Just A New Fractal Detail In The Big Picture (2015)

Their dumbness is what prevents them from using our infrastructure.

> That's some reasons why I think run-away AI won't be dangerous to us unless we let it build or take over our physical world which won't happen given our current caution towards subject.

Given how infrastructure is constantly being taken over by people who shouldn't be able to, I fail to see where you see any caution.

Also, the primary threat from AI IMO is not some "increased intelligence" in and of itself, but the lack of fatigue, the much greater bandwidth (you can only know so many people well, a machine isn't really limited in the number of people once it has the ability), and that machines might have a lower drive to resist being used against people than people have (which already is too low).

The EU has traditionally been more supportive of the rights of indivduals as opposed to those of companies and other organisations, in the belief that companies emerge from human behaviour and even though they operate in their own (ie commercial) environment, they exist to serve the population.

The US has by contrast given companies more freedom to operate, in the belief that by doing so they provide an environment for individuals to fulfil themselves. I fully believe a post-Brexit UK government will go the same way.

I think the key term here is "motivation". If the motivations of the company and individual align, all's well. If not, who's going to cover your back?

Rather, it's knowledge. And to get more knowledge you need to manipulate matter. Pure intelligence is useful up to a point but the you need to go out and get more data. That's the limiting factor I think of. The only thing you can do with just intelligence is math (that can have some use later on) and philosophy (that's totally useless).

> The best is physicists which had an average IQ of 130.

Incredible amount of people have IQ over 130. If that was key factor to writing stuxnet there'd be new one each day.

> An average person can barely figure out how to operate their email client, they won't be building stuxnet anytime soon.

IMHO that's mostly because they lack knowledge and any reason to care. You really don't need IQ above 130 to do technical things and having IQ of 150 or even 200 doesn't help you all that much, it seems, with pushing boundaries of human capacity. Progress is made mostly by fairly common intelligent people talking to each other.

> It is not. People make fortunes with just slightly better statistical models, or slightly better information.

And they lose fortunes with significantly better statistical models, and better information (up to but excluding insider trading). When you sum up everything it's as random as it gets. Take a look at Warren Buffet bet against hedge funds.

In all fairness additional information could help, but data is not information and where good models are unavailable and processes are chaotic you are almost just as likely to infer correct information from data as incorrect.

> But we could potentially bootstrap nanotechnology really quickly from existing biology. There are already labs that will make proteins on demand from DNA. The problem is it's just so complicated.

For me it looks more like building multi-core CPU in times of Blaise Pascal. Foundation theory is here, even some tech, but we have no idea how much technicalities lie ahead of us to figure to get to our dreams.

The reason nobody can predict the next value is that at some point something unpredictable happens that can wipe out all your previous gains and some more.

Market can remain insane (meaning not behaving according to a model) longer than you can remain solvent.

We already have vast amounts of knowledge on the internet. I can download all of wikipedia in an hour, and fit it on a flash drive. Most of the world's scientific papers and books are digitized and available.

The limiting factor is no longer knowledge. It's the ability to absorb knowledge. To be able to instantly find an obscure paper from 1930 that's relevant to your current thought, or know some random fact from some article you read years ago, etc. That's something AIs would have a huge advantage over humans at.

>Incredible amount of people have IQ over 130. If that was key factor to writing stuxnet there'd be new one each day.

Stuxnet wasn't written by one person. It was probably a huge team of intelligent people, who worked possibly for years.

But who says an AI can't be equivalent to a group of humans? If it has enough computing power, it could make copies of itself. And unlike humans it can communicate thoughts and plans instantly to it's other "selves".

And who says it has to work at the same speed humans work. Human brains run at maybe 100 hz. AI's built out of silicone could work thousands of times faster. Doing the same work, just in much shorter time.

>And they lose fortunes with significantly better statistical models, and better information (up to but excluding insider trading). When you sum up everything it's as random as it gets. Take a look at Warren Buffet bet against hedge funds.

Look it's simple math. If you can predict prices 1% more accurately than anyone else, then you can make a huge amount of money in the long run. Stock prices aren't random. They are determined by real events, mainly how much profit the company makes.

>but we have no idea how much technicalities lie ahead of us to figure to get to our dreams.

That's the point though. A superintelligent AI could figure out exactly what those technicalities are, and find the shortest route to that tech. If you went back in time to Blaise Pascal, if you took the right books and plans from the future, you could get them to CPUs in mere years. We could advance technologically much faster than we are, the limiting factor is the speed of invention, which is slow.