Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
so what's your great solution? To make it where we can't advance technology anymore? Or to force people to just keep hiring people regardless of if there is work or not?
Utopia or dystopia.
Only a world wide disaster would stop the tech. So we can either have an egalitarian society where the bounty is shared, and everyone lives well without a need to work. Or we can have one where the massive wealth increase is controlled by a few, and most people are sustained in high-tech low-impact ghettos until they die off.
The aim is to build a computer with ten billion neurons and a hundred-trillion synapses that consumes just one kilowatt of power and occupies less than two litres in volume.
Now keep in mind, every 18 months that number will double. One major breakthrough no one has mentioned here yet is the crosspoint technology that Intel has developed. Basically think really really fast and large flash memory. Its a major breakthrough for AI research. It plus these chips from IBM are moving this forward even faster then I was expecting.
Additionally for the person going on about how it takes 25 years to train a CPA....bwahahaha.
Its not going to take us 25 years to train a computer system to be a CPA. And once we have 1 doing it, it will take moments to allow ALL of them to do it.
Imagine your CPA, we make 1,000 AI's to replace 1,000 CPA's. The CPA's train them. every day they gain 3 years of experience. Call it 1 year due to duplication of information.
In a month The AI's have replaced all CPA's in ability.
Additionally....the AI's never sleep, don't eat, and don't make mistakes because they are tired.
Timeframe for this? Well....before the crosspoint memory from Intel, and IBM's stuff I was thinking 20 years or so. now? Less. Not sure how much less yet, but the trends to create a system with enough capability to do a serious AI is on track. But what if its just the same ol linear results we have been getting....IE moores law.
Human brain: 1 billion neurons.
IBM research: 16 million neurons.
Using moores law as a guide....(18 month doubling, we will give numbers in 36 month jumps (so 4X)
2018 32 million neurons
2021 128 million neurons
2024 512 million neurons
2027 2 billion neurons
2030 8 billion neurons (Very possible that at this point it can be done-remove a lot of what we use neurons for and you will have a expert system that will work well.
2033 32 billion neurons
2036 128 billion neurons (from here on out you can literally emulate a human brain)
So 2036 assuming no unusual breakthroughs, and 1-1 emulation requirements. The reality? We're seeing some amazing technological breakthroughs. And the 1-1 emulation requirement is vastly overstated. over 100 million neurons are used for vision alone, and we already have self driving cars on the road. My geuss? 2030. Based upon the above data. (Not tossing sticks on the floor, but rather on past data, and the knowledge of the topic)
And you are wrong, over and over. It won't even be hard.
You keep saying that but you never explain even at a high level how this "won't even be hard" thing can be accomplished.
I, on the other hand, have dived into detailed explanation (but of course not the entirety) of why it cannot be done. If you can match 10% of the thought I put into this discussion, perhaps it can move forward, because right now it's getting rather pointless.
Quote:
Originally Posted by rruff
Computer processing power is currently at mouse level, and they can still beat every human on the planet in many tasks. When the hardware gets up to human level, there won't be much if anything left where humans have an edge.
Your explanation makes me think you don't actually understand what processing power means. I've already explained why it doesn't matter. So perhaps it's time for you to tell me why you are so hung on this processing power thingy.
Quote:
Originally Posted by rruff
Besides, how many people make a living now from their innovation and creativity?
Many. Most of the technology, healthcare, engineering, arts, marketing, sales, etc; jobs involve that. What kind of jobs have you done?
.
In a month The AI's have replaced all CPA's in ability.
One thing I'd like to mention is that most work supports a consumer capitalist economic system. CPAs have jobs because of profit seeking businesses and taxation. These will eventually disappear and with them, the infrastructure that supports them.
Think of the incredible amount of resource consumption due to one 1st world human. From the cars, the house, the cars, the roads, the utilities, the food, the planes, etc. Now compare that to a person who basically lives in a VR pod. How many resources do they consume? How many service workers are necessary to support their lifestyle?
If you belong to the oligarchs club, just how many resources are you willing to waste on people who no longer have a useful function?
Are you making a joke? Because I don't see why an employer would want his workers to be whimsical.
You don't get it. This "whim" is part of decision making. It's what hardware is incapable of doing without human programming. An employer may not want a whimsical employee, but a thing that is utterly incapable of making decisions are also not fit for higher level positions.
