Artificial intelligence (AI) is one of those technologies that excites the public and business imagination alike. Long since a favourite theme in science-fiction, it is now gaining traction in everyday practical scenarios.
In 2017, we saw a considerable rise in the adoption of AI around the world and across industries, with businesses using it to improve operations, generate new innovations and boost customer experience.
With financial services, telecoms and high tech leading the way in bringing AI into the mainstream, and other areas such as automotive, healthcare, energy and retail also embracing it, we expect the rapid growth of AI to continue in 2018 as companies strive to get the most value and competitive advantage from the data they capture. Let’s take a look at what’s been powering the rise of AI recently and what might be around the corner.
Data is the fuel for AI. As data collection, analysis and storage abilities dramatically improved over recent years, most companies found themselves with a huge potential resource and yet under equipped to wrestle with such high volumes of information.
In 2017, that started to change, as the people skills made headway in catching up with the technology. There remains a lot of complexity around how data is handled and used, but businesses are starting to see a deeper understanding of the specific skills needed to help companies bear fruit from data, and how these can be mapped onto ‘personas’ they can train up, or recruit.
Businesses are steadily learning how data scientists and AI developers work differently from traditional app developers, the tools they need, and how to bring them into app dev teams cohesively.
>See also: Cyber security and AI predictions 2018
For vendors in this space, the challenge is to make AI more easily accessible for all developers. Some are building machine learning frameworks to help organisations apply AI across use cases.
In 2018, vendors and enterprises alike need to continue to broaden their skill set, training existing developers and bringing new data scientists on board. To target the shortage of data scientists (and other data worker personas) at its roots, as a society we need to engage and inspire young people to boost demand in the education system, as well as supply that education. With a continued push expected this year and beyond, AI can fulfil its potential ‘superpower’ status for developers, helping them tap rich new sources of innovation.
One use case that has become more prevalent in the last twelve months is chatbots. Chatbots’ increase in popularity stems from businesses’ desire to give users the same experience online as they would get in-store – be they retail banking customers, healthcare patients, retail shoppers, or whoever else.
>See also: AI: from hype to reality in healthcare
Chatbots are great because they can respond very quickly to customers and deliver personalised care by taking advantage of data analysis and algorithms to determine the category a user falls into.
Among others building out their chatbot strategy, Twitter just announced a new enterprise API to power chatbots and help enable “more natural conversational experiences”. Look out for more of this in 2018.
Machine learning (ML) frameworks
One of the biggest surprises in this area of 2017 was to see the consolidation of machine learning frameworks beginning. There have been AI and ML libraries and projects for decades, but in the past 5-7 years we’ve seen a lot more investment in building deep learning libraries and artificial neural networks. Some of the big players in this area have poured tremendous amounts of resource and efforts into this, pulling in scarce talent.
With these corporations having access to much greater amounts of infrastructure for storage and compute power than universities could supply, they have transitioned many researchers from the academic space to the corporate space. With such heavy investment, it was easy to expect these key players to want to keep their assets closed for a few more years in order to monetise the value they have amassed.
However, in 2017 we already saw companies starting to open up their frameworks with recent collaborations like Onyx from Facebook and Microsoft, and Gluon from AWS and Microsoft. Why? There is obvious benefit to be gained collectively from building together, such as faster innovation and more freedom of choice for developers, and there is also a large-scale push more generally in the direction of open source in the tech industry today. It could also be said that in the context of today’s huge wave of cloud computing, even these high value frameworks can simply be drivers of workloads on top of the cloud infrastructures these big players have.
The intelligent application
With almost 8.5 billion mobile connections globally and counting, (see GSMA data), many of us use a smartphone, if not multiple smartphones, daily. Probably half or more of your mobile applications will have AI functionality, either directly embedded into the app or supporting it in the back-end. For example, your keyboard learns how you – and everyone else – interacts with it to improve how it works.
If you want to buy something through an online retailer, it will make recommendations based on your purchase history as well as typical buying habits using an AI engine. Ride-sharing apps and navigation apps use AI to calculate how to connect various users on a route. Intelligent applications are likely to continue to gather steam in 2018.
Any new technology may bring pitfalls that have not been fully anticipated. We’re finding with AI that the algorithms are only as good as the training data they are using, which can have negative consequences. The issue of how humans can pass on their own biases and prejudices to algorithms with damaging results, in anything from crime prediction to language translation, became better known in 2017.
This should be given a lot more attention, and hope to see real action to address this challenge in 2018. Vendors need to think about how they can offer tools and tutorials to help less experienced data scientists, developers and businesses at large gain a better understanding of data and the human impact of AI. Together, we can develop a more structural approach to the problem.
To be continued…
The application of artificial intelligence and machine learning will solve business problems and bring new ideas to life to continue into 2018 as companies strive to get the most business value and competitive advantage from their existing data. Watch this space.
Sourced by Kim Palko, principal product manager, Data and Analytics, Red Hat JBoss, and Matthew Farrellee, Emerging Technology & Strategy, CTO Office,
Article by channel:
Everything you need to know about Digital Transformation
The best articles, news and events direct to your inbox
Read more articles tagged: