Why hasn’t anyone lost their jobs yet?
Why hasn’t any company other than google and facebook used it in a significant capacity?
What are neural nets ?
It’s our first step to making “Artificial Intelligence” or machine learn things, on their own.
Say you show a neural nut a bunch of apples, it will know what an apple looks like in all situations.
many jobs are lost every day, much more than are generated - it’s not something to make spectacle about (check stats from Mercedes & BMW, just a tip of an iceberg)
on the other hand, nobody want’s to be held accountable if AI officially fucks up (look what’s happening to Musk)
it’s just the 4th stage of industrialization
practices & exercises are done in 3rd world still
just enjoy the blissful ignorance
Ai that write Ai spell the end of many jobs.
Many jobs simply won’t need to exist.
say you can 3d print a set of robot arms like these -> in 1 or 2 runs, with minimal assembly
using tech like this -
Now imagine the robot arm / chef thing is also a 3d printer and can tend to a garden.
there goes 80% of all jobs.
(robots grow and prepare food for people at home)
now imagine you 3d print open source hardware and can also directly recycle the hardware and print it again.
‘transmutation’ 
Input energy -> sort materials -> 3d print new device.
Digital Alchemy.
yeah, a Detroit’s dream, flying cars & a fountain of youth…
yet not more than 10% will get the chance to taste it, while others will never own a home, neither a land, nor water… or if they get lucky they’ll get a momentary illusion of being a child again
you’ll print it and then it will seat on the shelf… 90% done
When you “say something to your computer,” that’s a neural-net in action. It is a most interesting technology that Ray Kurzweil (who also invented music synthesizers and flat-bed scanners) has done a lot to study and to popularize. The essential idea is to try to mimic how the brain works, with its network of neurons. Unsurprisingly, this research has also taught us a lot of things about how brains do, actually, work.
If you’re that sort of nerd – or you just like to buy nerdy t-shirts – http://kurzweilai.net is a fun (and nerdily informative) place to visit.
Neural-net technology has proven to be very good at things like recognizing speech, identifying parts of photographs, and other applications where it proved to be quite difficult to write a procedural algorithm to do the same thing. (We tried that conventional approach first.) On the other hand, it also makes mistakes.
If you’re the sort of nerd – as I am – who likes to try to figure out how to make “an inanimate chip of highly-refined sand” do something useful and to do so all by itself, then the subject is actually quite engaging. We observe the thing doing something and (maybe) doing it correctly, but we can’t point our finger directly at the reason why it does it. 