Ask Slashdot: What Will Language Be Like In a Future 'Human-Machine Era'? (lithme.eu) 56
Long-time Slashdot reader united_notions is trying to envision "the 'human-machine era', a time when the tech has moved out of our hands and into our ears, eyes, and brains."
Real-time captioning of conversation. Highly accurate instant translation. Auto voice mimicry making it sound like you speaking the translation. Real-time AR facial augmentation making it also look like you speaking the translation. Meanwhile, super-intelligent Turing-passing chatbots that look real and can talk tirelessly about any topic, in different languages, in anyone's voice. Then, a little further into the future, brain-machine interfaces that turn your thoughts into language, saving you the effort of talking at all...
Slashdot has long reported on the development of all these technologies. They are coming.
When these are not futuristic but widespread everyday devices, what will language and interaction actually be like?
Would you trust instant auto-translation while shopping? On a date? At a hospital? How much would you interact with virtual characters? Debate with them? Learn a new language from them? Socialise with them, or more? Would you wear a device that lets you communicate without talking?
And with all this new tech, would you trust tech companies with the bountiful new data they gather?
Meanwhile, what about the people who get left behind as these shiny new gadgets spread? As always with new tech, they will be prohibitively expensive for many. And despite rapid improvements, still for some years progress will be slower for smaller languages around the world – and much slower still for sign language – despite the hype.
"Language in the Human-Machine Era" is an EU-funded research network putting together all these pieces. Watch our animations setting out future scenarios, read our open access forecast report, and contribute to our big survey!
Slashdot has long reported on the development of all these technologies. They are coming.
When these are not futuristic but widespread everyday devices, what will language and interaction actually be like?
Would you trust instant auto-translation while shopping? On a date? At a hospital? How much would you interact with virtual characters? Debate with them? Learn a new language from them? Socialise with them, or more? Would you wear a device that lets you communicate without talking?
And with all this new tech, would you trust tech companies with the bountiful new data they gather?
Meanwhile, what about the people who get left behind as these shiny new gadgets spread? As always with new tech, they will be prohibitively expensive for many. And despite rapid improvements, still for some years progress will be slower for smaller languages around the world – and much slower still for sign language – despite the hype.
"Language in the Human-Machine Era" is an EU-funded research network putting together all these pieces. Watch our animations setting out future scenarios, read our open access forecast report, and contribute to our big survey!
Re: (Score:3)
Don't forget:
"People who bought $thing also purchased $other_thing."
Re:Idocracy [controlled by cunning liars] (Score:3)
Funny coincidence... The FP is on point and related to a piece I just wrote elsewhere... https://wt.social/post/fightin... [wt.social]
Since the Slashdot editor doesn't discourage it, I'll go ahead an include the content, too:
I think we need to know the enemy first, so here's the "ontology of lies" framework I've developed over the years:
Level 0: Self-contradiction. Nothing needs to be checked to know there's a lie there. The only amusing wrinkle is when one lie is contradicted with another lie. More common than you'd think, but that may be because the low-level liars are so indifferent or hostile to the truth.
Level 1: Counterfactual. "Any fool can check the facts." I thought that was a famous quote, but websearch came up dry. And it's still true. And anyone who lies about facts that can be checked is a fool.
Level 2: Partial truth. This is where the lawyers and politicians "work" most of the time. And it actually is hard work to understand the facts clearly enough to know which of them can be omitted. The higher form of this level is deliberately creating false beliefs with a pattern of partially true statements.
Level 3: Deframing. This is where the big money is. These liars can detach meaning and reality and leave you wondering if the bright thing you noticed at noon really was the sun. Many different techniques used by Level-3 liars, but the simplest and most famous is probably the loaded question: "Have you stopped beating your wife yet?" Such a question has established the lie in its premises so any answer becomes a "confession" to the false accusation. Push polls are the current form you probably encounter most often.
If you've been paying attention, you may recognize the higher levels came from something Robert Heinlein wrote many years ago, though his example of Level 3 was more like telling a partial truth in a persuasive way. However, the question of the liar's motivation is important, since a lie is supposed to include an element of the liar's volition. If the liar is just mistaken, is that still a lie? I think so, but it gets complicated when you consider such things as "tone of the statement" and "intellectual malfeasance".
For now this is just a kind of stub discussion for ease of reference. I hope it will be sufficiently searchable that I can find it the next time I want such a link. But the mysterious thing to me is why websearch can't find any of the previous times I've written about the topic... The entire topic "ontology of lies" seems to be some sort of black hole? (Slightly more success by searching for lists of types of lies, but all the lists I looked at were epistemologically quite shallow.)
But I notice the Slashdot editor mangled the list...
Re: (Score:3)
Interesting. Could these be referred to as "The Three Laws of Idiotics"?
Re: (Score:2)
Interesting branding, but I don't think the higher level liars are idiots... However they may be unified be the objective of controlling other people, even when the other people aren't idiots? Some folks keep telling me that some of the supporters of TFG are not idiots.
Re: (Score:2)
Re: (Score:3)
...Alternating with "Your $thing warranty is expiring."
