For hundreds of years, the answer to progress and technology destroying jobs was "Go get an education." Now, automation and AI is targeting educated jobs,
That's rather interesting. And yeah, that's going to shift a bit.
But... since about 2000 it's not been enough to get a degree. Before, ANY sort of degree put you a step above the masses and meant you were smart (and/or wealthy) and could get trained to do a slew of jobs. Now it's more like... "Go get a USEFUL education". Cue standard refrain about underwater basket weaving 101. This bit my brother in the ass. There's really not much call for a BA in anthropology. With AI, what's defined as "useful" is going to change.
I think the issue is that not everyone is smart enough to go get a useful degree. Kids "dropping down" to an easier major was a common theme during my Uni days. And some people can't even get into the schools. People are getting smarter or more educated. Something is increasing the average IQ scores. But I'm not sure it's fast enough. There's always high demand for knowledge workers with specific skills. (And right now unemployment is good in general). After AI takes (more) jobs, it's not going to be good for the GINI coefficient (inequality). The answer is still the same, but it's getting harder to make happen.
Eat the Rich will come into play, so this will all ultimately reset itself
Naw, bread and circuses. Movies, Games, Internet. Hell Cannibis is probably going to get legalized everywhere. We might as well call it "Soma" and embrace the brave new world. As long as people have enough to scrape by and some entertainment to distract them, they're not going to eat the rich. There are enough bleeding hearts in the rich upper elite that we still have welfare.
Software that learns. The bar is pretty low because it's not that special. Making software that learns complex things quickly is the difference between a rock you can use as a hammer and a pneumatic jackhammer. Just like how "life" includes everything from bacteria to humans.
it is often trivial to feed the system garbage until it is useless.
Just like people.
A basic feedback loop would at least help point it correctly.
BRILLIANT! You should tell someone over at google that maybe they could feed the output of their neural net back into the inputs. Could be a real game-changer.
(I expect this sort of ignorance from the general populace, but on slashdot this is ridiculous. Come on people)
an elite-focused system without even a pretense of noblesse oblige and a terribly smug disdain for the "wrong people" almost all of whom are poor
A lot of people who believe that AI is going to take over a bunch of jobs are talking about Universal Basic Income. That is, we have welfare to support everyone, including the less fortunate. That's pretty much exactly noblesse oblige via the taxes to pay for said system. And this whole article is a tech-company CEO talking and warning about it. And what distain? Who openly shits on poor people?
What? No, it's exactly like automated looms. What once took skilled labor can now use menial labor and the output is greatly increased. Automated looms made those expensive guild weavers obsolete and destroyed their jobs. A new job of tending to those machines opened up and was largely filled by street urchins. The industrialist owners got ludicrously rich.
In exactly the same way, specialized knowledge-workers are going to be replaced by a few IT staff (or IT staff at a third-party company) maintaining the machines which do their tasks. Like general practitioners giving initial diagnosis, HR, and paralegals.
We HAVE been here before. And that's exactly the concern. THREE generations of soul-crushing 50% unemployment (for those in the industry), riots, what amounts to terrorism, open rebellion, and sending in the army to put it down. That's pretty much the worst-case scenario. We've got to do better.
Hmmm, that's right. It's not simple a matter of AI replacing workers. The question is how many paralegals does the world need, at what price-point, and how many of them will be AI. Drop the price-point and people find novel uses.
How many doctors do we need? Well, if they were cheap enough I could use a doctor any time I wasn't feeling good. Which would be a LOT more frequent than my current views on the medical industry. Which is pretty much just a long string of expletives about insurance and costs and reproducability and guesswork.
How many office mail clerks do we need? Since they're extinct with the invention of email, NONE. But raises the point that, with new tech does come new jobs. And every place has an IT department, almost like the new office mail room. Just with a lot less employees. Maybe we'll have office AI wranglers.
How many HR staff do we need? As long as I get my check, and there are no issues with vacation, I don't think there are any. I'm not a real big fan of their involvement with hiring either.
How many scientists, programmers, engineers do we need? (Not that I think AI is anywhere near this level, but hey, the tools are getting better. All scientists need to learn to program these days) I think the answer is that there's an unlimited amount of stuff to do for these sort of fields. As long as there's a limit of human knowledge, or things need to get made, the job pool for these guys will be around despite how advanced the field's capabilities become.
