Thursday, December 28, 2017

AI SPECIAL.... Artificial intelligence is actually real

Artificial intelligence is actually real

The gazillion bytes of photographs and videos/frames have suddenly become a rich source of information.

A lot has been said about the impact of artificial intelligence (AI) and its ability to transform our lives. In the spectrum between enthusiasts and doomsday predictors, I am an enthusiast who believes that AI will transform business and decision-making.

While computers are great at rule-based programs, the human brain performs way better with pattern matching and intuition. With AI, though, machines are now getting better at pattern-matching and are fundamentally changing how we comprehend data but are still a far cry from how the human brain functions.

Several innovations are possible today because of advances in AI — not just in 
algorithms, but because data and computing power are both growing exponentially. Particularly exciting are the ways in which AI is impacting e-commerce.

AI-driven 
computer vision will have far-reaching implications for how commerce is conducted online and offline. Computer vision allows a granular understanding of images and videos/frames, the same way programs have been able to dissect text.

The gazillion bytes of photographs and videos/frames have suddenly become a rich source of information.

Feeding this information into clustering algorithms and recommendation engines, for example, gives insights not only to overall trends but also individual customer preferences.

Online assisted selling is also getting a huge boost through AI, bringing the online journey closer to the offline retail experience. Automating customer service and a personalised stylist experience through a bot implementation are improving.

While bots work better in more narrowly defined use cases — for instance, more in vertical e-commerce (fashion) than horizontal e-commerce — and there is a lot more to be done, the promise and potential are real. Virtual fitting rooms, through a combination of augmented reality (AR) and AI, are now becoming a reality, especially with AR software development kits (SDKs) becoming more commonplace.

Not only is AI enhancing customer experience, it is also rapidly driving efficiencies in e-commerce back-end operations. For example, enhanced efficiencies are being driven through the use of computer vision-based quality checks for sizing, product quality and colour matching. Another area of application is demand forecasting both at the aggregate level (for the buying, planning process) and at the regional level (forward deployment of the inventory for improved regional utilisation).

Given how technologies like AI are changing fundamental business processes, we need to prepare ourselves and our organisations to leverage the advantages.

First, senior management needs to spend real time to get educated on the topic. This would enable them to ask the right questions and engage more deeply and realistically with their teams. Next, we have to start thinking about data very differently. Data that seemed ‘un-analysable’ can now become valuable training sets that give rise to interesting insights.

For example, heat maps of browsing patterns were so difficult to analyse that it didn’t make sense to store them. An eyeballing exercise was carried out to make sense of the data. But with structured analysis, this information could help give us valuable insights and almost automate a user interface (UR) redesign for higher engagement.

Finally, getting the right talent is critical, especially as data scientists, engineers and product managers with relevant AI experience are a scarce resource globally. One needs to attract the right folks as well as create a culture for them to thrive in.
Doing these will set up both new age and existing businesses for success in the AI era.

(By: Ananth Narayanan; The writer is CEO, Myntra-Jabong)

ET11DEC17 

No comments: