AI News, AI News: Track The Latest Artificial Intelligence Trends And Leading ... artificial intelligence
Seasoned explorers: How experienced TMT organizations are navigating AI Insights from Deloitte’s State of AI in the Enterprise, 2nd Edition survey
They are more likely to have a companywide AI strategy, put a stronger focus on training, recruit both technical and nontechnical talent, show a strong capability to experiment, and consume AI through multiple means—for example, enterprise software, open-source tools, and AI-as-a-service.
Instead of asking themselves “Should we pursue AI?” many companies are now asking themselves “How should we pursue AI?” Enthusiasm and use are both increasing as organizations look to unlock the value that AI implementations can potentially provide.
Investment is following enthusiasm, and the global AI industry has received over US$24 billion in investments over the past three years.1 Additionally, technology giants acquired 115 AI-focused startups in 2017, 44 percent more than in 2016.2 The investment fervor isn’t limited to acquisitions: Organizations are looking to improve their own research and development capabilities as well.
This includes building a team of 1,000 AI engineers and researchers by 2020.3 Amid this wave, we wanted to understand how AI technologies are beginning to transform early-adopter companies.
In this edition, we wanted to take our analysis a step further and look at the most experienced AI adopters—those companies that have built multiple AI systems and show a high level of maturity in selecting appropriate technologies, identifying use cases, building and integrating AI solutions, and staffing.
To uncover what’s happening at the “leading edge of the leading edge,” we grouped organizations into three segments, based on the number of AI production deployments undertaken and how respondents rated their enterprise’s expertise across the various measures (see figure 1):
Seasoned companies overwhelmingly regard AI adoption as “very important” or “critically important” to their business success: A remarkable 94 percent of executives at Seasoned companies believe this, versus 82 percent of the Skilled and 52 percent of the Starters.
There are many paths to value offered by AI technologies, including short-term efforts to drive efficiency and others for longer-term transformation.7 Indeed, our three groups of TMT companies have differing views on how quickly their AI-powered transformations may take place.
Since Seasoned adopters view AI as a strategic imperative, it’s unsurprising that they are investing in AI more heavily than others and accelerating their spending at a higher rate (see figure 3).
The Seasoned firms are also serious about tracking metrics for their AI initiatives: About six in 10 are measuring and tracking their ROI as well as cost-savings targets, customer-related targets, and revenue targets, and half are tracking metrics for productivity targets.
Accurate, high-quality, curated data is the fuel that powers AI systems.8 Potential data struggles for organizations include locating and accessing the right internal and external data sources, cleaning and aggregating data that may come from disparate systems, and ensuring that it’s free of bias.9 Protecting the security and privacy of that data presents another critical challenge.
At the same time, 86 percent of respondents feel their organizations are somewhat or fully prepared for these risks—a figure so high as to suggest that organizations may not fully appreciate all the potential risks.
While Seasoned adopters are heavy AI experimenters, they’re generally guided by a strong organizational strategy: 58 percent say they have a comprehensive, detailed, companywide strategy in place for AI adoption—compared with just 33 percent of Skilled and 23 percent of Starters.
In a Forbes Insights executive survey, respondents cited a strategic vision for AI as the single most important contributor to successfully integrating the technology.14 Seasoned adopters are also undertaking a variety of activities that indicate their organizations’ emerging strategic vision for AI.
One example comes from Microsoft, which recently acquired the startup Lobe, with the goal of making deep learning simpler and more accessible, without coding.15 Tapping into AI-as-a-service is a key way that the Seasoned and Skilled fuel experimentation (see figure 8).
According to a recent Deloitte study, companies are 2.6 times more likely to prefer acquiring advanced innovation capabilities such as AI through cloud-based services versus on-premise solutions, and three times more likely to prefer as-a-service to building these capabilities themselves.16 Major cloud providers all now offer a variety of AI capabilities as services, including data preparation, speech-to-text, text-to-speech, natural language understanding, and image analysis.17 There are services for building chatbots and conversational interfaces, and for developing and training machine learning models, without having to learn complex algorithms.
Cloud-based deep learning services can provide access to specialized processors on-demand for handling the extremely heavy computational requirements.18 Service-based solutions give enterprises rapid access to the newest advanced technologies.
For example, Salesforce Einstein applies machine learning to historic sales data and predicts which prospects are most likely to close.20 These systems can be used off the shelf by employees who aren’t necessarily AI experts, and they’re regularly updated with technology advancements.
Seasoned adopters are more likely than others to use automated machine learning, now available via both commercial and open-source tools.21 —Ruchir Puri, IBM fellow, CTO, and chief architect for Watson, IBM Inadequate talent and skills often hinder AI adoption.
This is followed closely by a need for business leaders who can interpret AI results and take actions based on them, and by change management experts who can implement strategies to help people adapt to the organizational changes that AI is bringing.
Despite the potential disruption, the majority believe that AI technologies will have a positive effect on employees and newly added talent.23 Seasoned executives in particular are overwhelmingly optimistic: 92 percent agree that AI empowers their employees to make better decisions, 96 percent believe AI will enhance employee job performance and satisfaction, and 90 percent say that human workers and AI will augment each other, encouraging new ways of working.
recent Deloitte human capital trends report notes that, despite increasing clarity around the skills needed in a world where humans work side by side with machines, roughly half of respondents have no plan in place to cultivate these skills.24 There are some strong indications that the TMT AI adopters are approaching workforce training seriously and taking action (see figure 10).
The Seasoned organizations are aggressively educating their nontechnical workforce as well—two-thirds are training employees to take on alternative roles within the company, and nearly two-thirds are showing employees how to use AI in their jobs.
Today, organizations are using AI for areas ranging from IT optimization and driving automation and efficiency, to quality control and improving customer service, to creating new products and pursuing new markets.
3 ways retailers are winning with artificial intelligence
With better information, enabled by AI, retailers can create personalized experiences both online and in-store—and with NRF reporting that 79% of consumers make at least half of their purchases in-store, successful brick and mortar retailers are eager to exploit AI tools to improve the in-store experience.
In-store innovations include smart shelves that display detailed product information, multi-lingual robots that guide shoppers, and virtual fitting kiosks to ‘try on’ clothes.