• During IBM Think, IBM made several AI-related announcements, some designed for enterprises with complex requirements, and others geared towards helping businesses deploy their first AI solution.
• Although IBM’s new capabilities and tools in support of deep learning are impressive, and position IBM as a thought leader, it’s the steps IBM is taking to help companies just getting started with AI that truly move the market forward.
IBM Think was promoted as an event that would bring together the greatest minds in AI. It featured technologies such as virtual assistants, machine learning (ML), and deep learning (DL), and also touched on hot button issues such as ethics and AI. During her keynote, CEO Ginni Rometty discussed the transformational role that AI will have on the IT market going forward, and she introduced Watson’s Law, describing it as a follow on to Moore’s Law and Metcalfe’s Law.
With these assertions, IBM set the bar high when it came to IBM Think 2018 and its impact on the AI market. So did IBM move the proverbial needle when it comes to AI? In order to truly evaluate the symposium’s impact, we need to first gauge where we are now – what exists today and what is lacking – and where the market needs to go. The question is, did IBM Think move us any closer to where we aspire to be?
What Exists Today
There is no dearth of AI-based solutions or vendors in the market today. Numerous players, from North America, to Europe and Asia, offer a variety of solutions that incorporate AI technology. Well-known market leaders, such as SAP, Amazon, and Microsoft offer platforms that include natural language processing, image recognition, and solutions in support of ML. Indian systems integrators have launched AI platforms, such as Wipro Holmes and Infosys Nia, and are using them to deliver more insightful and actionable vertical solutions to customers. Players such as Salesforce have incorporated AI into offerings to improve sales processes and extract customer insights. And countless specialized vendors provide point solutions, such as chatbots or industry specific applications.
But let’s consider where businesses are falling short when it comes to AI. Of course, there are the innovative early adopters in every industry, but most companies are still in search of their first AI project. They are exploring the various solution and vendor options, observing how others in their industry are implementing AI, and trying to gauge how to adopt AI in a way that will yield the most effective business results. As organizations work towards educating themselves on AI, most are struggling with a well-known market challenge: a skills gap. Many simply do not have the internal talent to confidently evaluate and incorporate AI into their business processes. In order for the industry to gain momentum, this hurdle needs to be overcome. Of course there are other challenges, such as access to data and processing power, but when it comes to AI, improving ease of adoption and ease of use will go a long way in moving the broader market forward.
Where We Aspire to Be
Ideally, we are striving to be in an environment where all businesses enjoy the enhanced insights, customer support, efficiency and productivity that AI promises to deliver. Businesses will have the internal expertise, such as data scientists and developers, as well as the tools to quickly create customized models or AI-infused applications themselves, or they will be able to choose from a range of AI solutions that are so easy to use that even lines of business professionals can implement them. We aren’t there yet, but the steps taken by vendors to point us in this direction have the potential to greatly shape our journey.
IBM Think AI Announcements
During IBM Think, IBM made several AI-related announcements, some designed for enterprises with complex requirements, and others geared towards helping businesses deploy their first AI project. Below are just a few of the noteworthy announcements IBM made during the show:
• Deep Learning as a Service. With this new service, IBM makes it easier for businesses with limited AI resources to develop deep learning models by automating the fine-tuning step of the deep learning process.
• IBM and Apple Partnership. The two companies expanded their existing partnership by announcing that developers can now build and train models using Watson, convert the model into Core ML, and insert them into an Apple app.
• Data Science Elite Team. IBM created a new team that works onsite with customers to help them identify AI use cases that can be up and running within weeks; the service is available free of charge.
• Watson Assistant. Watson Assistant is similar to Alexa or Siri in concept, but is designed to be incorporated into products and services offered by enterprises.
• IBM Watson Data Kits. IBM’s new kits include industry specific data that developers can incorporate into their AI models, speeding the development and deployment of new models. .
• NVIDIA Partnership. IBM and NVIDIA demonstrated the potential of combining IBM POWER9 servers with NIVIDIA Tesla V2100 BPUs, resulting in faster processing.
For additional analysis on these announcements see IBM Think 2018: Putting AI to Work with Data, Lots of Data
Did IBM Move the Needle?
Although IBM’s new capabilities and tools in support of DL are impressive, and position IBM as a thought leader, it’s the steps IBM is taking to help companies just getting started with AI that truly move the market forward. As the industry forges ahead, it needs the cutting-edge solutions that expand AI’s vision and potential, but it shouldn’t undermine the less glamorous announcements – those that are designed to help businesses still on the fence – because they are the ones with the potential to have the greatest impact overall.