OracleVoice: AI Is The Wave, And CIOs Must Learn To Surf

Jun 6, 2016 @ 06:00 AM
AI Is The Wave, And CIOs Must Learn To Surf
Share to email
Oracle
Tweet This
As compute power increases, so will the capabilities of machine-learning models. "> As compute power increases, so will the capabilities of machine-learning models.
potential repercussions of artificial intelligence, you should, too.
After years percolating in the backwaters of IT, AI is catching a dynamic wave of interest and investment. Self-driving cars and Go-playing computers have grabbed the public’s attention. But AI’s potential to dramatically improve cost/benefit equations is why “CIOs should start looking at how this is going to change the business they’re in and potentially disrupt the types of applications they’re using,” says IDC Research Director David Schubmehl.
Source: iStockphoto
If robotics is the face of AI, the beating heart is what’s known as machine learning—the ability to program a computer to recognize patterns and build models that let it make decisions or generate predictions. The use of machine learning algorithms in business isn’t new. They power such digital age stalwarts as recommendation engines and spam filters.
Several factors account for the recent explosion in the capability and use of machine learning. First is raw horsepower: Moore’s Law helped, but so did the introduction of GPUs, number-crunching chips built for video games that turned out to be especially suited to machine learning models. Second is the ability to store large amounts of data (enter the cloud). And third is the huge, escalating volume of data itself. Serendipitously, data is what machine learning systems feed on. “The more data you have, the better your model is going to be,” Schubmehl says.
With the help of long-gestating AI techniques such as neural networking, which imitates the functioning of the human brain, machine learning algorithms evolved into sophisticated “cognitive systems” capable of “deep learning,” enabling advanced functions such as natural language processing and image recognition. As AI’s capabilities grew, so did its use in business.
Early adopters in the financial services and insurance industries used machine learning models to do “predictive analytics,” which helped improve forecasting, recommendations, and risk analysis. Many New Economy companies, including Google and Uber, owe their business models to insight-generating, machine-learning-powered systems.
Today, HR, marketing, and other horizontal business functions use AI-based tools to support their expanding enterprise responsibilities. Machine learning models make recruiting and talent management applications faster, smarter, and more accurate. Marketing’s “customer experience” mandate calls for deep and wide data nets, such as so-called “ customer journey maps ,” that play to machine learning’s strengths.
On the negative side, the potential for AI to usurp relatively sophisticated job roles has a few critics, and more than a few employees, worried. Predictions that AI will eliminate all lawyers are probably just wishful thinking. But predictive analytics systems are dramatically cutting the time and effort that lawyers and their clerks spend combing through emails and examining documents. Such litigation-support tools can make a single lawyer as effective as “teams of lawyers and teams of paralegals,” Schubmehl says.
The Future of IT
To a remarkable extent, AI is the future of IT. The market for cognitive systems will exceed $40 billion by 2020, according to IDC. Predictive capabilities will be added to most kinds of enterprise software, helping business analysts, salespeople, and logistics experts sift through reams of structured and unstructured data to recommend non-obvious courses of action.
And that includes infrastructure software as well as business apps. Oracle is incorporating machine learning algorithms into every aspect of its Oracle Management Cloud , “fundamentally changing the way customers interact with these services,” says Oracle Senior Vice President Amit Ganesh. Oracle Management Cloud provides tools that monitor and report on IT infrastructure activities such as network log-ons and application performance. “These algorithms can continuously learn an application’s behavior and proactively provide meaningful insights when there’s a deviation from the norm,” Ganesh says. “This shifts the burden from users asking the right questions to these applications finding the right context-sensitive answers that the user needs to know.”
As compute power increases, so will the capabilities of machine-learning models. In the next year or so, look for a “hockey stick increase in compute-power-for-unit-dollar, more than we’ve ever seen,” says Toby Redshaw, a technology consultant and former CIO of American Express. That performance boost—made possible in part by smarter, tighter integration of software, hardware, and underlying storage—will empower companies to “do things with data they never thought possible,” Redshaw predicts.
Meantime, the data spigot is about to open wider as well, as the sensor-laden devices, appliances, machines, and other elements of the Internet of Things generate orders of magnitude more information. And that big data explosion will make machine learning algorithms that much more valuable.
“Because you now have enough of that data, you can use machine learning effectively,” says Fred Davis, a tech entrepreneur who helped launch one of the first Internet search engines, Ask Jeeves . “You can get insights that you might never have gotten from human-designed algorithms because the computer can actually find patterns in these things.”
Among other AI trends CIOs need to stay on top of:
AI in the cloud. Several providers offer AI tools as cloud services, giving small and midsize businesses access to this cutting-edge technology. At the same time, AI techniques are being incorporated into cloud services to increase their effectiveness and appeal. For instance, Oracle recently acquired a startup that will add machine learning capabilities to its data-as-a-service offering, such as matching data to an individual using multiple digital devices.
AI as open source. Many AI tools, such as Alchemy and H20.ai , are available as open source. The stakes got raised last year when Google open-sourced its AI engine, TensorFlow ; Facebook open-sourced the blueprints for its AI hardware server, Big Sur ; and Tesla founder Elon Musk launched OpenAI , a not-for-profit research lab intended “to advance digital intelligence in the way that is most likely to benefit humanity as a whole.”
AI in security. Spam filters use “supervised” machine learning models, where the algorithm is given parameters to follow in its data analysis. Unsupervised machine learning models, though they’re more complex and require more horsepower, promise to expand the pattern-recognition capabilities and, therefore, the effectiveness of security applications. “The whole hacking world is a data analysis problem that benefits from machine learning,” says consultant Redshaw.
Get Involved
So what should CIOs be doing right now with AI? First, get into the conversation.
“The people who are driving this conversation have not been the IT department—it’s been the line of business,” IDC’s Schubmehl says. Reminiscent of how cloud computing first entered the enterprise, “in some cases CIOs are getting bypassed,” he says.
Second, look for help. “This stuff is conceptually simple but really, really hard to do,” Redshaw says. “Get a partner that has done a lot in this space so you can learn.”
Schubmehl recommends that CIOs create a specialized group within their IT organizations. “I know of a couple of large companies that have set up cognitive swat teams, AI swat teams, that are providing information and advice to the line-of-business units about this, acting as advisers under the auspices of the CIO,” he says.
Third, just do it. Entrepreneur Davis estimates that half of companies are “playing” with AI, trying to figure out how it fits into their enterprises, “but very few companies have actually started using it.” Likely jumping off points are financial decision-making, customer relationship management, and big data analysis, he says.
Where AI lands on a CIO’s priority list depends on the company’s industry, financial resources, technical skills, and strategy. But CIOs delay investing in AI at their peril, Redshaw says. All other competitive factors being equal, if you don’t at least get acquainted with AI-based business technologies, he says, you’re “going to be the guy at the gun fight with a knife.”
Learn more on Oracle.com:
Internet of Things Will Need AI to Work Correctly
">
CIOs take note: When even the White House wants to take a serious look at the potential repercussions of artificial intelligence , you should, too.
After years percolating in the backwaters of IT, AI is catching a dynamic wave of interest and investment. Self-driving cars and Go-playing computers have grabbed the public’s attention. But AI’s potential to dramatically improve cost/benefit equations is why “CIOs should start looking at how this is going to change the business they’re in and potentially disrupt the types of applications they’re using,” says IDC Research Director David Schubmehl.
Source: iStockphoto
If robotics is the face of AI, the beating heart is what’s known as machine learning—the ability to program a computer to recognize patterns and build models that let it make decisions or generate predictions. The use of machine learning algorithms in business isn’t new. They power such digital age stalwarts as recommendation engines and spam filters.
Several factors account for the recent explosion in the capability and use of machine learning. First is raw horsepower: Moore’s Law helped, but so did the introduction of GPUs, number-crunching chips built for video games that turned out to be especially suited to machine learning models. Second is the ability to store large amounts of data (enter the cloud). And third is the huge, escalating volume of data itself. Serendipitously, data is what machine learning systems feed on. “The more data you have, the better your model is going to be,” Schubmehl says.
With the help of long-gestating AI techniques such as neural networking, which imitates the functioning of the human brain, machine learning algorithms evolved into sophisticated “cognitive systems” capable of “deep learning,” enabling advanced functions such as natural language processing and image recognition. As AI’s capabilities grew, so did its use in business.
Early adopters in the financial services and insurance industries used machine learning models to do “predictive analytics,” which helped improve forecasting, recommendations, and risk analysis. Many New Economy companies, including Google and Uber, owe their business models to insight-generating, machine-learning-powered systems.
Today, HR, marketing, and other horizontal business functions use AI-based tools to support their expanding enterprise responsibilities. Machine learning models make recruiting and talent management applications faster, smarter, and more accurate. Marketing’s “customer experience” mandate calls for deep and wide data nets, such as so-called “ customer journey maps ,” that play to machine learning’s strengths.
