HOW AI IS TRANSFORMING CLOUD COMPUTING
Samira Kabbour
CMO
Table of Contents
Cloud computing has already permeated every facet of online activity. However, recent developments in AI and the increasing sophistication of programmers, presages a new age of cloud computing. This article looks at how that technology emerges and how its set to impact our lives.
Every tech guru knows about the potential of cloud technology and how it has already affected how businesses and citizens store data and existing workloads. But because the cloud is such a new technology, companies have to think about how it will continue to evolve over time. Trends like the rise of mobile in place of computers and the Internet of Things have made small changes to cloud technology. But now the big dream is how artificial intelligence (AI) could improve cloud technology just as cloud technology has improved AI development.
IBM, one of the biggest cloud companies out there, states that the fusion of AI and cloud computing “promises to be both a source of innovation and a means to accelerate change.” The cloud can help provide AIs with the information which they need to learn, while the AI can provide information which can give a cloud more data. This symbiotic relationship can transform our development of AI, and the efforts of cloud companies like IBM to delve into AI research shows that these are not empty words.
Can AIs finally speak?
One of the biggest transformation in AI development has been how tech companies can create AI which can finally learn. One well-known example of this new machine learning occurred earlier this year when an AI defeated the world’s best Go player. Instead of brute-forcing the best moves like Deep Blue did in chess about 20 years ago, the AI learned by playing millions of games with itself and figuring out strategies which even Go players had not considered.
Of course, machine learning has far more useful practical purposes than playing games. One of the biggest fields is in the field of conversation, where voice-responsive AI systems can respond to human commands.
While we already have personal assistants like Cortana which can respond to voice commands, technology companies are interested in developing AI systems which can learn new words and how to respond differently. While there is a still of lot of work to be done here as the disastrous Tay experiment from earlier this year shows, the goal is to construct an AI which can communicate like a normal human.
Cloud computing could help immensely with this goal. The many, disparate servers which are part of cloud technology hold the data which an AI can access and use to make decisions and learn things like how to hold a conversation. But as the AI learns this, it can impart this new data back to the cloud, which can thus help other AIs learn as well.
Combining AI and the Cloud
As noted above, the potentials of the AI and the cloud means that companies which specialize in one of those two are putting more work into getting involved with both. As CIO notes, many of these cloud AI technologies take on two forms. They are cloud machine learning platforms like Google Cloud Machine Learning which combine machine learning with the cloud but do not have deep learning frameworks, or they are AI cloud services like IBM Watson.
The latter in particular is interesting because of the numerous applications through which businesses can use AI cloud services. For example, Wired recently reported about how organizations are relying on IBM Watson to help fight cybercrime. This is not as simple as plugging Watson into a USB and letting it go to work. Researcher have to teach Watson various parts of how to deal with hackers and criminals, where it becomes steadily more effective as it stores information through the network.
What is incredible about this learning process is that while Watson knows so much, there is still an important role for humans to play. Watson could read far more reports than a human could, but he still makes basic, fundamental mistakes like thinking that “ransomware” indicates a place. The researchers help Watson and guide the data so that he learns to think correctly.
In every step of the way, the AI, the cloud technology and humans are all needed. The AI is needed to learn, the cloud technology is needed so that the AI can access more data about cybercrime than could be stored on a server, and humans help the AI when it makes mistakes. This sort of cooperation and technological development can apply to almost field which we can think of today, as well as some which we may not even conceive of now.
An AI and cloud fusion in the future
Combining AI, machine learning, and the data stored with technology means that both AI and humans can analyze and gather more data than ever before. Tech experts have indicated that 2017 could be the year when AI becomes a ubiquitous part of our daily lives, and AI capabilities will only be improved with the development of cloud technology.
So pay attention to companies like Google and IBM as they work on combining the two. The result will be a world which transforms how we view both AI and the cloud.