Machine Learning is the new kid on the block.
This is of special interest to me since I specialized in that field 15 years ago to “unspecialized” three years later after being discouraged by the lack of real market of the discipline. Back then (early 2000’s) ML applications were always academically inspired with little true business value delivered. I could see the potential of the field, but I experienced hands-on the difficulties of truly implementing ML in business.
The tide seems to have turned with a few critical success behind us (e.g. descent voice recognition, handwriting recognition, automatic translation & more recently self-driven cars). One of the key success factor is the scale of data being able to be processed by modern capacity.
But like many new buzz Machine Learning has simply been steadily improving and just crossed a critical mass barrier.
One way to see this is to look at Microsoft’s history with the field of Machine learning. That article relates the beginnings, in 1992 at Microsoft Research, and the constant impact it had on Microsoft products: content-based spam detector, SQL Server Data Mining, Kinect and the rest. The most recent impact on Microsoft Products has of course been Microsoft Azure ML.
(A visible absent in the line-up of course is Clippy, Microsoft Office’s 1997 assistant I remember reading as the first true productization of ML at Microsoft)
As you go through this historic, you’ll realise the recent buzz truly is simply a line in the sand that has been crossed while progress had been constant if accelerated.