A 5min introduction that highlights what it is, the challenges to scale, the impacts on society – both bad and good
The technical term for building input A→ output B software is supervised learning.
So what can A→B do? Andrew Ng: Here’s one rule of thumb that speaks to its disruptiveness:If a typical person can do a mental task with less than one second of thought, we can probably automate it using AI either now or in the near future.
Eg — is this image a cat? is this behavior suspicious?…
Challenge for scale:
1-It requires a huge amount of data.
In 2012, the New York Times reported that a cluster of 16,000 computers dedicated to trained itself to recognize a cat based on 10 million digital images.
2- computational power
3- Other issue: talent. there is currently a war for the scarce AI talent that can do the work.
According to a recent survey by the World Economic Forum, by the year 2020, automation, and AI will lead to a loss of 7 million jobs in 15 countries. The irony is that a majority of the jobs lost will be for people who can least afford it.
the same survey does highlight a creation of 2 million new, high-paying jobs in AI. That’s less than what it would destroy.
there are other societal benefits that go beyond the job market – for example in medicine.