Web Directions ai 2017 Technology/Business/Design of AI

For those not versed in the technologies and approaches, machine learning and AI looks a lot like a magic black box, increasingly capable of understanding natural human language, transcribing speech and speaking back to us) and learning many of the things we do. But how is all this possible? What's going on behind the APIs?

In this session Alasdair Hamilton from Australian AI focused startup Remi.ai gives us the 5000ft view of the current AI landscape, to help us understand the foundations other technologies we'll discuss throughout the day.

Not too long ago, even exploring the feasibility of adding AI into a digital product would likely have required adding machine learning and data science expertise to a team, a not inconsiderable expense.But the rise of AI APIs like IBM Watson, Google Cloud Platform, and Azure and AWS AI APIs among others has made incorporating intelligence into products and services more accessible.

In this session, Justin Anderson from Ansarada shares insights from their recent research into the capabilities, strengths and challenges of a number of the most widely available AI API platforms.

So you've decided to explore adding intelligence to your product. Just what will be required? Is it the sort of feature that off-the-shelf AI APIs like those Justin Anderson has surveyed, or do we need to go deeper (indeed, is what we're trying to build even feasible, or is it still sci-fi?)

In this session, Andy Huang, Head of AI at Servian, walks us through an "AI decision tree", to help decide whether what you're imagining is feasible, and what technologies are most appropriate for making it a reality.

We'll wrap up this morning's session with a question and answer session. There'll be the chance to discuss the presentations at your table, then time for questions from the floor.

Have you ever sat down to write something, got stuck, and realised that you were watching cat videos instead of researching the secret life of plants? Well, you are not alone. Switching contexts between writing and research puts a big cognitive load on the writer, making this everyday task remarkably difficult for the average citizen.

Today, with the proliferation of AI technology services, any developer can easily tap into decades of hardcore cognitive science research via services like IBM Watson, AWS, Google Cloud and Microsoft Azure. Could these shiny new APIs help to break writer’s block, and reduce cognitive… oh look, a cute cat video!

Join Anna for a frank and fun exploration of Ephox’s foray into creating “one of the most practical applications of Watson technology” (IBM) for an audience of millions of Textbox.io and TinyMCE users.

In many places, AI is more or less synonymous with chatbots, whether text or voice based. But we've been doing voice based telephone systems for quite some time now, even if we've not thought of them in terms of AI.

Jennifer Wilson has been working with Inference Voice Recognition (IVR) systems since long before the current wave of AI enthusiasm. In this session, she'll share lessons learned about the user experience of such systems, and how we can apply them to today's voice driven chat interfaces.

Whether through WhatsApp, Messenger, iMessage or even plain old SMS, chatting has become perhaps the predominant human to human form of non in-person communication. So why not take advantage of the same modality for communicating with online products and services?

In this session, Henry Cho, Head of UX and AI for Upwire, takes us through key ideas in designing text-based chat interfaces.

Once again, we'll wrap up this session with a question and answer session. Hone your questions for our design leaders.

As with any emerging technology, making the business case for adopting and exploring the opportunities of AI presents potential challenges.

In this session, Gumtree's Magda Cortez will share operating principals and frameworks that will help you communicate investment into AI products.

There will be a healthy dose of skepticism, practical advice and proven approaches to turn concepts into products.

If, as has been suggested, "data is the new oil", the raw material for the current AI and machine learning boom, how do you get it? Fundamentally, startups don't have Big Data unless it comes from someone else. Yes they can science the data but how can they get the data and in so doing create a winning value proposition within a target market?

In this session, Chris Bayley, founder of Cover Genius, shares some of their experience building RentalCover.com and BrightWrite, two successful AI-based businesses within the insurance industry.

Trained on thousands of previous enquiries made by customers over the current live chat feature, UBank recently launched RoboChat. RoboChat complements their existing live chat interface, reducing the time it takes to apply for a home loan online, and is capable of providing answers to thousands of questions 24 hours a day.

In this session, UBank's Peter O'Malley will cover how their team built RobotChat, and the business value it brings, as well as sharing some lessons learned along the way that you can apply to many situations where live chat makes sense as a channel.

We'll wrap up the day with questions for our business focused speakers. There'll be the chance to discuss the presentations at your table, then time for questions from the floor.