On the negative side, the potential for AI to usurp relatively sophisticated job roles has a few critics, and more than a few employees, worried. Predictions that AI will eliminate all lawyers are probably just wishful thinking. But predictive analytics systems are dramatically cutting the time and effort that lawyers and their clerks spend combing through emails and examining documents. Such litigation-support tools can make a single lawyer as effective as “teams of lawyers and teams of paralegals,” Schubmehl says.
The Future of IT
To a remarkable extent, AI is the future of IT. The market for cognitive systems will exceed $40 billion by 2020, according to IDC. Predictive capabilities will be added to most kinds of enterprise software, helping business analysts, salespeople, and logistics experts sift through reams of structured and unstructured data to recommend non-obvious courses of action.
And that includes infrastructure software as well as business apps. Oracle is incorporating machine learning algorithms into every aspect of its Oracle Management Cloud , “fundamentally changing the way customers interact with these services,” says Oracle Senior Vice President Amit Ganesh. Oracle Management Cloud provides tools that monitor and report on IT infrastructure activities such as network log-ons and application performance. “These algorithms can continuously learn an application’s behavior and proactively provide meaningful insights when there’s a deviation from the norm,” Ganesh says. “This shifts the burden from users asking the right questions to these applications finding the right context-sensitive answers that the user needs to know.”
As compute power increases, so will the capabilities of machine-learning models. In the next year or so, look for a “hockey stick increase in compute-power-for-unit-dollar, more than we’ve ever seen,” says Toby Redshaw, a technology consultant and former CIO of American Express. That performance boost—made possible in part by smarter, tighter integration of software, hardware, and underlying storage—will empower companies to “do things with data they never thought possible,” Redshaw predicts.
Meantime, the data spigot is about to open wider as well, as the sensor-laden devices, appliances, machines, and other elements of the Internet of Things generate orders of magnitude more information. And that big data explosion will make machine learning algorithms that much more valuable.
“Because you now have enough of that data, you can use machine learning effectively,” says Fred Davis, a tech entrepreneur who helped launch one of the first Internet search engines, Ask Jeeves . “You can get insights that you might never have gotten from human-designed algorithms because the computer can actually find patterns in these things.”
Among other AI trends CIOs need to stay on top of:
AI in the cloud. Several providers offer AI tools as cloud services, giving small and midsize businesses access to this cutting-edge technology. At the same time, AI techniques are being incorporated into cloud services to increase their effectiveness and appeal. For instance, Oracle recently acquired a startup that will add machine learning capabilities to its data-as-a-service offering, such as matching data to an individual using multiple digital devices.
AI as open source. Many AI tools, such as Alchemy and H20.ai , are available as open source. The stakes got raised last year when Google open-sourced its AI engine, TensorFlow ; Facebook open-sourced the blueprints for its AI hardware server, Big Sur ; and Tesla founder Elon Musk launched OpenAI , a not-for-profit research lab intended “to advance digital intelligence in the way that is most likely to benefit humanity as a whole.”
AI in security. Spam filters use “supervised” machine learning models, where the algorithm is given parameters to follow in its data analysis. Unsupervised machine learning models, though they’re more complex and require more horsepower, promise to expand the pattern-recognition capabilities and, therefore, the effectiveness of security applications. “The whole hacking world is a data analysis problem that benefits from machine learning,” says consultant Redshaw.
Get Involved
So what should CIOs be doing right now with AI? First, get into the conversation.
“The people who are driving this conversation have not been the IT department—it’s been the line of business,” IDC’s Schubmehl says. Reminiscent of how cloud computing first entered the enterprise, “in some cases CIOs are getting bypassed,” he says.
Second, look for help. “This stuff is conceptually simple but really, really hard to do,” Redshaw says. “Get a partner that has done a lot in this space so you can learn.”
Schubmehl recommends that CIOs create a specialized group within their IT organizations. “I know of a couple of large companies that have set up cognitive swat teams, AI swat teams, that are providing information and advice to the line-of-business units about this, acting as advisers under the auspices of the CIO,” he says.
Third, just do it. Entrepreneur Davis estimates that half of companies are “playing” with AI, trying to figure out how it fits into their enterprises, “but very few companies have actually started using it.” Likely jumping off points are financial decision-making, customer relationship management, and big data analysis, he says.
Where AI lands on a CIO’s priority list depends on the company’s industry, financial resources, technical skills, and strategy. But CIOs delay investing in AI at their peril, Redshaw says. All other competitive factors being equal, if you don’t at least get acquainted with AI-based business technologies, he says, you’re “going to be the guy at the gun fight with a knife.”
Learn more on Oracle.com: