You are asked to prepare a 3-minute presentation of the paper, explaining about the technology (WHY -
Question:
You are asked to prepare a 3-minute presentation of the paper, explaining about the technology (WHY - what was the design motive behind, WHAT - its components, HOW- how it works).s a reference, see what an elevator pitchis for explaining an idea to a critical person (who could potentially fund your idea if you convince him/her), when you randomly meet him/her in an elevator.
Abstract:
Artificial intelligence (AI) is best leveraged for specific types of applications that will benefit the most from this technology. Some examples include fraud detection, predictive marketing, machine monitoring (Internet of Things), and inventory management.
Artificial intelligence (AI) is best leveraged for specific types of applications that will benefit the most from this technology. Some examples include fraud detection, predictive marketing, machine monitoring (Internet of Things), and inventory management.
Those are just some obvious examples. Many enterprises that use AI do so effectively, and thus waste money. Keep in mind that AI technology can be costly in terms of processing time and storage, cloud or not.
Back in 1985, AI was a shiny new object, and IT departments with big budgets over-leveraged AI systems. AI was applied to all sorts of use cases where it was not needed. We actually run that risk again today.
Artificial intelligence has become more affordable through the use of cloud platforms, making it a shiny new object once again. The affordable cost, coupled with cloud providers that promote artificial intelligence as having wide value, raises valid concerns that the technology will be misapplied. This already seems to be a pattern. Be aware that the value won't be realized if artificial intelligence is applied to systems that can't benefit from making predictions from patterns found in data, which applies to most applications within enterprises.
So, what's the bottom line with artificial intelligence and the cloud? Most big-budget IT shops in the late 1980s learned the expensive way that not all AI applications that can be applied should be applied. Now that cloud providers' offer artificial intelligence services within their public clouds, AI is within reach of most enterprise budgets. It's time for a new round of lessons. However, enterprises that look for applications for this technology can, in some cases, find game changers for the business. There is actual value there for businesses, if correctly applied.
Google Cloud artificial intelligence and Amazon artificial intelligence are the best examples of public cloud AI options. Both apply artificial intelligence technology within their respective clouds to drive interest in application development on their cloud services. Typical offers pair the ability to leverage artificial intelligence services on the cheap with big data management systems that provide the source of data, and thus the source of patterns.
To realize an effective return on investment, you need to consider all aspects of your requirements and how the public cloud provider can best meet those requirements. This goes beyond artificial intelligence to the way in which data, middleware, and analytical services work together to solve real business problems.
Artificial intelligence systems offered up by public cloud providers include software developer kits (SDKs) and APIs that allow developers to embed AI within applications. This bridges the gap between the capabilities of artificial intelligence and the actual application of this technology. An example would be the ability to determine if a loan application is fraudulent, based upon past and current patterns, as applied to the data that's within the loan application.
Artificial Intelligence Patterns
All artificial intelligence models are not the same. They are all defined to learn, but they provide different solution patterns. Most cloud providers, including Microsoft, Google, and Amazon provide support for three types of predictions. They have different names, but they boil down to three types: binary prediction, category prediction, and value prediction
Binary predictions deal with “yes” or “no” responses, i.e., does the order contain data that the artificial intelligence application previously flagged as fraudulent? Or, based upon data that comes from a recommendation engine that's AI-enabled, will a customer be likely to buy an “up sell” product?
The types of applications we leverage for these types of predictions are more numerous than the other types of predictions considering that the responses are much less complex: “yes” or “no”. Thus, these types of artificial intelligence use cases often find themselves in typical business processes, such as order processing, credit check systems, and recommendation engines used to recommend videos, music, or products to users based upon gathered data and learned responses. I found that the Finance and Manufacturing verticals are the industries that can benefit most from this aspect of artificial intelligence (see Figure 1).
Category prediction means that we can look at a data set and, based upon learned information, place that information into a particular category. This is useful when very different types of data are being analyzed, and a category should be applied so that data can be better understood and processed.
For instance, insurance companies place different instances of claims in specific categories based upon what's been learned over the years. An example would be to define the likely cause of an accident, even if the information is not a part of the data, such as “alcohol likely involved,” “likely fraudulent,” or “likely weather-related.” The AI system makes these assignments based upon past learning, such as the time of day that the accident occurred, as well as location, the type of damage done, age of driver, etc.
