3 Key Takeaways from the AI World Conference 2016
Recently, on November 7-9, 2016 the nation’s largest AI Business Conference, AI World was held in San Francisco. This year’s event was a great success with over 2,200 attendees and more than 65 sponsors and exhibitors. The conference and expo focused on how global 2000 businesses are using artificial intelligence machine learning and intelligent technologies to build a competitive advantage, drive new business opportunities and accelerate innovate efforts and gave executives the opportunity to learn more about how this technology can be used in their companies.
Conference founder and CEO of Trends Equity Eliot Weinman said, “We are extremely pleased to announce that we more than doubled our attendance expectations for our inaugural AI World 2016 program”.
Members of the eZdia team, Rahul Shah, CTO and Alok Jain, Co-CEO attended the event and spent their time interacting with industry leaders working on artificial intelligence and learned a lot about the AI industry.
3 Key Things Learned From the AI World Conference and Expo
1) Convert Calculations to Predictions
Today’s modern world includes a lot of automation and in many cases it begins with calculations based on certain parameters that never change. Information shared at the AI World conference indicate that there will soon be a movement away from these calculations and more towards predictions. In fact, artificial intelligence is enhancing the ability to make accurate predictions.
For example: Google maps traffic predictions, which constantly changed based on data that is continuously being updated or Google’s Smart Inbox that allows users to auto-reply to simple questions such as “when are you available for a meeting”? Simply by checking your calendars for available dates, Smart Inbox can automatically reply to these emails, saving users a lot of time.
Similarly, CrewMachine uses smart predictions to keep project managers informed of important information such as the probability of a project being completed on time based on the time commitments and availability of the resources. It can even predict how much money clients will have to pay in upcoming months for completion of due tasks.
2) Many Things Can’t be Coded They Have to be Learnt
Think about how a writer writes product descriptions. Information about specific attributes can’t be coded through the system, they have to be learned by people and their behavior. There are some tasks that require brain-based or cognitive abilities which have more to do with the mechanisms of how we learn. problem-solve, and pay attention rather than with any actual knowledge.
3) Machines are Good at Supervised Learning
Machines are good at supervised learning and a lot of innovation needs to happen in an unsupervised learning space. For example, when machine writes an eCommerce product description, it is trusting that the attributes are accurate and it is expecting the information in certain structured way. Along with all the data, the logic behind the data also needs to be provided and that’s supervised learning. For example, machine can be taught that when a person’s name ends with the letter ‘a’, it is probably a female name; it works as a logic behind the data.
As you can see, our team learned a lot at AI World and the expo itself was extremely successful. So much in fact that the organizers are planning to expand the conference and expo in 2017 and have already booked a date and place for next years conference.