Getty Images/iStockphoto
AI set to be cloud computing’s killer app
Major cloud suppliers are already offering artificial intelligence capabilities that are expected to account for a greater share of the cloud computing market in years to come
Artificial intelligence (AI) will become the killer application that will drive cloud computing forward, according to a top executive from the Asia Cloud Computing Association (ACCA).
Speaking at the Cloud Expo Asia conference in Singapore this week, Bernie Trudel, chairman of the ACCA, said although AI only accounts for 1% of the global cloud computing market today, its share of the overall IT market is growing at 52%.
“We’re starting to see AI having a significant impact on cloud computing,” he said. “If you extrapolate what the analysts are saying, there’s faster growth in AI, with 10% of cloud revenue expected to come from AI by 2025.”
Trudel noted that although major cloud suppliers – including Amazon Web Services (AWS), Microsoft, Google and IBM – are already offering AI capabilities, the market for cloud-based AI services is still in its infancy.
“AWS is certainly seen as the clear leader in cloud computing, but they’re really playing catch-up from an artificial intelligence perspective,” he said, noting that besides the Alexa intelligent assistant, AWS offers application programming interfaces (APIs) such as Lex for natural language processing, Polly for text-to-speech processing and Rekognition for image recognition and facial analysis.
Trudel said that AWS’s approach to AI was interesting, because the company also uses AI to help organisations easily determine the best open-source machine learning frameworks – such as TensorFlow or MXNet – to use for crunching different types of data.
He noted that AWS also offers publicly available datasets such as Common Crawl that contains petabytes of data collected over years of web crawling.
With Netflix as one of its biggest customers, AWS has also published a suite of Netflix-related applications and tools that can be used by its customers, according to Trudel.
Meanwhile, other cloud providers such as Google are not standing still. Trudel noted that the search giant has been doing a lot of work in AI, whether it is “helping developers choose the best algorithm for AI projects, using its Deepmind technology to build AI services or open-sourcing TensorFlow and Android to gain mindshare in AI developments”.
Read more about cloud and AI in APAC
- IBM’s Watson for Oncology will provide oncologists with medical information and treatment options to support clinical decision making.
- The Singapore government has been able to meet peak demands and address security issues through a cloud-based web-hosting platform powered by Amazon Web Services.
- Google’s Singapore cloud region is its third one in APAC, underscoring its ambition to challenge Amazon and Microsoft in a fast-growing public cloud market.
- Oracle is counting on its enterprise-grade cloud services to stand out from the crowded cloud computing market in the Asia-Pacific region.
- Pizza chain Domino’s is using a chatbot to take food orders in Australia, enabling workers to focus on what they do best – making pizzas.
In addition, Google is also using AI to improve the energy efficiency of its datacentres. “Now, national grid providers are also looking at leveraging the same [AI] models to drive efficiency in their networks,” he said.
Trudel said as the major cloud suppliers work towards delivering “general AI” services, beyond specific applications such as image recognition and speech recognition, the likes of Apple, Baidu, Alibaba, Intel, Tencent and Facebook are expected to join the fray with their own AI services.
Organisations across the Asia-Pacific region are already making use of AI to improve customer service through chatbots, plan treatment for cancer patients and prevent bus accidents.
Earlier this year, Japanese technology provider NEC said it was working with a Singapore bus company to predict the likelihood of bus accidents by analysing driver and telematics data using AI technology.
