Computer vision is where it all began. In 2012 Geoff Hinton and his team built a neural network that massively out-perfomed all other approaches to recognising images. This breakthrough started the deep learning revolution.

Now 10 years later, this technology is being commercialised by a group of hugely successful startups. The applications attend to be specific to particular industry and range from agriculture, retail, insurance and construction. Some examples:

• Ceres Imaging can accurately identify crops that need more or less water
• Standard makes systems for checkout free shopping
• Tractable a London based company that allows instant damage estimates for insurers

I have not been hugely involved in computer vision, but it is interesting to see its progress from experimental breakthrough technology, to an increasingly stable and mature set of tools, through to eventual commercialisation over about 10 years.

Making these businesses work is not about a general purpose model or algorithm so much as finding a niche application and specialising in it. I expect a similar trajectory for other areas (like NLP) over the next 5-10 years.

Why this matters: AI has long overpromised and under-delivered - that might be beginning to change.