Musk strongly agrees with me.

I'd take the world of an engineer over the word of a some dopey chess player trying to extrapolate his chess computer experiences to the field of AI. Chess is such a stupid game that computers could beat humans a decade ago. It was one of the first things to fall to computation.

Proficiency in chess is mostly spending years memorizing openings and strategies, with a bit of cleverness thrown in.

Most important reason Musks probably tells people AI is a big danger is to scare a lot of scientists into keeping everything out in the open. He said his big nightmare is some guy in a basement developing a super AI. But that is probably because that guy will then have enourmous financial leverage over companies like Tesla.

Seems that in AI you have the guys with all the data, and the ones with the algo's. The ones with the data (like Tesla) would want those algo's as cheap as possible. What better way to do that than to fearmonger scientists into keeping things in the open?

I started to seriously distrust Musk after discovering how ridicilous and completely impractical his hyperloop idea was. He is too smart to not know that. Seems a lot of these statements from him are clever ways to keep the hype up, so he can issue Tesla shares at a high valuation.

That said, Musk could be right. I don't think anybody has a clue on what is truly possible in the AI field, since it is so new. I listened to some MIT lecture on neural networks the other day that was pretty recent, and the guy lecturing said that the correct answer from cutting edge experts to like 80% of relatively basic questions about what is going on in the brain is 'I don't know'.

I don't understand what the big deal is about replacing people with robots. We've been replacing people with machines for a very long time.

Just look at farming....the amount of "farmer workers" that have been replaced by machines is TREMENDOUS!

So what if more people get replaced by machines and/or robots. It'll be long term beneficial for society as those people can move to more productive forms of work and/or humans can enjoy an increased amount of leisure. People complaining that the robots will take all the jobs just don't know history. Go read "The Rational Optimist" if you really want to look more into this.

I understand the argument. This time, there aren't any new avenues for human labor to be allocated toward.

Zuckerberg is proposing a universal income. That is really just a nice way of saying "let's have communism". We all make the same wage, no competition, no incentive.

If you look at "history".... it never ends well.

UBI is not the same as communism

You're kidding right?

You don't think in 20 years there will be more internet based/programming/app development etc. type jobs then there are today by a large amount?

You don't think all the baby boomers retiring aren't going to require more nurses/doctors/medical care?

You don't think as our population continues to grow and people get married later/divorced more often that household formation isn't going to demand more housing?

The other thing to consider is that when people have more free time, and/or desperate for an income source, and/or laid off from work and forced to find something else.....that's EXACTLY when new inventions come out, new industries are formed etc. etc.

You act like if someone gets fired they're just going to sit at home and twiddle their thumbs. Sure some idiots will....but a lot of people will go and create their own job.

(side note: in regards to your UBI...I'm 100% against it, as it'll destroy ambition and is nothing more then the government stealing from one person to give to another...but honestly it'll probably happen and maybe even in our lifetime, it's just too easy of a way for politicians to buy votes, I can't see the political class passing it up)

Musk the CEO who is constantly off the cuff? I thought you thought he was a cultist w "Kim Jong Il" like mind control powers! Now you're yearning to sit next to him in a photo op. Metaphorically speaking of course.

When it comes to AI, Kasporov is the expert, not Musk.

Ya, in econ it's called "reskilling".

I think UBI countries will probably see a talent drain. Because some basic calcs show that it is prohibitively expensive. So taxes need to be raised for sure. And once 75% of your voter base is hooked up on the UBI IV, they will vote for tax increases to keep the party going.

So you will have countries with UBI, and countries without UBI with low tax rates. Deadbeats will slowly go to UBI countries, and the winners will slowly go to non UBI countries, until the UBI countries collapse economically.

Then the UBI countries will fight a war with non UBI countries (because they are jealous of all that wealth), but they will lose. Because they couldn't get out of bed in time.

Then after the war, instead of giving the deadbeats UBI, they will just be hooked up on the entertainmentfeedmatron4000®. So they can consume and get nutrients injected 24/7 without ever getting out of the house. This is of course a clever ploy because they will slowly be harvested for energy by non UBI people.

Agree w/ post above.

It will be a very slow "vote with your feet" movement, though. Most nations don't take immigrants from people trying to pwn the system. And they will be happier with the EntertainmentFeedMatron4000® than trying to contribute more to society anyway.

Looks GTO to me.

Should probably break this off into a new thread about UBI if there isn't one already in existence, but UBI doesn't necessarily have to be a "talent drain" or effectively be communism.

USA's welfare state could actually be reformed in a massively +EV way with UBI. If you instill UBI, you get to cut a whole bunch of other government spending that no longer becomes necessary and is probably less efficient than direct cash. Could also have subsequent beneficial effects on healthcare and taxation (both of which, in their current states, are completely ****ed up and have much room for improvement).

I don't necessarily disagree with the previous few posts about UBI, but I certainly believe those views are too pessimistic and dismissive.

Better yet a new forum! Take this crap to politics where it belongs.

Social Media Consultant.

SEO specialist.

App developers.

Just because FB or Google does not employ them does not mean there is not other jobs that are being generated due to their innovation, creation, etc.

Well I'm sure you've all seen the news now that's it's front page of WSJ and a whole bunch of other sites.

http://www.autonews.com/article/2017...l-intelligence


"Tesla Boss Warns of Artificial Intelligence" by Tim Higgins of WSJ

Direct quotes from Musk:

"It is the biggest risk that we face as a civilization."

"the technology will threaten all human jobs and may even start a war"

"Right now the government doesn't even have insight....Once there is awareness people will be afraid, as they should be"

Wow, straight from a tech boy's mouth. Guess you need to be a revolutionary genius to state something so blindingly obvious before anyone will listen. Don't be worried society, you can become a LMFAO "social media consultant" and instead of being paid in real currency, you get paid with fake "likes and retweets" by social media bots. Sounds f*cking awesome.

Is it 100% sure computer power will increase enough to reach the level needed for general AI or is there a chance it will stagnate before that?

Related question: Is there a economic reason or somethign that incentives companies to reduces the speed of innovation. I see people make predictions based on Moores Law but is there any fundamental reason why we should expect it too continue. The fact that is has continued for soo long is annoying me cuz I cant see any reason.

It makes sense if there would be a monopoly tech company, they would prefer to keep making slight improvements of the product instead of go straight to the best version, but in competing markets everyone is incentivized to go straight to the best version right?

We should make sure and highlight that it is "General" AI (not AI in general, but a specific form/field of AI that replicates more or less full human thinking or actions up to a human level - just FYI). that being said, I do think that, barring some global crisis etc, General AI is not only more or less certain, but near to being here with us.

As for innovation, I can't think of any economic reason (unless some body or entity is subsidizing the elongation of technical innovation for an industry); there are certainly barriers (not the list of which is economy of resources available to invest in that innovation). there could be disincentives in place that somewhat inhibit the pathway of innovation (for example, EPA regulations for certain industries etc.) Ultimately though, with enough access to the technology and resources to deploy or develop it there is no reason a company should deter innovation other than ethical/moral ones.

Imo yes. If humans have general intelligence then clearly general intelligence is possible. As long as we are moving forward with technology at some point we will get there.

There are at least three possible ways to reach superhuman general intelligence:
1. Breed smarter humans/animals. We can already select like 3 IQ points per generation and soon we will understand genome so well that we can predict intelligence at embryo level, making generation times much shorter. Genetical Modification can likely speed up this also.
2. Make smart AI. With memory added to the deep neural networks they are Turing complete, ie they generalise. Harder and harder tasks are reaching super human performance, today it is playing GO, in a few decades it will be researching AI.
3. Simulate humans. Worst case we just simulate a human in the computer. Then we speed it up and create copies of it. Then we let those "humans" develop better simulations.

If I had to bet I would say p>0.9 for AGI in 30years.

Elon elaborates a bit here 55min into this:


Regarding Kasparov, imo he was great at chess, had some experience playing against one AI but is not really that well schooled in mathematics, game theory, computer science and machine learning. For example he claimed that computers can't play poker because they can't bluff. Imo this highlights a lack of understanding in this topic. I would say that Demis Hassabis is the expert today.

Elon Musk has no background in AI whatsoever.

He did start OpenAI. And he has people like Ilya Sutskever and Andrej Karpathy working for him.

http://www.vanityfair.com/news/2017/...epathy-company

Musk is also CEO of Neuralink where he wants to link the human mind with computers.

Saying Musk has no background in AI is like saying Steve Jobs has no background with cell phones.

AI is Musk's white whale, imo. Also he's a CEO of several companies. By your Jobbs analogy he's a rocket scientist and an AI Phd, only he studied physics and business at Berkely, not AI.

Musk is a CEO who surrounds himself with smart people and thrives to be cutting edge. But his AI doom is stupid and not rooted in his AI Phd.

What makes it that you don't like the answers? If the automation does a better job at a better price and the investment is of low risk then it would be irresponsible not to do it. Does it suck for the employees, yes. Are you working to make money or to be a charity?

Zuckerberg labels Musk as uneducated and his comments as irresponsible.