opt-ins, 19% believe these consent requests negatively impact their productivity and 32% are increasingly annoyed by them; 74% of respondents also believe that the technology industry still needs more regulations [Snow Software].

Personally identifiable information (PII) and intellectual property (IP) are now tied as the data categories with the highest potential impact to 43% of respondents; PII is of greater concern in Europe (49%), most likely due to the recent enforcement date of the General Data Protection Regulation (GDPR); in Asia-Pacific countries, intellectual property theft is of greater concern (51%) than PII; 52% of respondents claim IT is at fault for creating the most data leakage events; 61% of incidents are being discovered by the security team, up 14% from 2015 [McAfee].

24% of government agencies failed a compliance audit in the last year; 98% of U.S. agencies are using sensitive data within digitally transformative technologies and 70% are not protecting the data with encryption tools; 60% of U.S. agencies have encountered a data breach, 35% of the breaches occurred in the last year [Thales].

36% of UK adults trust companies and organizations with their personal data more since GDPR came into effect one year ago; 47% have exercised some of their GDPR privacy rights and 57% are also more likely to use websites that have a certification mark or seal to demonstrate GDPR compliance [TrustArc].

New jobs created, especially for AI experts

43 million jobs have been created in the past 5 years across the OECD – “the lesson of the past half-millennium is that technological change complements jobs rather than destroys them” [The Economist].

Job postings for AI positions in the U.S. increased 159% over the past year [CompTIA].

56% of the top AI talent pool in America (38 out of 68 of the 113 authors of the 30 papers that made it to the oral presentation stage of NIPS 2018) is composed of foreign nationals who chose to work in the United States; when examining the country of origin of these immigrant scientists, the largest supply (10 or 26%) comes China [MarcoPolo].

About 14% of digital experts have the high-level AI skills—people with specialized knowledge of AI skills and the ability to teach what they know; 70% are willing to relocate for work, slightly more than other digital experts; some of the job factors that AI experts value the most differ on the basis of where they live: AI experts in North America, for example, place the greatest value on opportunities for learning and skills training and good relationships with their manager and co-workers; AI experts in Europe prioritize opportunities for learning and skills training, a good work-life balance, and good relationships with colleagues; and AI experts in Latin America place the highest value on job factors that help them get ahead, including learning and skills training, career development, and opportunities to lead and take responsibility [BCG].

AI funding more than doubled from 2016 to 2017 and more than tripled from 2016 to 2018 with a total of 3,434 investors and $62 billion total all time funding; top 5 AI domains in terms of cumulative funding worldwide through March 2019: Machine learning applications ($28.5B), machine learning platforms ($14.4B), smart robots ($7.5B), computer vision platforms ($7.4B), natural language processing ($6.7B) [Venture Scanner and Statista].

Top 5 most valuable AI Startups: SenseTime ($4.5B, China, security), UiPath ($3B, US, RPA), Automation Anywhere ($2.6B, US, RPA), YITU Technology ($2.365, China, security), Graphcore ($1.7B, UK, semiconductors) [CB Insights].

Deep learning techniques that use artificial neural networks have the potential to create between $3.5 trillion and $5.8 trillion in value annually across nine business functions in 19 industries [MGI].

The AI software market will reach $118.6 billion by 2025, up from $9.5 billion in 2018 [Tractica].

Asia/Pacific spending on AI will reach nearly $5.5 billion in 2019, up almost 80% from 2018, risg to $15.06 billion in 2022 [IDC].

Middle East and Africa spending on AI will reach $290 million in 2019, up 42.5% from 2018, rising to $530 million in 2022 [IDC].

The wearable AI market will surpass $185 billion by 2026 [Acumen Research and Consulting].

Retail sales from chatbot-based interactions are forecast to almost double every year to $112 billion by 2023 from $7.3 billion in 2019; retailers can expect to cut costs by $439 billion a year in 2023, up from $7 million this year, as AI-powered chatbots get more sophisticated at responding to customers [Juniper Research].

The AI in cybersecurity market will reach $38.2 billion by 2026, up from $8.8 billion in 2019 [ResearchAndMarkets].

The AI in manufacturing market will reach $16 billion by 2025 [Global Market Insights]

The AI in manufacturing market  will reach $18.5 Billion by 2025 [Zion Market Research].

The AI in Internet of Things (IoT) devices market will reach $9.5 billion in North America by 2024 [ResearchAndMarkets].

The mobile AI market worldwide will reach $22.4 billion by 2024 [Zion Market Research].

Bots defeat humans at yet another game—this time by working together: Engineers trained a total of 30 virtual gamers on a capture-the-flag game in the shooter Quake III Arena. During training, the digital players learned the rules of the game themselves. In later matches with professional, human game testers, the machines won roughly three quarters of the time [WSJ and Science Magazine].

A deep neural network that was trained for simple visual object detection has spontaneously developed  a human-like number sense [phys.org,ScienceAdvances].

Can a robot create art? Developed for the WSJ Future of Everything Festival, an autonomous robot spent two days painting advertising posters featuring the various event speakers including Martha Stewart, Trevor Noah and Jonathan Van Ness [The&Partnership and Traction3D].

In a pun contest pitting the AI against (human) humorists, AI beat humans only 10% of the time [Wired].

The country song “You Can’t Take My Door” was created by training a neural network to learn country music hits and then produce one of its own. The song was then arranged and performed by humans. The video below reflects all of the colorful imagery in the song [languidsquid.com].

“Duplex, which Google first showed off last year as a technological marvel using A.I., is still largely operated by humans. While A.I. services like Google’s are meant to help us, their part-machine, part-human approach could contribute to a mounting problem: the struggle to decipher the real from the fake, from bogus reviews and online disinformation to bots posing as people” [New York Times].

“The first challenge [for autonomous cars], no human safety driver, has not been met by a single experimental deployment of autonomous vehicles on public roads anywhere in the world” [Rodney Brooks].

The growing promise of AI in healthcare

The term “artificial intelligence” appears in nearly 90,000 publications from biomedical literature and research on the PubMed database [singularityhub].

Using reinforcement learning to determine appropriate amounts of morphine to give patients in intensive care [Arxiv].

The FDA has approved for the first time an AI-based chest X-ray solution, HealthPNX, an AI alert for pneumothoraces based on chest X-rays, from Zebra Medical Vision[Zebra].

A deep neural network was used to predict and identify the structural features that are associated with knee pain. The radiologist review revealed that about 86% of the cases that were predicted correctly had effusion-synovitis within the regions that were most associated with pain [Biorxiv.org].

A deep learning algorithm detects malignant lung nodules on low-dose chest computed tomography (LDCT) scans with a performance meeting or exceeding that of expert radiologists [Nature Medicine].

An explainable deep-learning algorithm for the detection of acute intracranial haemorrhage from small datasets [NIH].

Deep learning algorithms can identify with a great degree of accuracy whether a 5-day-old, in vitro fertilized human embryo has a high potential to progress to a successful pregnancy [Weill Cornell Medicine].

85% of people in the UK support the use of AI in diagnostics and treatment, and 86% say they were happy for their anonymized health data to be shared to better diagnose medical conditions [Microsoft].

The growing but limited practice of AI in healthcare

AI-assisted breast density measurements are already in use for screening mammograms performed at Massachusetts General Hospital (MGH), helping predict more accurately a woman's future risk of breast cancer [RSNA].

Walklake, a health checking robot, takes just 3 seconds to diagnose a variety of ailments in children, including conjunctivitis, and hand, foot and mouth disease. Over 2000 preschools in China, with children aged between 2 and 6, are using Walklake every morning to check the health status of their students [NewScientist].


Read the rest of this article here