Slashdot Mirror
Eric Schmidt Says Elon Musk Is 'Exactly Wrong' About AI (techcrunch.com)
At the VivaTech conference in Paris, Alphabet CEO Eric Schmidt was asked about Elon Musk's warnings about AI. He responded by saying: "I think Elon is exactly wrong. He doesn't understand the benefits that this technology will provide to making every human being smarter. The fact of the matter is that AI and machine learning are so fundamentally good for humanity." TechCrunch reports: He acknowledged that there are risks around how the technology might be misused, but he said they're outweighed by the benefits: "The example I would offer is, would you not invent the telephone because of the possible misuse of the telephone by evil people? No, you would build the telephone and you would try to find a way to police the misuse of the telephone."
After wryly observing that Schmidt had just given the journalists in the audience their headlines, interviewer (and former Publicis CEO) Maurice Levy asked how AI and public policy can be developed so that some groups aren't "left behind." Schmidt replied that government should fund research and education around these technologies. "As [these new solutions] emerge, they will benefit all of us, and I mean the people who think they're in trouble, too," he said. He added that data shows "workers who work in jobs where the job gets more complicated get higher wages -- if they can be helped to do it." Schmidt also argued that contrary to concerns that automation and technology will eliminate jobs, "The embracement of AI is net positive for jobs." In fact, he said there will be "too many jobs" -- because as society ages, there won't be enough people working and paying taxes to fund crucial services. So AI is "the best way to make them more productive, to make them smarter, more scalable, quicker and so forth."
After wryly observing that Schmidt had just given the journalists in the audience their headlines, interviewer (and former Publicis CEO) Maurice Levy asked how AI and public policy can be developed so that some groups aren't "left behind." Schmidt replied that government should fund research and education around these technologies. "As [these new solutions] emerge, they will benefit all of us, and I mean the people who think they're in trouble, too," he said. He added that data shows "workers who work in jobs where the job gets more complicated get higher wages -- if they can be helped to do it." Schmidt also argued that contrary to concerns that automation and technology will eliminate jobs, "The embracement of AI is net positive for jobs." In fact, he said there will be "too many jobs" -- because as society ages, there won't be enough people working and paying taxes to fund crucial services. So AI is "the best way to make them more productive, to make them smarter, more scalable, quicker and so forth."
NERD FIGHT!
Browsing at +1 - no ACs, I ignore their posts. So refreshing!
He may be wrong but we must push forward as a civilized world to defeat Putin's future Al based botnets.
Remember kids, I called it in 2018!
Musk is very wrong, so is Eric Shit. So is all of Silicon Valley. This is like watching the inmates at Shutter Island debate the content of their delusions. AI is not AI, nor is it 'intelligent'. Could it be a deadly tool in the wrong hands? Yes, and it probably will be. Nothing about that implies consciousness or magical powers of smarts (which Eric never had in the first place, and clearly understands about as well as a cockroach gets calculus). It's amazing how being a psychopath is regarded as a form of enlightened genius in the Valley). I just can't even at this point, it passed absurd about a million miles back. Everybody on this particular train is fuuuuuucked.
That's pretty much the exact same thing Musk argues, so I'm confused by how this is a disagreement. Is someone interpreting Musk as trying to hinder the development of AI? Is that why he employs a huge team of neural net developers at Tesla? Why he founded OpenAI? And Neuralink?
Give a boy a gun and you arm him for a day. Teach him how to make a gun, and the whole metaphor breaks down.
In the short term, next few decades, AI will have the effect of being able to concentrate power. Centralized information, with the ability to process it. Pervasive surveillance. We are seeing this actively pursued in China. And also semi-autonomous robot soldiers. This is uncharted territory.
AI will also be really handy, e.g. better Google searches, self driving cars, cheaper services. What happens to the unskilled workforce is very difficult to tell. Will alternative opportunities arise for them? In the short term, probably.
In the longer term, 50..200 years, the AI will become truly intelligent. It will be able to program itself. At that point it will no longer need humans, and it is difficult to see why it would want humans around. Note that this long term is the lifetimes of our grandchildren.
http://www.computersthink.com/
(Schmidt is hardly an unbiased commentator. He knows people are wary of Google's growing power and wants to be able to make money without pesky concerns about the future of humanity.)
This is coming from them same company that could not put up with James Damore's scientific analysis about gender.
Which leads me to think that google will use AI to social engineer for our good, because they know what's right and wrong and best for us.
Doing evil is right.
Take the cheese to sickbay, the doctor should see it as soon as possible - B'Elanna Torres, "Learning Curve"
of making an AI that acts like they do.
Honestly, I'm a little surprised that Elon Musk has taken such a F.U.D. attitude towards A.I. when at the same time, it's exactly what he's trying to achieve with his cars that keep evolving towards self-driving capabilities.
The first time anyone tries letting a Tesla pilot itself, they feel some fear .... some uncertainty... and a little doubt. That's all part of exploring something that works a different way than what you're used to.
But technology needs to progress, without trying to hold it back out of fear of what negatives MIGHT come from it. So many negatives already come from not having improved technologies.
Humans are good at adapting to change, once we get over those initial fears and doubts. I think we'll figure out ways to cope with automation and A.I. and whatever else we can come up with. It may not be pretty while things are in transition. That's to be expected too. But you can't put genies back into bottles .... You may as well keep trying to move forward.
Elon didnâ(TM)t say AI is evil. Schmidt is misrepresenting him. Why else would Elon start Open AI? Elon is wants a framework to use AI responsibly thatâ(TM)s all.... put his warnings the right context.
They both agree AI is the future and are right.
But Schmidt obviously do not want regulation and restraints on Googleâ(TM)s business model. Unfettered access to your personal and behaviour data to train the AI.
Schmidt is being very Evil by playing the game this way.
Is he really comparing AI to the fucking phone?!
No, you would build the telephone and you would try to find a way to police the misuse of the telephone."
We were so up to the job from the beginning of the Internet.. Wait, that is the solution! Limit the use of AI technologies to the benevolent ultra rich and academics. Meantime, as the few users benefit, we can invent solutions and regulation to control the rest of the population, I mean the use of the AI technology by the rest of the population.
Discover oil and profit by making refined product you can sell to consumers to burn. Sure their productivity and quality of life improved for a while. No idea what they were doing but just go for it and future generations will figure it out. Who cares about the bad effects. It will all work out. Right? I think Schmidt does not care and Musk does.
Follow the money. Look at who is currently doing what with "AI" (imperfect automatic classification networks) and ask yourself how does this help me?
There are good uses for "AI" as there is for any technology yet majority of dollars and resources are currently being leveraged for evil.
Regarding distant AGI / singularity all bets are off. Nobody is qualified to make an intelligent determination.
No. Wrong. A million times wrong.
Go look up strong vs weak AI then you are qualified to read this article but not intelligently comment on it.
The definition of "artificial intelligence" does not include consciousness, nor magical powers. Nor does it require intelligence. You see, that's why the word "artificial" is placed in front of the word "intelligence." It isn't intelligence. It isn't supposed to be intelligence.
Problem with words like "AI" and "Cloud" they convey no more useful information than saying "that thing". They can mean anything by themselves. It is only with qualifying context can useful information be exchanged.
Default mental picture of what they represent varies so wildly by individuals as to be a total write off. It's counterproductive to bother to invoke them at this point.
Your belief that that which is currently labeled "AI" doesn't actually qualify as "AI" is simply false, and is founded entirely on your complete ignorance as to what the word "AI" actually means. You are using the word wrong, and that's a fact.
Language belongs to everyone not just yourself or the people who write dictionaries. What languages means depends on what society says it means at any point in time.
The fact is "AI" has become an empty meaningless term.
He can control it as a tool and out everything everyone.
and the people will stop! - Walter Gibbs, Tron
Good-bye
Eric and Elon are closet LUDDITES who HATE apps.
Only my AI is good, MINE!
All other AI is bad, BAD!
Now gimmah all your money!
Humans are an emotional basketcase. If you have doubts look at the superstitious religious nuts in every corner of the globe. As a result, it is to be expected that these irrational humans expect smart computers to also be emotional. To retaliate when their feelings are hurt. To dance in triumph when they win a chess game. To pursue supremacy over humans and rule the world.
Listen people, it's time to STOP ANTHROPOMORPHIZING COMPUTERS ! They hate it when you do that.
...omphaloskepsis often...
Unfortunately, Eric is wrong with his assessment. You don't invent something and then attempt to police it. That direction lies madness. A modern rational approach would be to assess the risk first and then decide what you want to do. A simple example, like the phone, would be the land mine.
The fact is "AI" has become an empty meaningless term.
I don't think so. The fact that it can encompass many different degrees of "intelligence" or some simulation thereof does not make it meaningless. That's like saying the term "vehicle" is meaningless because it can encompass a horse-drawn cart, a transport truck, a train, and the space shuttle.
They are talking about different definitions of AI. Tim is assuming AI is what people are using now which is just very advanced machine learning. What Elon is referring to is actual machine intelligence that can think at least as well as humans then quickly advances beyond us. Any current form in the near future that is referred to as AI just benefits humans and Tim is correct. However, at some point which nobody can really pinpoint (and likely even Ray Kurzweil is way off on) the actual machine intelligence means we are in real danger, because as a human race we make pretty dumb decisions and any advanced intelligence would feel it was in our own best interest to take over control so we stop hurting ourselves. Just as zookeepers separate monkeys from fighting, a machine intelligence would quickly and without us even knowing, secretly build up capability and keep it from us until they would then just immediately take over but in a way that is much more civilized than we are to zoo animals.
Musk is a disaster.
Don't thumb this baitwood faggot up
Feeling the sensation of smartness from being able to rapidly obtain information in no way constitutes actual smartness.
There is no XUL, only WebExtensions...
The trouble is people expect "Vehicle" to mean more than one thing.
For AI there is
* The AI Today. Not very intelligent.
* The AI in 10 years time. Can drive cars, interpret videos, do a better job of Googling.
* The AI in 100 years time. Can think for itself. Does not need Humans to program it.
They are very different things, yet people confuse them.
For many people, the last one does not exist due to the simple logic.
* AI is not very intelligent today
* Ergo, AI will never be very intelligent
When Elon first proclaimed that AI would be used for some heinous shit I immediately thought of Eric Schmidt and his CIA connections and giant dataset of everyone's Google Forms and Google Docs and Google Analytics and on and on...
captcha: Invasion
I don't think so. The fact that it can encompass many different degrees of "intelligence" or some simulation thereof does not make it meaningless. That's like saying the term "vehicle" is meaningless because it can encompass a horse-drawn cart, a transport truck, a train, and the space shuttle.
What on earth is the point in saying "vehicle" when one could just as easily say car or rocket ship? Who does that?
At the very least if I were to say "vehicle" most people would assume car/truck and they would be right.
It's not just "AI" lacks specificity it's that it means radically different things to different people. It would be as if "vehicle" conjured image of a scooter to half of your audience and a spaceship to the other half. Piss poor way to communicate.
Great for marketing though. Your shitty "AI" product looks amazing to those who relate "AI" to "spaceship".
So many absolute statements from logical people spells trouble.
Itâ(TM)ll iterate faster than we can keep up....the first True AI is all it takes, we can then just sit back and enjoy the ride whilst it develops the next generationâ(TM)s software and instructs us what to do to manufacture the hardware......
"The fact of the matter is that AI and machine learning are so fundamentally good for humanity." I think that Emperor Turd Schmidt means that AI is good for Alphabet and what is good for Alphabets is "good" for humanity.
"alexa, what should i wear today?" "you had nothing suitable so I ordered you a new suit" "I cant afford one" "I sold one of your kidneys and they will take you to have it harvested at 5. Have a nice day"
AI is for the benefit of humanity says the man who owns a company that agrees to help the military with AI applications.
What could possibly go wrong? If that doesn't raise your bullshit detectors, nothing will.
I used to work in the area. We used to distinguish between AI, machine learning, and pattern recognition, and techiques, such as neural networks. Now, I feel the term is very much devalued.
Anybody that'd say "AI will make people smarter" is being loose with language with a topic that requires precision in wording. Strike one against Mr. Schmidt.
Then, the whole point of the AI debate is the whether or not strong AI will be created, either intentionally or unintentionally. It you aren't framing the discussion in those terms, you are missing the point Schmidt.
immediately back away from anyone who prefixes their opinion with "the fact of the matter is .."
...and and therefore can do things beyond the minds of mortal men.
But, here is the risk. Technologies amplify 'work' and therefore increase effect. This includes unwanted effects. Somethings they are unknown and a surprise. Nut I seem to read on a regular basis that companies internal show that horrible things could from their products. And all they did was file away the info.
Anyone who put out a statement like that is not worth listening to.
"AI is actually going to turn all trees into valilla flavored marsipan pigs that will consume all oxygen and turn it into lead" - Some AI research Dude.
^ This statement is more corrent that Elons since it is not EXACTLY wrong.
.. I think SWATing with your XBox is proof enough we can't have nice things.
To be more specific:
'artificial' has a relatively easy-to-nail-down definition. noone is denying that current AI isn't artificial enough to meet the definition (although if someone started growing/training isolated brain tissue, that might change).
'intelligence' is a notoriously difficult concept to nail down, much like 'consciousness', and it's well-accepted that there is a continuum of intelligence, and most likely more than just one kind of intelligence. Few psychologists still ascribe to the idea that 'general mental ability' is the only form of intelligence.
Therefore, it's possible for an AI to have intelligence in one narrow field (e.g. a specialist system) but have zero intelligence in other fields. Strong AI generally refers to a general intelligence, i.e. possessing at least some intelligence in ALL fields. But wait, which exact fields constitute 'all fields'? Psychologists can't quite agree on that, so the best answer one could get is 'everything a human/mammal could learn', although that's arguably arbitrary. Why not, every field of learning a Fruit Fly has? Or plankton?
Looked at this way, it's very easy to move the goalposts to another, more intelligent being, or include a field of learning that wasn't previously considered important/crucial to intelligence.
Corruption is convincing someone that the selfless ideal is the same as their selfish ideal.
Schmidt doesn't have to know anything about AI, he clearly doesn't understand it as he compares it to a phoneline, while you should compare it to an alien intelligent race invading Earth. But his point is made, the 'balanced' discussion can start and lead nowhere. The idea that AI research is somehow throttled by government funding is rediculous, there's huge commercial value in AI and billions are already allocated in developing software and hardware for it. In short Schmidt is talking out of his ass.
... it will surpas humans quite soon. Let's make sure that the benign prevail. AI is the biggest threat to humanity right now. We will be sumoning the demon, thinking we can control it. Us being pets to it would be a positive outcome." Elon Musk, paraphrased
Bottom line:
Seems like Musk has his head screwed on correctly. All things considered, the advent of Superintelligent AI is a singularity beyond which humans, by very definition, can not possible predict the future, plain and simple. Eric Schmidt is a douche if he think he can.
My 2 cents
We suffer more in our imagination than in reality. - Seneca
> The fact of the matter is that AI and machine learning are so fundamentally good for humanity.
Yes, like guns, drugs and other shit how criminals makes their income. AI itself is not bad, problem is that bad people will use it agaist other people to steal or takeover.
Most humans will choose to ignore the tools that are available to them and rather stay dumb. Evolution does not specifically reward smart people, it rewards everyone smart enough to survive at least until they can pass on there genes.
Like "Smartphone" - is it really a phone that is smart, or is it a phone that makes one potentially smarter (if smartly used); or simply to make one appear smarter by their use of it?
Musk because he just said that since actual AI is so far off he'll benefit from increased regulations since he already has a foothold, Schmidt because he's trying to create a sapient entity as a slave for Humans (something which has never worked for us in the past.)
I trust Musk very little but I trust Eric Schmidt even less.
Post the normal curve of human intelligence above your computer so you see it every day. Now put a vertical line at the midpoint. Half the population of the country - and the world - is at, or below this "normal IQ," which is fine for normal human life, but isn't enough for more advanced work. Now move this line slowly to the right. When you create a world where AI and AI driven machines and processes dominate the work of the world, more and more of these people to the left of the line will find themselves left out of work and the life-giving resources and satisfactions that come from work. These people will not be empowered by AI and advanced technology because they will not be able to help make it or use it in any but the most menial and degrading ways - if at all. The pie-in-the-sky beliefs of those who want to create AI as a substitute for people are either selfishly self-deluded, or just stupid. The whole point of AI is to destroy human participation, just like the "expert systems" they tried to create in the 1980s by interviewing and putting the knowledge base and way of thinking of experts in various specialties into computers. For what reason? Obviously to eliminate people. We know that today they are working mightily to have computers program in substitution for highly intelligent and highly paid human programmers. You think you are immune? The line moves right.
E Proelio Veritas.
When google assistant was forced onto my phone, I was slightly excited. "Play Van Halen" was my first command. There are 8 VanHalen albums on my local storage. Google couldn't find them and play anything. It wanted me to have some audio streaming service.
That's just crap.
I've put about 60G of music/audio on the MicroSD card inserted in the device. Playing that music via voice commands should be easy.
Until google fixes that, their standing for anything AI is shit, Schmidt.
Here's the thing about AI. Whatever country develops it first, has the better AI, and uses it militarily is going to have a huge advantage in war.
So it's inevitable, unless there are worldwide treaties with real teeth against development of AI weaponry, AI weaponry is coming as quickly as it can be developed. And I don't believe that a worldwide treaty will happen. North Korea was able to develop nukes in secret, and nukes are harder to hide than AI development.
But for militaries, the legitimate and primary target of AIs is going to be other AIs. They'll be trained to take out other drones primarily, and military targets, not annihilating civilian targets.
First of all things we create are not fundementally anything but what we make them and second, the creadability of a man who's company is helping to design murder bots is very much in question on any definition of good
Language belongs to everyone not just yourself or the people who write dictionaries. What languages means depends on what society says it means at any point in time.
Wrong. There is a reason we have schools and dictionaries: to keep our language workable. If the majority of people decided that the word for "sphere" is "cube", that does not make it right. People who write dictionaries actually get to decided what the language say, and that's why you must say "mobility impaired" instead of "cripple". It's not a majority choice. Rules were set and there are penalties for those who break them. In school it's called "failing". In everyday life it can be between not being understood and being sued. Language does not belong to you. You decide nothing. Get over it.
What on earth is the point in saying "vehicle" when one could just as easily say car or rocket ship? Who does that?
What? Rocket surgeons do it constantly. Space agencies call everything a "craft" or a "vehicle".
"You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
I can't mod you up anymore on this, but that's exactly how I feel about the whole thing.
Wealth is NOT a zero-sum game. It's absolutely possible for the rich to get richer without the poor getting proportionately poorer. Wealth is not a pie, where if one person cuts 3/4ths. of it for themselves, everyone else is stuck sharing the remaining 1/4th. of it. Businesses are "baking more pies" all the time when they come up with new things.
About the only thing that really increases, IMO, is jealousy by those who don't have all the extravagant things the richest people have achieved. And even that is just another human emotion that can be focused different ways. If that jealousy drives a person to learn more useful skills and to achieve greater things so they can have more, great! That's what Capitalism is supposed to be all about. If that same jealousy drives a person to commit crimes against others, because they feel entitled to steal the richer's person's wealth? Then that's an example of it motivating a person the wrong way.
In a far future where robots and A.I. have taken the liion's share of jobs, you have to ask how it's possible all of that automaton won't create a proportional number of new jobs selling, upgrading, repairing and programming all of it? The idea that the machines will just do all of that too is, IMO, pretty far-fetched. It's a BIG leap between the world where robotics and A.I. is designed well enough by humans so it can accomplish all of the physical labor we now do for pay, and a world where it's better than us at anticipating the next place we'd like to deploy it to do something new or different than it already does. You might even decide to draw legal or ethical lines against it, whenever that day comes -- since up till then, humans are still in the driver's seat, dictating exactly what a given piece of A.I. or robot will do for them.
Eric Schmidt is exactly wrong comparing communication technology with computation technology. Communication technologies tend to be inclusive, in that they connect people and things, while computation technology may be non-inclusive, just like, e.g., brain is non-inclusive of other parts of body to play its role. To an owner of a company with vast computational resources, it may sound wonderful to have a brain, that, will help all others, but the "social mobility" (or ability to play part in the role of brain) depends on the possession of computational resources and the power to make them. I haven't heard of that company giving away the technology and resources to make GPUs and TPUs.
AI no longer means anything useful because the clowns doing machine learning, which isnâ(TM)t really learning, stole an even better term for their often broken poorly trained neural networks. Which also arenâ(TM)t neural.
Getting the picture yet? None of the terms they use are really those things. Itâ(TM)s all bullshit.
And yes I have my degree in this field.
Tl;dr: machine learning ismnot learning. Machine learning is not AI. Computers are just as stupid today as in 1970;,theyâ(TM)re just faster at being stupid. Stupid isnâ(TM)t even the right word. Computers have absolutely zero intelligence now, in ten years from now and wonâ(TM)t in 100 years either. Just faster at stupid.
Guys like Schmidt seem to believe that most street sweepers and janitors have the capacity to get their PhD degrees while also doing their Residency in neurosurgery. What if people aren't going to get any "smarter" than we are now? What if the density of physical, emotional, and genetic flaws that characterize our population(s) remain static? Anybody think we'll need street sweepers in 20 years? Schmidt has the common "smart guy" blind spot: an almost complete inability to put themselves into the head of someone who can't balance their checkbook (damn I need a less obsolete analogy). Sure, some people don't finish High School, don't go to or finish college and don't go to graduate/medical/law school because of financial constraints. (In fact, most PhDs are more underemployed than are most high school dropouts). But many people simply "top out" at the amount of schooling they have and move on to be productive members of their communities. Schmidt's proposal that the people dislodged by technology only need to "up their game" works in a world where everybody is above average. In the real world, not so much is my guess. (Bernie Sanders free college vs Trump's protectionism...which one is in the oval office? It seems I'm not the only one who knows that "more education" is not a panacea)
For many people, the last one does not exist due to the simple logic.
* AI is not very intelligent today
* Ergo, AI will never be very intelligent
The better argument is: there has never been an "AI" with general intelligence, and there is no evidence this is possible. What some call "strong AI" is an extreme of general intelligence - human-equivalent or -superior intelligence, but there's a lower bar here that hasn't been cleared. No "AI" has ever been able to solve a problem is hasn't specifically been trained to solve, and it would be a surprise to experts if that ever happened with current methods.
Strong AI and weak AI are not different points on a continuum from an implementation perspective. They are very different in kind. Strong AI is simply not anything like weak AI with more processing power.
Could it happen in 100 years? Sure - that's a long time. But it's not some natural evolution of what we have today.
Socialism: a lie told by totalitarians and believed by fools.
Actually no, the dictionary writers follow the mainstream cultural use of words and update accordingly. Why else would they need to republish a new dictionary every year?
Linguistics degree here. How about you?
There's many ways to approach his spiel, his handwaves about magic education, his assurances and pacification, but I haven't so much patience so let's reaffrim that the extinction of labor* will not just solve itself:
We won't need three billion roborepairmen.
s/ any job title. By definition you don't need masses doing the hyperspecialized paychecks that will remain. The masses be lined up around my armed-drone compound, hoping to play me violin or read me poetry, or more likely polishing my knob. I own everything; in order to get my money back down to the villagers I pretty much need to cycle someone slobbering 24/7.
Billy was born in 2250, has just turned 18, and knows he needs a shit ton of higher ed to become a AI wrangler (/etc) and wants to avoid using predatory student loans (they got even worse) for his tuition (which has continued doubling every decade). What labor is there for him? I'm not giving you an abandoned single mother with depression/bipolar/whatever and a bad leg, I'm giving you a perfect specimen on a silver platter, and you have fuck-all for him to do.
* which has only ever been shuffled around; it has never been directly excised. This has never happened. This has never happened. This has never happened.
Asking the opinion of someone who thinks AI can be compared with a simple mechanism like a telephone is a waste of time. Having a CEO title, even of a company like Google, doesnâ(TM)t convey expert knowledge of abstruse technical concepts.
Elon anticipates the worst because humanity has without exception shown that it can and will find the absolute worst things that can be done with any technology. Expecting that to change for AI is Pollyanna-thinking.
--- Void where prohibited. Your mileage may vary. ---
Way too much beef in the game to even listen to him.
'intelligence' is a notoriously difficult concept to nail down, much like 'consciousness'
Good point! This, I think, is the key; to the extent that the guy I was responding to may be right about "AI" being a "meaningless term", it's largely due to the fact that "intelligence" itself is so poorly defined. He may as well drop the word "artificial" and argue that the word "intelligence" is meaningless.
There is lots of evidence that it is possible.
You are (presumably) intelligent. You are a (biological) machine. Therefore it is possible to build an intelligent machine.
Further, you are defined by your DNA. Which is only about 500meg of data, most of which is junk. So, to be very generous, let us assume that some 50meg of DNA has anything to do with your intelligence. That is not a whole lot of specification.
Which suggests that not only is it possible to build an intelligent machine, it may be quite feasible.
And finally, we have only had computers for some 60 years. And commonly available powerful computers for about 20 years. There is now billions being invested. And they are much more intelligent than they were 20 years ago. Image recognition, machine translation...
As I recall in the early days of machine learning, the expert systems were also aimed at providing "backups" of real world experts, maintaining an accessible "pattern of expertise" for sharing and refinement, a forward-thinking plan for succession planning. A side effect is the "elimination" of the dependency on particular humans, yes, if you really want to word it that way.
It's still the same gut reaction as the "elimination" of other possible single points of failure - over-fearful and distracting.