And I think there's a similar deal with artists. Maybe even politicians if you REALLY squint and view them like "people who deal with interactions between groups of people". Certain types of needs are elastic and we'll take as much as we can get.
The AI will let you do anything you want. They're tools that open a lot of possibilities. They can automate tasks, but they still need someone to set them up and deal with the output.
Your boss will can your ass the first moment he can. get away with it....Although, that said, there are a lot of people that could be replaced with a dozen lines of bash script.
Also, wtf would a block-chain.... network... be all about? How are they using "network" here?
to upend web-based businesses such as Facebook and Twitter
First off, Twitter isn't web-based, it's Internet-based through and through. And there's a compelling argument that facebook isn't really web-based anymore either. Most of their traffic is through phones and their application, which bypasses the web.
Both of these are only networks in the sense that they have social networks (and whatever CDN they run on).
And decentralized facebook has been tried. despora? I think? It didn't go anywhere because facebook has a critical mass of users. Users are the product, why would you switch to a system with an inferior product?
Decentralized twitter... is... fuck man, I still don't even understand this fad. email is too complicated for some people? Newsletters just aren't hip enough? If you made a "I'll sign up for the first 200 characters of your email newsletter, but that's it" button, you'd effectively have the same damn thing. You don't even have to trust the senders, you could just filter shit on your end.
The writers should take some caution when using a historical or cultural event as a backdrop.
Maybe if it was presented as the focus of the story, or it was a serious piece of work. As a comic-book fantasy-piece, it's not really even trying to convey "This is the truth". Let me put it this way, if someone is 15 and such an idiot that they can't see past the layer of hollywood, they're going to be fucked when it comes to actual political propaganda. Which is a hell of a lot more subtle and can be a lot more pervasive. Hell, at 15 you could even consider this a sort of inoculation to bullshit. At some point they're going to hear more about the 6-day war, and they'll learn that, along with the magical flying woman, not everything they had heard was true. Which is a damn important lesson. I guess I'm saying that we need to lie to teenagers, at least a little, to teach them to question the source and be at least a little cynical. Otherwise we're setting them up for failure. Of course some dumb fucks will blindly believe what they see in the movies. You can't help that.
What you're proposing is that no-one can use any historical or cultural event in a movie, or a book, game, really any sort of media, without "being sensitive" to... something or other. Political finger-pointing? Making people look like "the bad guy"? Maybe for PG movies, but pg-13 and up is fine and should be expected.....we're going to have movies rated PG-13 for cultural insensitivity aren't we? Or rated R for blatant propaganda. (haha, rated NC-17 for the truth)
ok, sure, "forecast curve". But there are multiple variables allowing forecasts for things like "where should I move rook?" and "should this person go get screened for cancer?" and "Is this a picture of a cow?" and the datasets seem large enough to beat the best players in Chess and Go, predict cancer, and identify objects. Sooooooo.... no excuses, they're being used today for real effect.
And, what about DeepMind Alpha? The thing they made right after DeepMind won big at Go. It doesn't need any dataset, it makes it's own, and beat the previous DeepMind after a rather short amount of learning.
because their domain of knowledge is pinpoint size.
Sure. Nearly all of them have limited scope. But many of them have a scope large enough to give them a hell of a utility. Especially the ones that do the bulk of certain office-worker jobs. They're coming for the paralegals.
I don't want to oversell it though. AI is a useful tool. We don't have any sort of "general intelligence" yet. But to deny that these thing learns and teach themselves.... and that they're not intelligent.... just seems deluded. Like wtf is intelligence if not this?
Google is making self-driving cars... which is hot right now.
And Facebook was at the heart of social media... which was hot back in.... 2006?
Yahoo was hot back when searching the Internet was a big deal. Back in the 90's.
Cloud computing peaked around 2011. But, YES, if you set up a company revolving around self-driving cars, cloud computing, social media, or even search, you are a tech company. The field itself isn't new. But unless you bring something new to the table, you're not going to do very well. This might shock you, but people are still research and advancing the technology of internal combustion engines and making them more efficient, despite being around for... what? A century? That's still tech. And if you made a company dedicated to improving the technology of that old-ass invention, you'd have yourself a tech company.
I think you might be expecting a revolutionary game changer with every new business. We're living in the middle of a technological singularity. Just like the industrial revolution, it's coming in waves. We've got computers, personal computers, the Internet, hand-held computers, and I think artificial intelligence will be another one, if it's not already. But there are still people working on building better computers. It's still a pretty new field, all things considered in the broader scheme of "the economy".
And in the 1600's governments had a very cozy relationship with their East Indie Trading companies and the like. And these days we've seen, through the diplomatic cable leaks, that government is somewhat cozy with industry today. The point is, government-industry cooperation doesn't always work out for the best.
Let's refine that. Government funded basic research opens doors to industry and advances technology.
Also, are you REALLY sure you want to try and spin the whole "social credit score" as a positive thing? Most people freak out at that and blow the red-scare whistle. I mean, I understand China is making a credit-score system just like America has with it's **wink-wink** private industry **nudge-nudge** Experion-TransUnion-whoeverthethirdguyis, just with a government-Loyalty aspect. Not that we know what Experion's algorithm is. And of course private companies are using it for profit. It's made to figure out who to give loans to. It's just... kind of a weird example to throw out there as most people would say that's a bad thing.
Got into a debate with my brother about crypto currencies. He was appalled about how much power they use globally. He's kind of an eco-nut.
There are has statistics about SETI@Home like "Since its launch on May 17, 1999, the project has logged over two million years of aggregate computing time..... With over 145,000 active computers in the system (1.4 million total) in 233 countries, as of 23 June 2013, SETI@home had the ability to compute over 668 teraFLOPS."
Buuuut, how much power consumption is that? And I've really no idea about how much power consumption lies behind all the servers processing VISA transactions, or Amazon's servers.
Syncing a phone with a desktop. Like you walk within wifi range and it logs in, syncs music, text, downloads, calendar (moot for me, as I use google calandar, but it opens up third-party tools).
I'd also like to just have the phone show up as a window on my desktop. So you walk up, open an application, and you can then use your mouse and keyboard on a window that has the phone's screen and you can enter the code and run stuff, etc.
And while I'll always have a desktop for crunching power reasons, most people could get by with the processing power and storage space of a phone. Even at work. But it's torturous and cruel to make them use such a small screen and swiping. So blue-tooth keyboard and mouse.... and wireless or dockable video port and you can have most employees just walk up to any terminal and it's their computer they're typing on and seeing.
I'd also like easy man-in-the-middle HDMI recorderr and IoT tools. Like something that flips on a relay and logs power usage to inbetween a wall socket. Boom, anything you can plug in to turn on is now an IoT device. Anything that turns itself on and or draws different rates of power is loggable with... phone messages or whatever.
And organizations of people don't have all the rights that individual people have. Nor the risks. Which is why we form corporations.
We can execute a person for a crime. We don't execute groups of people. At least we're not supposed to.
We can revoke the charter of a big corporation if we think it is no longer serving the interests of society,
Name me a corporation with actual assets that had it's charter revoked.
Corporate personshood is a thing that's gone this way and that over the course of time. I think it would be better for the USA and the world if corporations had less rights. Taking away corporations' rights does not remove our individual rights.
How do you propose to take away the voices of corporations? Shall we pass a law preventing individuals from speaking on behalf of corporations?
No, we legislate that corporations don't have free speech, they only have commercial speech. Like in advertisements. Because while they employ people, corporations are not people.
Advertisements have all sorts of restrictions on what they can say. They can't blatantly LIE. Because that would be fraud. For some things they face COMPELLED speech. All that sped-up medical information at the end of medication ads.
In that way, if a person lobbied their politician, they can say whatever they want. But if a corporate lobbyist lied to a politician, they could be sued.
AI as it currently exists is no more exciting than the assembly line.
... in the 1800's. Yes. It's pretty much exactly like that. It looks to be another phase of the technological singularity which is the computer revolution. Just like the industrial revolution came in a couple waves and CHANGED EVERYTHING computers, the internet, hand-held devices, and now AI have and are going to change a lot of things when it comes to employment, business, and how things are done.
Robotics is great for automation of tasks. The type of AI we have now is great for expert systems and chewing through large amounts of data. The combination of machine learning and robotics have exciting prospects for eliminating mundane jobs.
Yep. Yep (and also things other than expert systems, but sure, close enough). and... No? The combination of robotics and AI is pretty cool for things like Baxter, but less on the AI and more because it's easy to set up. Separately though, Robotics HAS ALREADY been eaten manufacturing jobs since 2000. Output is up, employment is down. That's the worry that the WORKERS have with AI. The bosses? Hey, they think AI is great. But the sort of change that AI looks like it's going to bring about doesn't require robotics. It's stuff it can do in a server farm across the globe. Liiiiiiike, doctors. Doctors get paid a lot. Especially to translate your symptoms into a diagnosis. If a computer could do that better (and people trusted it to do better), a lot of general practitioners would be out of a job.
However we are no closer to hard AI today than we were forty years ago.
Bullshit
we naively assumed that human ingenuity was rendering artificial sentience into something that was right around the corner.
Whoa buddy, way to make the leap from AI to "sentience". You have to remember that society is diverse. The people actually working with the stuff knew it's limits. The people in hollywood are generally idiots. The general masses sadly are more influenced by hollywood than scientists. Which is why it's a tedious chore to call bullshit on all the idiots every time this comes up.
I am no more impressed with computers winning at Go and Chess than I am impressed that a hydraulic press can exert several (thousand) times my strength.
Eh, sure. ok. Both are useful tools. Imagine seeing the first hydrolic press. That'd be a pretty fucking awesome development.
The software that is winning at Chess and Go are, in fact, little smarter than that same hydraulic press.
Wrong. The software is smart like the press is strong.
The software knows from analyzing millions of games that humans have played what winning strategies are, and combines that with brute force strength to know where to optimize its searches.
How the hell did you use "brute force" and "optimize it's searches" in the same breath without an anurism? ANYWAY, glossing over that, YES you're sorta right. DeepBlue and DeepMind analyzed a TON of games and found winning strategies. (JUUUUUUST LIIIIIIIKE HUMANS!) But then there came DeepMindAlpha which DOES NOT require any learning sets and promptly beat the version of DeepMind which did learn from training data. DeepmindAlpha self-taught. No prior experience. "From scratch". If this is your main holdup, then you need to cure your ignorance on the current state of affairs.
We are not close to hard AI.
Probably true. But I'd say we're "closer".
We are not close to soft AI.
Utter bullshit. We had that with the Atari 2600. A single "if" statement. Goombas walking back and
Come on people, the world of tech is wide and deep. You don't have to define what Ubuntu is, but you do have to define what the hell this new thing is.
Putting "Real" in quotes followed by "would be" implies that there is no real artificial intelligence, and the tools we have available are somehow "fake".
While I understand that people argue over the definition of what artificial intelligence really is, people would generally agree the definition of artificial intelligence certainly includes real, existing, here-and-now tools. They are real. AI is a real thing.
And no, a counter-example to the bogus definition (e.g. the traditional skit: "a True scotsman eats haggis", "NO a TRUE scotsman wears a kilt!") is simply perpetuating the fallacy, and is not required for the fallacy to exist in the first place. The fallacy is a problem because EVERYONE will have some arbitrary definition or goalpost
Your arbitrarily strict definition of intelligence hinges on some sort of nuance about "Totally new scenarios" in italics and everything. And yet you can't describe any such scenario you've been through. So what? You don't qualify as intelligent?
But sure, you want a definition? Let's go with Merriam Webster:
Definition of artificial intelligence
1 : a branch of computer science dealing with the simulation of intelligent behavior in computers
2 : the capability of a machine to imitate intelligent human behavior
Google's:
the theory and development of computer systems able to perform tasks that normally require human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages.
Or dictionary.com:
the capacity of a computer to perform operations analogous to learning and decision making in humans, as by an expert system, a program for CAD or CAM, or a program for the perception and recognition of shapes in computer vision systems.
Personally, I'd just boil it down to "It can learn". Maybe tack on "and updates it's behavior accordingly". That's the interesting part after all. Not that my definition really matters. By any sane definition, we have real AI tools. They learn, they make decisions, and (imperfectly) simulate intelligent behavior. If you want to fear-monger about the future of AI, hey go for it. It's certainly something we should talk about. We should likewise talk about the hype-train. But denying the current state of research and the decades of progress isn't helpful.
Slashdot Mirror
For hundreds of years, the answer to progress and technology destroying jobs was "Go get an education." Now, automation and AI is targeting educated jobs,
That's rather interesting. And yeah, that's going to shift a bit.
But... since about 2000 it's not been enough to get a degree. Before, ANY sort of degree put you a step above the masses and meant you were smart (and/or wealthy) and could get trained to do a slew of jobs. Now it's more like... "Go get a USEFUL education". Cue standard refrain about underwater basket weaving 101. This bit my brother in the ass. There's really not much call for a BA in anthropology. With AI, what's defined as "useful" is going to change.
I think the issue is that not everyone is smart enough to go get a useful degree. Kids "dropping down" to an easier major was a common theme during my Uni days. And some people can't even get into the schools. People are getting smarter or more educated. Something is increasing the average IQ scores. But I'm not sure it's fast enough. There's always high demand for knowledge workers with specific skills. (And right now unemployment is good in general). After AI takes (more) jobs, it's not going to be good for the GINI coefficient (inequality). The answer is still the same, but it's getting harder to make happen.
Eat the Rich will come into play, so this will all ultimately reset itself
Naw, bread and circuses. Movies, Games, Internet. Hell Cannibis is probably going to get legalized everywhere. We might as well call it "Soma" and embrace the brave new world. As long as people have enough to scrape by and some entertainment to distract them, they're not going to eat the rich. There are enough bleeding hearts in the rich upper elite that we still have welfare.
Define AI
Software that learns. The bar is pretty low because it's not that special. Making software that learns complex things quickly is the difference between a rock you can use as a hammer and a pneumatic jackhammer. Just like how "life" includes everything from bacteria to humans.
Show me true AI that is self learning.
Dangerously close to No True Scotsman. But sure, here you go, AlphaGoZero.
it is often trivial to feed the system garbage until it is useless.
Just like people.
A basic feedback loop would at least help point it correctly.
BRILLIANT! You should tell someone over at google that maybe they could feed the output of their neural net back into the inputs. Could be a real game-changer.
(I expect this sort of ignorance from the general populace, but on slashdot this is ridiculous. Come on people)
leftist
Shibboleth
an elite-focused system without even a pretense of noblesse oblige and a terribly smug disdain for the "wrong people" almost all of whom are poor
A lot of people who believe that AI is going to take over a bunch of jobs are talking about Universal Basic Income. That is, we have welfare to support everyone, including the less fortunate. That's pretty much exactly noblesse oblige via the taxes to pay for said system. And this whole article is a tech-company CEO talking and warning about it. And what distain? Who openly shits on poor people?
Who exactly is "them"?
What? No, it's exactly like automated looms. What once took skilled labor can now use menial labor and the output is greatly increased. Automated looms made those expensive guild weavers obsolete and destroyed their jobs. A new job of tending to those machines opened up and was largely filled by street urchins. The industrialist owners got ludicrously rich.
In exactly the same way, specialized knowledge-workers are going to be replaced by a few IT staff (or IT staff at a third-party company) maintaining the machines which do their tasks. Like general practitioners giving initial diagnosis, HR, and paralegals.
We HAVE been here before. And that's exactly the concern. THREE generations of soul-crushing 50% unemployment (for those in the industry), riots, what amounts to terrorism, open rebellion, and sending in the army to put it down. That's pretty much the worst-case scenario. We've got to do better.
Hmmm, that's right. It's not simple a matter of AI replacing workers. The question is how many paralegals does the world need, at what price-point, and how many of them will be AI. Drop the price-point and people find novel uses.
How many doctors do we need? Well, if they were cheap enough I could use a doctor any time I wasn't feeling good. Which would be a LOT more frequent than my current views on the medical industry. Which is pretty much just a long string of expletives about insurance and costs and reproducability and guesswork.
How many office mail clerks do we need? Since they're extinct with the invention of email, NONE. But raises the point that, with new tech does come new jobs. And every place has an IT department, almost like the new office mail room. Just with a lot less employees. Maybe we'll have office AI wranglers.
How many HR staff do we need? As long as I get my check, and there are no issues with vacation, I don't think there are any. I'm not a real big fan of their involvement with hiring either.
How many scientists, programmers, engineers do we need? (Not that I think AI is anywhere near this level, but hey, the tools are getting better. All scientists need to learn to program these days) I think the answer is that there's an unlimited amount of stuff to do for these sort of fields. As long as there's a limit of human knowledge, or things need to get made, the job pool for these guys will be around despite how advanced the field's capabilities become.
And I think there's a similar deal with artists. Maybe even politicians if you REALLY squint and view them like "people who deal with interactions between groups of people". Certain types of needs are elastic and we'll take as much as we can get.
The AI will let you do anything you want. They're tools that open a lot of possibilities. They can automate tasks, but they still need someone to set them up and deal with the output.
Your boss will can your ass the first moment he can. get away with it. ...Although, that said, there are a lot of people that could be replaced with a dozen lines of bash script.
blockchain-based networks
uuuugggghhhhh.
Also, wtf would a block-chain.... network... be all about? How are they using "network" here?
to upend web-based businesses such as Facebook and Twitter
First off, Twitter isn't web-based, it's Internet-based through and through. And there's a compelling argument that facebook isn't really web-based anymore either. Most of their traffic is through phones and their application, which bypasses the web.
Both of these are only networks in the sense that they have social networks (and whatever CDN they run on).
And decentralized facebook has been tried. despora? I think? It didn't go anywhere because facebook has a critical mass of users. Users are the product, why would you switch to a system with an inferior product?
Decentralized twitter... is... fuck man, I still don't even understand this fad. email is too complicated for some people? Newsletters just aren't hip enough? If you made a "I'll sign up for the first 200 characters of your email newsletter, but that's it" button, you'd effectively have the same damn thing. You don't even have to trust the senders, you could just filter shit on your end.
I completely trust foxnews to give me news with a spin.
The writers should take some caution when using a historical or cultural event as a backdrop.
Maybe if it was presented as the focus of the story, or it was a serious piece of work. As a comic-book fantasy-piece, it's not really even trying to convey "This is the truth". Let me put it this way, if someone is 15 and such an idiot that they can't see past the layer of hollywood, they're going to be fucked when it comes to actual political propaganda. Which is a hell of a lot more subtle and can be a lot more pervasive. Hell, at 15 you could even consider this a sort of inoculation to bullshit. At some point they're going to hear more about the 6-day war, and they'll learn that, along with the magical flying woman, not everything they had heard was true. Which is a damn important lesson. I guess I'm saying that we need to lie to teenagers, at least a little, to teach them to question the source and be at least a little cynical. Otherwise we're setting them up for failure. Of course some dumb fucks will blindly believe what they see in the movies. You can't help that.
What you're proposing is that no-one can use any historical or cultural event in a movie, or a book, game, really any sort of media, without "being sensitive" to... something or other. Political finger-pointing? Making people look like "the bad guy"? Maybe for PG movies, but pg-13 and up is fine and should be expected. ....we're going to have movies rated PG-13 for cultural insensitivity aren't we? Or rated R for blatant propaganda. (haha, rated NC-17 for the truth)
ok, sure, "forecast curve". But there are multiple variables allowing forecasts for things like "where should I move rook?" and "should this person go get screened for cancer?" and "Is this a picture of a cow?" and the datasets seem large enough to beat the best players in Chess and Go, predict cancer, and identify objects. Sooooooo.... no excuses, they're being used today for real effect.
And, what about DeepMind Alpha? The thing they made right after DeepMind won big at Go. It doesn't need any dataset, it makes it's own, and beat the previous DeepMind after a rather short amount of learning.
because their domain of knowledge is pinpoint size.
Sure. Nearly all of them have limited scope. But many of them have a scope large enough to give them a hell of a utility. Especially the ones that do the bulk of certain office-worker jobs. They're coming for the paralegals.
I don't want to oversell it though. AI is a useful tool. We don't have any sort of "general intelligence" yet. But to deny that these thing learns and teach themselves.... and that they're not intelligent.... just seems deluded. Like wtf is intelligence if not this?
Google is making self-driving cars... which is hot right now.
And Facebook was at the heart of social media... which was hot back in.... 2006?
Yahoo was hot back when searching the Internet was a big deal. Back in the 90's.
Cloud computing peaked around 2011. But, YES, if you set up a company revolving around self-driving cars, cloud computing, social media, or even search, you are a tech company. The field itself isn't new. But unless you bring something new to the table, you're not going to do very well. This might shock you, but people are still research and advancing the technology of internal combustion engines and making them more efficient, despite being around for... what? A century? That's still tech. And if you made a company dedicated to improving the technology of that old-ass invention, you'd have yourself a tech company.
I think you might be expecting a revolutionary game changer with every new business. We're living in the middle of a technological singularity. Just like the industrial revolution, it's coming in waves. We've got computers, personal computers, the Internet, hand-held computers, and I think artificial intelligence will be another one, if it's not already. But there are still people working on building better computers. It's still a pretty new field, all things considered in the broader scheme of "the economy".
What would you call an algorithm that can figure out solutions on it's own? Self-modifying? Learning? Automated?
Yep, that whooshed me. Waaaay too subtle dude.
And in the 1600's governments had a very cozy relationship with their East Indie Trading companies and the like. And these days we've seen, through the diplomatic cable leaks, that government is somewhat cozy with industry today. The point is, government-industry cooperation doesn't always work out for the best.
Let's refine that. Government funded basic research opens doors to industry and advances technology.
Also, are you REALLY sure you want to try and spin the whole "social credit score" as a positive thing? Most people freak out at that and blow the red-scare whistle. I mean, I understand China is making a credit-score system just like America has with it's **wink-wink** private industry **nudge-nudge** Experion-TransUnion-whoeverthethirdguyis, just with a government-Loyalty aspect. Not that we know what Experion's algorithm is. And of course private companies are using it for profit. It's made to figure out who to give loans to. It's just... kind of a weird example to throw out there as most people would say that's a bad thing.
China using AI to manage pigs?
There's a joke about capitalists in here somewhere, but I can't find it in this pigsty!
Price and computational utility aside, they sound GREAT for researching how biological neural networks work.
Got into a debate with my brother about crypto currencies. He was appalled about how much power they use globally. He's kind of an eco-nut.
There are has statistics about SETI@Home like "Since its launch on May 17, 1999, the project has logged over two million years of aggregate computing time..... With over 145,000 active computers in the system (1.4 million total) in 233 countries, as of 23 June 2013, SETI@home had the ability to compute over 668 teraFLOPS."
Buuuut, how much power consumption is that? And I've really no idea about how much power consumption lies behind all the servers processing VISA transactions, or Amazon's servers.
-wear : A suffix applied to mean the garb associated thereof.
Slacks, n : A name for fancy pants, as opposed to informal jeans or sweatpants. The sort you wear with a suit.
Dude asked where his "flame-resistant suit" was. Other dude made a garment pun.
There. The joke is dead now. It died in a fire.
Syncing a phone with a desktop. Like you walk within wifi range and it logs in, syncs music, text, downloads, calendar (moot for me, as I use google calandar, but it opens up third-party tools).
I'd also like to just have the phone show up as a window on my desktop. So you walk up, open an application, and you can then use your mouse and keyboard on a window that has the phone's screen and you can enter the code and run stuff, etc.
And while I'll always have a desktop for crunching power reasons, most people could get by with the processing power and storage space of a phone. Even at work. But it's torturous and cruel to make them use such a small screen and swiping. So blue-tooth keyboard and mouse.... and wireless or dockable video port and you can have most employees just walk up to any terminal and it's their computer they're typing on and seeing.
I'd also like easy man-in-the-middle HDMI recorderr and IoT tools. Like something that flips on a relay and logs power usage to inbetween a wall socket. Boom, anything you can plug in to turn on is now an IoT device. Anything that turns itself on and or draws different rates of power is loggable with... phone messages or whatever.
For those in power, sure.
But the entire concept of hate speech was to help the powerless.
And organizations of people don't have all the rights that individual people have. Nor the risks. Which is why we form corporations.
We can execute a person for a crime. We don't execute groups of people. At least we're not supposed to.
We can revoke the charter of a big corporation if we think it is no longer serving the interests of society,
Name me a corporation with actual assets that had it's charter revoked.
Corporate personshood is a thing that's gone this way and that over the course of time. I think it would be better for the USA and the world if corporations had less rights. Taking away corporations' rights does not remove our individual rights.
How do you propose to take away the voices of corporations? Shall we pass a law preventing individuals from speaking on behalf of corporations?
No, we legislate that corporations don't have free speech, they only have commercial speech. Like in advertisements. Because while they employ people, corporations are not people.
Advertisements have all sorts of restrictions on what they can say. They can't blatantly LIE. Because that would be fraud. For some things they face COMPELLED speech. All that sped-up medical information at the end of medication ads.
In that way, if a person lobbied their politician, they can say whatever they want. But if a corporate lobbyist lied to a politician, they could be sued.
AI as it currently exists is no more exciting than the assembly line.
... in the 1800's. Yes. It's pretty much exactly like that. It looks to be another phase of the technological singularity which is the computer revolution. Just like the industrial revolution came in a couple waves and CHANGED EVERYTHING computers, the internet, hand-held devices, and now AI have and are going to change a lot of things when it comes to employment, business, and how things are done.
Robotics is great for automation of tasks. The type of AI we have now is great for expert systems and chewing through large amounts of data. The combination of machine learning and robotics have exciting prospects for eliminating mundane jobs.
Yep. Yep (and also things other than expert systems, but sure, close enough). and... No? The combination of robotics and AI is pretty cool for things like Baxter, but less on the AI and more because it's easy to set up. Separately though, Robotics HAS ALREADY been eaten manufacturing jobs since 2000. Output is up, employment is down. That's the worry that the WORKERS have with AI. The bosses? Hey, they think AI is great. But the sort of change that AI looks like it's going to bring about doesn't require robotics. It's stuff it can do in a server farm across the globe. Liiiiiiike, doctors. Doctors get paid a lot. Especially to translate your symptoms into a diagnosis. If a computer could do that better (and people trusted it to do better), a lot of general practitioners would be out of a job.
However we are no closer to hard AI today than we were forty years ago.
Bullshit
we naively assumed that human ingenuity was rendering artificial sentience into something that was right around the corner.
Whoa buddy, way to make the leap from AI to "sentience". You have to remember that society is diverse. The people actually working with the stuff knew it's limits. The people in hollywood are generally idiots. The general masses sadly are more influenced by hollywood than scientists. Which is why it's a tedious chore to call bullshit on all the idiots every time this comes up.
I am no more impressed with computers winning at Go and Chess than I am impressed that a hydraulic press can exert several (thousand) times my strength.
Eh, sure. ok. Both are useful tools. Imagine seeing the first hydrolic press. That'd be a pretty fucking awesome development.
The software that is winning at Chess and Go are, in fact, little smarter than that same hydraulic press.
Wrong. The software is smart like the press is strong.
The software knows from analyzing millions of games that humans have played what winning strategies are, and combines that with brute force strength to know where to optimize its searches.
How the hell did you use "brute force" and "optimize it's searches" in the same breath without an anurism? ANYWAY, glossing over that, YES you're sorta right. DeepBlue and DeepMind analyzed a TON of games and found winning strategies. (JUUUUUUST LIIIIIIIKE HUMANS!) But then there came DeepMindAlpha which DOES NOT require any learning sets and promptly beat the version of DeepMind which did learn from training data. DeepmindAlpha self-taught. No prior experience. "From scratch". If this is your main holdup, then you need to cure your ignorance on the current state of affairs.
We are not close to hard AI.
Probably true. But I'd say we're "closer".
We are not close to soft AI.
Utter bullshit. We had that with the Atari 2600. A single "if" statement. Goombas walking back and
Presumably something that installs programs?
But hip and new and calls them apps?
Come on people, the world of tech is wide and deep. You don't have to define what Ubuntu is, but you do have to define what the hell this new thing is.
Putting "Real" in quotes followed by "would be" implies that there is no real artificial intelligence, and the tools we have available are somehow "fake".
While I understand that people argue over the definition of what artificial intelligence really is, people would generally agree the definition of artificial intelligence certainly includes real, existing, here-and-now tools. They are real. AI is a real thing.
And no, a counter-example to the bogus definition (e.g. the traditional skit: "a True scotsman eats haggis", "NO a TRUE scotsman wears a kilt!") is simply perpetuating the fallacy, and is not required for the fallacy to exist in the first place. The fallacy is a problem because EVERYONE will have some arbitrary definition or goalpost
Your arbitrarily strict definition of intelligence hinges on some sort of nuance about "Totally new scenarios" in italics and everything. And yet you can't describe any such scenario you've been through. So what? You don't qualify as intelligent?
But sure, you want a definition? Let's go with Merriam Webster:
Definition of artificial intelligence
1 : a branch of computer science dealing with the simulation of intelligent behavior in computers
2 : the capability of a machine to imitate intelligent human behavior
Google's:
the theory and development of computer systems able to perform tasks that normally require human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages.
Or dictionary.com:
the capacity of a computer to perform operations analogous to learning and decision making in humans, as by an expert system, a program for CAD or CAM, or a program for the perception and recognition of shapes in computer vision systems.
Personally, I'd just boil it down to "It can learn". Maybe tack on "and updates it's behavior accordingly". That's the interesting part after all. Not that my definition really matters. By any sane definition, we have real AI tools. They learn, they make decisions, and (imperfectly) simulate intelligent behavior. If you want to fear-monger about the future of AI, hey go for it. It's certainly something we should talk about. We should likewise talk about the hype-train. But denying the current state of research and the decades of progress isn't helpful.