Welcome to the first AI Review of 2018!
I’ve had a few weeks’ break from the Review, taking some time for my loved ones over Christmas and New Year, then wrestling against the January workload tsunami. But I’m back with a vengeance now and ready to kick off 2018 with a great chunk of AI news for you.
So let’s start with the five most socially-shared articles of the week.
As the article states, McKinsey & Co estimate that the market for vehicle-gathered data will be worth $750 billion a year by 2030. So it’s little surprise that Tesla is running ahead of the pack on this. A lot of the data gathered from the auto giant is effectively ‘crowdsourced’, coming in from Tesla vehicles and their drivers, using an array of sensors for picking up this info. What it then does with that raw data, involving AI, makes for interesting reading.
This is basically a blow-by-blow account of how to teach a neural network how to code HTML and CSS websites based on a simple picture of a design mockup. Depending on your interests and skillset, this will either fascinate or confuse you.
Pretty awesome infographic here, via CB Insights research. Definitely worth your attention, this one. But so is our own AI infographic.
Yes! This is the sort of Black Mirror-style stuff that really gets my nerd blood pumping. ML has been able to read MRIs and generate visualisations for a while, but this research from scientists at Kyoto University takes this one step further. The team’s use of deep neural networks now allows the decoding of more sophisticated hierarchical images compared to the simple binary images AI was capable of decoding before.
So, basically, these neural networks can actually detect objects within images. Whilst, to the layperson, the results may look somewhat rudimentary, the significance of this development is not to be sniffed at. Think how far this technology will have come in just a couple of years… and, indeed, where it might lead us.
A whole district in Beijing will become a technology park devoted solely to AI development research. Should you be under any misconceptions about how seriously China, or indeed the rest of the world, is taking AI, this should put those to bed. China is, of course, planning to be the world’s industry leader in AI, which may or may not be a good thing (considering some of their dubious practices with data on their citizens).
My Pick of the Week
That’s your lot for this week, and I’ll just leave you with my favourite Elon Musk quote of the week, his rebuttal of the ‘news’ he attended a Silicon Valley sex party: