Session Description: Blackboard is pleased to welcome Hilary Mason, noted Machine Learning expert and Data Scientist to the 2017 DevCon Keynote. In this session, Hilary will take us on a journey, beginning with the origins of Machine Learning and weaving through to the present state of affairs in the discipline, before embarking on a wild journey to where Machine Learning and Artificial Intelligence will go in the years to come. Attendees should expect to leave this session with knowledge of Machine Learning and how it is applied today, and excitement and insight into the future of smarter education through data-driven design and implementation.
Speaker: Hilary Mason, Fast Forward Labs, @hmason, fastforwardlabs.com
- We are able to build things now that 5 years ago we thought were never possible via Machine Learning!
- First big data project was census, in 1888 a patent was filed for counting people. A machine was developed “tabulating machine company” ended up becoming IBM. The punch cards persisted into the 1900’s.
- “Teenager Teaches A.I. to Rap like Kanye“
Fred Brooks – “In 1969, I went to te Provost, Prof. Charles J. Morrow, and I said, “We are ready and have the necessaries to build an intelligence amplifying system in which the mind and matching will cooperate on tackling hard problems. Who on this faculty most deservices to have his intelligence amplified?”
- 1952 Bell Labs, Claude Shannon, Machine Learning Mouse
- In 2017, the world is becoming computable.
- In many ways, what we are doing now isn’t unique but rather we can do it faster and easier.
- Artificial intelligence hold great potential for both students and teachers – but only if we used wisely. – The Conversation
- How AI is Already Changing Business – Harvard Business Review
- Scientific American – How people read about information on the social web. Based on clicks.
- AI? > Maching Learning > Data Science > Analytics > “Big Data”
- Going from counting to predicting with data = data science.
- “Big Data” – once you build th stack, it’s not just one thing, you now have a bunch of things = business intelligence, analytics, business intelligence, product data science, R&D
- Your data can offer the opportunity to see the future, and a lot of people ignore this…
- Google Maps – A map with traffic data is an example, you can make a decision after seeing the data and you don’t have to think at all about the technology behind the scenes… you just use it.
- When we as community build great products and that data fades into the background.
- The biggest opportunities may surprise you.
- What higher value problem than we have than education.
- Data science is a new professional role and a “team sport”.
- Executives and product managers must embrace data, too.
- We are augmenting human capabilities – not replacing a human with Machine Learning…
- Strategies: first, invest in automation to support an existing process, then, invest in scaling to allow entirely new products and applications.
- Often there are ethic issues with data that engieers are often not trained in.
- Data culture is vital! “Data Driven – Creating a Data Culture” Free book.
Tips for Machine Learning
- Research Activity
- Change in Economics
- A Capability becomes a Commodity
- New Data is Available (Wikipedia)
- Plus a bit of adventure, some coffee, and bad ideas…
Natural Language Generation
- AP’s robot journalists are writing their own stories now – The Verge
- The real impact will be in making complex data simple.
- Deep Learning – Tools like image analysis makes rich media computable.
- Summarization is another example, finding similar content across large about of text.
- Small messy data + intuition is useful!
It’s “intuition” AND “data” it’s not one or the other.