Category predictions have many different types of applications, such as when we need to place additional meaning around the data, and the direct correlation data is not in the existing database. Finance, manufacturing, and retail are all verticals that can use this category prediction-type of technology. We found that the Finance and Healthcare verticals are the industries that can benefit most from this aspect of artificial intelligence (see Figure 1).
Value predictions are more complex but also more insightful. They actually tell you quantitatively about likely outcomes from the data culled through, again, using learning models to find patterns in the data.
Say we want to find out how many units of a product are likely to sell in the next month. Good information to know, because since it allows tighter manufacturing planning, additional revenue will likely be generated as related to the objectives for the quarter plan, and perhaps the enterprise can economize the cost of travel as sales people follow up on leads. We found that the Finance and Healthcare verticals are the industries that can benefit most from this aspect of artificial intelligence (see Figure 1), but also Manufacturing and Government. The Government, in terms of defense operations, such as threat assessments, are good use cases. Manufacturing, in terms production optimization.
You need to keep in mind that AI and Machine Learning are highly dependent upon high volumes of quality data. Enterprises often have neglected the collection of data over time and will not find as much value with AI in the cloud.
How are they expected to build quality models based on history, if no single source of the truth for history exists? Most are building go-forward models but the samples are temporally very short. So how valid is the model? This is the key question you need to address before investing in AI.
Artificial Intelligence in the Clouds
There are many open source and proprietary artificial intelligence systems that have been around for years, and they support the types of predictions described above. However, the cost of these systems, in terms of hardware and software, was once prohibitive for most enterprises. Even if an enterprise could afford it, they rarely had the artificial intelligence talent required to design the prediction models and to deal with the data science as well. Cost and in-house talent was a prohibitive factor in 1985, when AI first hit the technology market, and it's still an issue in 2017.
Enter today's cloud-based artificial intelligence solutions. They are all very different, but with some commonality and some advantages and limitations.
First the advantages: These systems are cheap to operate. On average, you only have to pay a few dollars per hour to drive your very own AI application, such as the ones outlined above. This is perhaps the largest advantage of AI in the cloud, and is really bringing back AI as a core enabling technology.
Public clouds also provide cheap data storage. You can leverage true databases or storage systems as the input of the data into the artificial intelligence-enabled applications.
Finally, they all provide SDKs and APIs that allow you to embed AI functionality directly into the applications, and they support most programming languages.
The real value of AI technology is its use from within applications. For instance, the ability to determine in real-time if a loan application is likely fraudulent and to provide a process to immediately deal with the issue, perhaps allowing the applicant to fix any errors and resubmit. These types of predictions are more operations- and transaction-focused.
Now the disadvantages: The artificial intelligence systems that reside on particular public clouds are pretty much bound to those clouds. So, if you use an artificial intelligence system on cloud A, then the data storage mechanism on cloud A will typically be natively supported. However, your enterprise database is not supported unless you provide data integration between your on-premises data storage system and those in the cloud.
The key value for the cloud provider is that, if you're looking to take advantage of the native artificial intelligence system, it will be in your best interest to take advantage of the native storage systems and native databases as well. Also, the applications live better on the cloud platform if they will frequently talk to the artificial intelligence models which, in turn, often talk to the data. Get the hook?
Of course, if you're looking to move the data, applications, and other processes to the cloud anyway, you're fine. The artificial intelligence system can be accessed as a native cloud service.
However, if you're working with hybrid-or multicloud-based deployments, and most are, the separation of the data from the artificial intelligence engine will be problematic in terms of performance, cost, and usability. Clearly, AI could be a cloud provider's loss leader that is designed to attach enterprises more firmly to their cloud.
Although artificial intelligence is being sold as a shiny new tool, it's actually technology that's been evolving for years. Current IT economics now allow us to consider the applications of AI that can provide value to the enterprise and apply the power of AI when it makes sense.
Artificial intelligence is a new reality because of cloud computing. The danger here is the overuse of this technology for applications that frankly don't need AI. Perhaps we can do some predictive learning as well.