Hacker News Reader

Comment by mchahn | Hacker News Reader

It was fascinating when the author experienced a weird bit of nausea when thinking about the implications of AI. I'm not sure how I feel about it yet. It does scare me a bit.

View on HN

I saw this tweet earlier:

In 13 years, computers progressed from beating Chess to mastering Go. I hope that in 13 more they will learn to not steal focus while I type

Think about that we're constantly pushing toward our own replacement.

considering the possibility of augmentation, supplantation seems like somewhat of an assumption.

When a student surpasses the teacher, that's supposed to be a good thing.

That's not really relatable because a student can only replace one teacher. When it comes to software, it's more so like saying that one student can replace all the teachers. We're seeing this now with self driving vehicles and the trucking industry. Once the software gets good enough and it's approved, there'll be a massive push for it's adoption and I'm sure that within roughly 3 years, about 70% of truckers will be replaced from that single breakthrough. Then comes taxi drives, maybe pilots and train operators in another 8 years or so. It'll be a cascade of lay-offs and that's just with this technology. So while it's a good thing that a student surpasses the teacher, this isn't exactly that same scenario.

Sure, but this phenomenon is not new. It's called creative destruction: innovators discover a market advantage, creating new industries while destroying the old.

Don't we normally call that having children?

How much of humanity genes will skynet inherit ?

Not genes, memes.

memetics

Important distinction: children get their genes from us and share many of our values by default. Computers do not share many of our values by default. Instead they do what they are programmed to do.

But the problem is that computers do what you say, not what you mean. If I write a function called be_nice_to_people(), the fact that I gave my function that name does nothing to affect the implementation. Instead my computers behavior will depend on the specific details of the implementation. And since being nice to people is a behavior that's extremely hard to precisely specify, by default creating an AI that's smart enough to replace humans is likely to result in a bad outcome.

Reply on news.ycombinator.com