More on this book
Community
Kindle Notes & Highlights
Read between
August 16 - September 16, 2019
In short, open data has tremendous potential, and governments have good reason to expand the range of data they release.
But policymakers also need to work with companies and citizens to address issues of privacy and security.[1298]
In 2016, the EU created a landmark law called the General Data Protection Regulation (GDPR).[1306] Under this law, companies who are breached have to pay fines of up to 20 million Euros or 4% of their annual revenue, whichever is more.[1307]
As data protection laws spread, some insurance companies are starting to offer data breach insurance.[1315] Like with normal health and auto insurance, companies would pay a small amount every year, and in exchange the insurer would cover the costs if a devastating data breach happens.[1316]
A self-driving car needs two things: information about its environment and a strategy for how to move through it.[1322]
Self-driving cars come with a huge collection of sensors and data
“inertial navigation systems,”[1324]
Waymo.
LIDAR,
Then it removes plans that would bring it too close to an obstacle, and it ranks the remaining plans by safety and speed.
The learning car
machine learning
at its simplest: a computer making predictions based on observed patterns.[1333]
Ford invested $1 billion in the self-driving software startup Argo AI
GM bought a similar startup called Cruise Automation.[1339]
Google could make Uber and Lyft’s lives even harder by restricting access to Google Maps APIs (which, if you recall, the Uber and Lyft apps use heavily).
self-driving “pod”
“pod”
The startup Nuro figured that, while self-driving cars aren’t yet safe enough to shuttle humans around at high speeds, they’re perfectly fine for delivering groceries;
Nuro’s delivery pods have launched successfully in Phoenix.[1349] These companies are betting that they can quickly master this limited use case of self-driving cars and learn enough that they can beat the Ubers and Waymos of the world in more advanced use cases.[1350]
pod side,
None of these technologies have moved past the testing phase yet, but that’s likely to change in the coming years.
Speed bumps
A Tesla in its self-driving
“autopilot”
Second is legal.
India banned self-driving cars in 2017 to protect driving jobs,[1355]
ethical problems.
To bring transparency to these ethical dilemmas, philosophers and technologists have called for “algorithmic transparency”: the principle that self-driving cars’ algorithms be made public.[1360]
For instance, think about the ATM, which became popular in the 1970s. Most customers no longer needed to talk to a teller inside a bank’s branch office. Many people assumed that this would eliminate the teller job altogether. Right? Wrong.[1364] Thanks to the ATM, banks needed fewer human tellers in their branches. But this made branches cheaper to operate, which led to banks opening more branches. And that, in turn, led banks to employ more tellers.
More broadly, automation could move people away from doing more manual tasks to doing more tasks that require the human mind.
Most scholars believe that high-skill workers will have a lot to gain from automation, while low-skilled workers will have more to lose.[1376]
One way to mitigate this problem is education.
In 2015, Deloitte estimated that manufacturing would add 3.5 million jobs by 2025 thanks to automation — but 2 million of those would go unfilled due to a lack of skilled workers.[1377]
Elon Musk projected that automation could send unemployment up to 30-40% (again, no one agrees on the impact of automation), so he proposed a “universal basic income.” Under this scheme, governments would send every resident a check, which he said would fight poverty and protect the economy from certain collapse. Musk’s universal basic income would be funded by a government tax on robots.[1379]
Bill Gates has long had a proposal to tax robots — or rather, to tax the companies that employ robots.
An AI bot called
Amelia
has proved surprisingly good at customer support for banks, insurers, ...
This highlight has been truncated due to consecutive passage length restrictions.
One Japanese insurance company replaced 34 human agents with IBM’s Watson AI, and large numbers of people in the mortgage financing industry have seen their jobs get destroyed by automation.[1383]
In this era of fake news, many people have stopped believing news they hear online, relying on video or audio to confirm that something really happened. But if video and audio can also be faked, we wouldn’t have anything left to trust.[1388] So how can you make fake video and audio footage?
Neural networks
your brain contains a
“neural network”:[1391]
To make computers more powerful,
computer scientists have tried to simulate your brain’s neural network inside a computer. It’s called an
“artificial neural ...
This highlight has been truncated due to consecutive passage length restrictions.
an artificial neural network keeps track of many variables and assigns each variable a “weight”:
As you give the artificial neural network feedback, it tweaks the weights to get closer to the right answer,
Neural networks are incredibly powerful:
