4 ms·

Some people think the inequality has gone down significantly (I'm in that camp), but others point out to things like Bezos and a few others owning more assets t

Some people think the inequality has gone down significantly (I'm in that camp), but others point out to things like Bezos and a few others owning more assets than 80% of the world combined. Regardless, AI is bound to increase inequality. Just yesterday we discussed this book [1] on probabilistic programming here on HN [2]. That book is, by all definitions, an advanced book about AI. How many people without a PhD can read that? How many people with a PhD can read it? Well, guess what, people who can and do read this book earn more than people who don't. That's AI driving inequality. Before AI will create the mythical HAL or other self-aware machines, it will enable humongous productivity for the top 0.01% most highly educated engineers. [1] https://arxiv.org/abs/1809.10756 https://arxiv.org/abs/1809.10756 [2] https://news.ycombinator.com/item?id=18109260 https://news.ycombinator.com/item?id=18109260

shmageggy 8y ago
The inequality generated by better tech largely doesn't go to engineers. They're still just knowledge workers, a class that has existed for a long time. It goes to shareholders and executives by virtue of reducing costs and generating more profit.
vorg 8y ago
Perhaps humankind needs a "Butlerian Jihad", as defined in the Dune novels, where all machines made in the image of a human brain are banned for all time.
- adrianN 8y ago
  Arguably we don't have machines that are made in the image of a human brain, so that wouldn't change things right now.
mindcrime 8y ago
Just yesterday we discussed this book [1] on probabilistic programming here on HN [2]. That book is, by all definitions, an advanced book about AI. How many people without a PhD can read that? How many people with a PhD can read it? From the abstract: This document is designed to be a first-year graduate-level introduction to probabilistic programming. It not only provides a thorough background for anyone wishing to use a probabilistic programming system, but also introduces the techniques needed to design and build these systems. It is aimed at people who have an undergraduate-level understanding of either or, ideally, both probabilistic machine learning and programming languages. Thanks to the Internet, how many people who have that prerequisite "undergraduate-level understanding of either or, ideally, both probabilistic machine learning and programming languages" now have access to this material, who otherwise might not have? And thanks to the Internet (Khan Academy, Coursera, OCW, Udacity, Fast.ai, 3blue1brown@youtube, etc.) how many people now have the ability to gain that prerequisite knowledge without investing tens of thousands of dollars in a university degree? Especially people who might not have been able to afford to take time off from working to attend University in the first place? While in the end you may be right that AI specifically does wind up driving inequality, I think there are strong arguments that technology / the 'net in a general sense also serve to advance equality.

shmageggy 8y ago

The inequality generated by better tech largely doesn't go to engineers. They're still just knowledge workers, a class that has existed for a long time. It goes to shareholders and executives by virtue of reducing costs and generating more profit.

vorg 8y ago

Perhaps humankind needs a "Butlerian Jihad", as defined in the Dune novels, where all machines made in the image of a human brain are banned for all time.

adrianN 8y ago
Arguably we don't have machines that are made in the image of a human brain, so that wouldn't change things right now.

mindcrime 8y ago

Just yesterday we discussed this book [1] on probabilistic programming here on HN [2]. That book is, by all definitions, an advanced book about AI. How many people without a PhD can read that? How many people with a PhD can read it? From the abstract: This document is designed to be a first-year graduate-level introduction to probabilistic programming. It not only provides a thorough background for anyone wishing to use a probabilistic programming system, but also introduces the techniques needed to design and build these systems. It is aimed at people who have an undergraduate-level understanding of either or, ideally, both probabilistic machine learning and programming languages. Thanks to the Internet, how many people who have that prerequisite "undergraduate-level understanding of either or, ideally, both probabilistic machine learning and programming languages" now have access to this material, who otherwise might not have? And thanks to the Internet (Khan Academy, Coursera, OCW, Udacity, Fast.ai, 3blue1brown@youtube, etc.) how many people now have the ability to gain that prerequisite knowledge without investing tens of thousands of dollars in a university degree? Especially people who might not have been able to afford to take time off from working to attend University in the first place? While in the end you may be right that AI specifically does wind up driving inequality, I think there are strong arguments that technology / the 'net in a general sense also serve to advance equality.