What jobs have you done? From your response, it sounds to me like you've only held manual intensive and repetitive jobs that don't require a high level of thinking. I could be wrong, but it comes off sounding like that.
Quote:
Originally Posted by rruff
But if you believe that a robot *can't* display the same kind of behavior as humans, you are underestimating computer capability and overestimating human.
You like to talk in very broad terms that doesn't really tell us anything.
.
I've already explained - computing power is not AI. It's like horsepower in a car, no matter how much juice you give it, it is never going to turn into a Transformer without some other new innovation.
All that article says is that we can make faster processor. Nothing more. It really has no meaning whatsoever if you're talking about AI.
.
I've already explained - computing power is not AI. It's like horsepower in a car, no matter how much juice you give it, it is never going to turn into a Transformer without some other new innovation.
All that article says is that we can make faster processor. Nothing more. It really has no meaning whatsoever if you're talking about AI.
.
Go watch the jeopardy video I posted earlier.
We dont have to have emotional whimsical AI's to replace humans. quite the opposite.
Also the article talks about simulating neurons and synapses specifically. This is the basic building blocks of a brain. I'm not sure why you are in such denial.
Moderator cut: .
Meh, I like hearing other opinions, and he has put some thought into this, the point about more horsepower not making your car into a transformer is relevant. Its wrong, but its nice to understand WHY people think that way.
Last edited by yellowbelle; 08-21-2015 at 08:32 PM..
Reason: orphaned - the quoted text was deleted
The Matrix, Terminator or Her? Which future are we going to get? lol.....
Processing power is not thinking.
As long as our computing architecture is based on the foundation of 0:1, all our computers are doing are a lot of "if, then" thinking. "If, then" calculations can work for chess, because there are only so many possible moves that can be made because there are rules of the game. It can win at jeopardy because there IS a right answer out of possible millions and a computer can process that.
Obviously humans think differently than that. Better than that. We cheat. Remember Kirk beat the Kobiyashi Meru simulation because he re-programmed it. When a computer is able to do that, I'll be more impressed.
I actually look forward to the day when computers replace mundane jobs like CPA. We'll miss them about as much as we miss the assembly lines where you worked 12 hours a day and occasionally a worker got his finger cut off and no one cared. Why people pine for that I don't know.
Dealing with money is probably something our computers could be much better at. If we took the emotion out of investing we'd have fewer crashes.
Status:
"Do not pass GO, do not collect $300 (used to be $200)"
(set 6 days ago)
Location: TN
600 posts, read 273,961 times
Reputation: 385
Well, here's my uninvited attempt to clear things up:
1. Here's the math on automation: yes, you can replace 100 machinists with 50 devs, 50 engineers, 10 lawyers and 5 buyers for the robot parts (115, sounds like more jobs). But what's really going on is that there were 100 facilities: instead of having 100x100=10,000 machinists, the 112 people can support all 100 facilities with maybe 2 operators at each facility (200 more jobs). Bottom line: we went from 10,000 jobs to 312.
2. Or think of it this way: no one's going to change their production system if it'll cost more money. You won't switch from 10 line workers to 10 engineers unless you know you'll eat the market share of a lot of companies (and effectively eat those jobs in the process).
3. Who says human thinking isn't based on 0s and 1s? I think current science is that the brain is based on a complex network of neurons - yet sure enough, each neuron either fires or it doesn't.
4. Automation may indeed wipe out a lot of white collar jobs. Look into legal document review: lots of folks with decent brains going through thousands of documents and flagging things as relevant, not relevant etc. Computers are tracking this, and I'm sure the natural result will be nothing short of legal advice: "based on input x, I think y is relevant. In y cases, successful people did this, and unsuccessful people did that."
5. Also, computers don't have to be perfect - if the computers get 5% wrong and nets you $95,000, while the best person gets 10% wrong and nets you $90,000, then the market may indeed favor computers.
6. Finally, a food-for-thought idea on Moore's law: maybe it's just a reflection of human limitation. Maybe progress could be much faster, but there's only so much risk people are willing to take when expanding a facility, ordering new machines, or just generally trying new things. So the doubling represents an equilibrium solution, nothing more.
Please register to post and access all features of our very popular forum. It is free and quick. Over $68,000 in prizes has already been given out to active posters on our forum. Additional giveaways are planned.
Detailed information about all U.S. cities, counties, and zip codes on our site: City-data.com.
Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
Are you making a joke? Because I don't see why an employer would want his workers to be whimsical.