Re: (Score:2)
I think we should be more concerned that what we're headed for is Idiocracy.
We're already there.
I saw a youtube video this morning where a couple of "influencers" were "reacting" to something cultural. Their reactions, that was the video.
My reaction: OMG. They can't even string a sentence of more than three words together.
Re: (Score:2)
How about language in Machine-Machine era? (Score:2)
"Puny humans, their game was over when they installd AI toilets!"
Re: (Score:2)
Mod parent Funny.
Who needs more language than pushing a button?
And what higher purpose can an AI possibly serve than washing my ass?
Re: (Score:2)
Do note that in order to skillfully wash your ass, an AI toilet needs a camera pointed at it to accurately aim at it.
A classic case of sacrificing privacy for convenience's sake.
Re: (Score:2)
Yeah, yeah, yeah, I should have known better than to go for the funny.
Some years ago I actually read a book about the secrets of toilets. It was produced with support from Toto, one of the big developers of washlets. There was a big section about how they got employees to volunteer to help with the testing of the prototypes. They eventually found the best angle for the spray. Was it 37 or 42 degrees? Something like that.
But nothing in the book about the brave volunteers who must have been lost to the danger
Re: (Score:2)
Well I for one welcome our water-spraying AI toilets overlords!
Re: (Score:2)
But we didn't get any Funny mods!
Well lets see now (Score:2, Funny)
I su10001 01at 1t w10010 01 lik1 so01th10g 1ke01011
Based on ... (Score:2)
Might be nice (Score:2)
saving you the effort of talking at all...
Considering how people today sound when they talk, this could be a good thing.
English or Chinese (Score:3)
As almost all of this research is done in English and Chinese, the default machine-human interaction will be one of those, or probably both, one dominant per region on Earth.
Gorram Firefly had it right.
Re: (Score:3)
Re: (Score:2)
Re:English or Chinese (Score:4, Insightful)
Having a bilingual world with a simple to understand language that computers and humans can parse alike, that can express every concept an evolved language like English or French, being that second language, would be a huge step forward. But it's just not practical.
I believe that the ambiguities of a language like English make it easier to lie, and a truly precise language would make it much harder to do so. People in power aren't going to like that, and would attempt to subvert it.
Re: (Score:3)
I believe that the ambiguities of a language like English make it easier to lie, and a truly precise language would make it much harder to do so.
Nonsense. Precision won't stop anyone from making completely false statements. Not that it matters, as it's trivial to lie while only saying true things, no ambiguity required. For example, it's laughably easy to lie and mislead with statistics.
People in power aren't going to like that, and would attempt to subvert it.
"They" would love for the ignorant masses to believe that it's difficult or impossible to lie in their newspeak.
Re: (Score:2)
Japanese is a fairly precise language, but euphemism and deliberate vagueness are very common when speaking it.
Incidentally, Japanese is one of the easier languages for AI to understand. It's very regular and while there are some fairly loose accents, it's relatively easy to parse. When I need to use voice control I usually default to Japanese for that reason. Google's assistant lets you seamlessly switch between languages, even half way through a conversation.
Re: (Score:1)
Actually some machine translation tools, be it text or spoken language, use Japanese as intermediate language.
Re: (Score:3)
English, for sure. Even aliens speak decent English. Don't you watch movies?
I could care less! (Score:1)
Nope (Score:2)
Re: (Score:2)
Simnple answer (Score:1)
The answer to all the questions posed is NEVER, and NO. It is (one of the) most useless scheme's ever conceived -- which just goes to demonstrate that most persons are complete utter idiots! Like other scheme's of this nature (there have been several) bankruptcy (or prison, or the gallows) is the outcome to be expected.
After playing with GPT-3 (Score:2)
Horrible? (Score:4, Insightful)
In some ways, I'm kinda glad to be in my 50's and already tiring of these types of advancements, despite being a developer by trade.
I'm becoming more jaded each passing year toward an industry that I see has little to no purpose in terms of the advancements made, except to capture data.
Data is the "new gold", which in turn equals less privacy, fewer secrets and more control by external forces who use that data to shape our lives.
We know - or at least should do - how much our private lives are already tracked. The mobile devices people carry pinpoint not only location, but habits.
The data can pretty much tell the type of activity a person is doing at any given time - that smart watch on your arm, heck, it probably _knows_ when you are having sex, if you are wearing it.
The smart TV you may be stupid enough to connect to the internet, knows your viewing habits, each click - every pause, every "channel" change.
The payments you make with your mobile device, are tracking your location, what you purchased and will know where you were before that and after - they can profile your entire day.
Once we start adding more advanced voice recognition, face and body tracking, metrics on interaction with AI ... hell, you are marked.
And who provides all of this?
The huge tech giants.
The real danger will happen - and it's starting to already - when actively opting _out_ of this crazy way of life, is firstly deemed socially odd and then ups the ante, to being socially dangerous to being considered a danger to society.
When it reaches a point where having no metrics tracked and having no "smart" devices about your person, means you cannot purchase anything nor interact in a meaningful way in society, is when ... well, what then?
What are we at that point? Just part of the machine?
I hate the way this is going, for so many reasons, but mostly because it is all being developed by a monopoly of tech companies who do NOT have your best interests at heart.
I predict revolution - an almost luddite anti-tech movement - at some point.
Re:Horrible? (Score:4, Interesting)
There is not going to be any revolution because people are too busy liking stuff on social networks and posting videos of themselves for others to like. And nobody really cares about being monitored online. The convenience far outweighs the loss of privacy. It's not just "the majority doesn't care", it's "nobody cares." Or at least nobody cares enough to make a difference.
Fear not tho, because our whole technological edifice is built on shaky foundations. As long as we are in peaceful times, our interdependent system of intertwined technologies is OK. But it is very sensitive to climate change, resources depletion, natural disasters, political unrest, and of course war. And the most recent generation of equipment is more brittle than ever, needs constant care, maintenance and regular replacement, partly due to planned obsolescence, and partly due to working with raw materials whose quality has lowered due to finding workarounds to resource depletion and supply chain issues. We are already reaching a point where network failures start happening and people are left out without internet access for months in a row. Network equipment companies have a backlog of orders of more than 10 months, meaning that if a customer has an issue with some network equipment, they will have to design an ad-hoc solution instead of yanking out the failed device and yanking in a replacement sent by the company.
At some point stuff is going to hit the fan and we will see more and more frequent interruptions and failures of what we currently are taking for granted - internet, phone networks, etc. The whole talk about AI and multiverse and all that is completely disconnected from the reality that surrounds us. The future is not going to be as digital as we seem to expect. Whether we like it or not, we are going back to an era of scarcity, of shortages of everything. And the more complex, interdependent things, are the things that are going to fail first.
In the long run we won't have any choice but to opt out of our crazy way of life - it's life itself who is going to opt us out of our crazy ways.
for the better maybe!
Re: (Score:2)
I predict revolution - an almost luddite anti-tech movement - at some point.
You don't have anything to worry about. This is all just silly science fiction nonsense.
Re: (Score:2)
I have been thinking about this within the context of corporatist libertarianism and third-party doctrine access of data by the government, and the only credible response is nullification of third-party doctrine which would redefine privacy rights.
And while libertarians would chaff at the idea business regulation, they would have to concede the government collecting the same data would be an affront to civil rights.
Except by third-party doctrine, there is no tangible differentiation between the two.
So fine,
Re: (Score:2)
We will lose our figures of speech (Score:2)
WFT? (Score:2)
I Have No Mouth But I Must Scream (Score:3)
"I Have No Mouth But I Must Scream" comes to mind for some reason. I read it long ago, and I'm not sure it applies...but the phrase seems apropos.
With a brain-computer interface ... (Score:3)
... the increase in bandwidth will be 100x to 1000x .... or something in that range. This means that groups of people having this will actually literally appear to have a hive-mind. From the PoV of a regular unmodified human they factually *will* be a hivemind. Traits of individuality will be so subtle amount these and probably expressed in manners regulars won't be able to pick up that they'll be alien to us, very much like dogs or cats might enjoy our company but can't fathom what we're discussing here on slashdot ... for example.
Regular language won't change, but bci users will have different ones on the radio spectrum.
Re: (Score:3)
the increase in bandwidth will be 100x to 1000x .... or something in that range.
There are other bottlenecks. Try hooking up a 1Gbps ethernet connection to a 80486 with a classic 40MB HDD and imagine how much use the increased bandwidth would be. And that is an example with compatible technologies.
Getting human brains to work 1000 times as fast and deal with all this increased bandwidth means solving huge organic engineering challenges. Increased power consumption, increased cooling requirements, etc.
Don't get me wrong, research into BCIs is fantastic and will bring many useful insights
English (Score:2)
Anti-manipulation is the future (Score:5, Insightful)
OP focuses on real-time language translation, which could be useful but won't impact day-to-day life very much for people in developed countries.
The real question is: What happens when these same technologies (Turing-passing chatbots, lifelike voice and facial synthesis) are put in the hands of marketers and propagandists? Eventually you flood every channel with so much shit that the only reasonable response will be to stop all interaction with unknown actors online.
As Daniel Kahneman points out, the risk here is that we humans don't even realize when we're being manipulated. So in the end, every sane person will want sophisticated systems to protect themselves from manipulation. I don't know what it looks like technologically but it will be necessary if we don't want to become mindless cogs, or tear apart our society from within.
Ask and you shall receive (Score:2)
https://www.youtube.com/watch?... [youtube.com]
Captioning? (Score:2)
Talking past the sale now, Slashdot? (Score:2)
Don't hold your breath.... (Score:2)
Machine translation is good for basic understanding and formal technical documents, just, but horrifically bad at casual conversation
This is also why chatbots are also really bad and cannot pass Turning tests reliably
Brian interfaces are horribly bad, they barely function at all even after years of development - again they will come but not yet
Strong AI is non-existant currently, it likely will arrive but not yet by quite a way
Ænglish (Score:2)
It will still be English, however, it will be the Autistic dialect of English, called Ænglish... https://www.youtube.com/watch?... [youtube.com]