Given the current state of language translating software, I think this is highly unlikely in the next 10 years. Anyone who is bilingual and has used a translation app knows how woefully inadequate it is for actual communication. The nuances of language are far too complex, not to mention the importance of non-verbal communication. How can software correctly translate the answer "Well, I'm not sure..." (which could signify hesitation to accept, passive disagreement, a pause for time to think, etc.) without understanding the context of the answer, and the tone of voice and facial expression of the speaker? Not in 10 years...
I think it's possible. Human brains are not wired to understand exponential growth. I don't think it will happen within 5 years, but 10 = very possible.
You've just said that human brains "are not wired to understand exponential growth" but then made a prediction based on understanding exponential growth. Plus, you haven't established the doubling factor for the growth - it might be exponential but needing 20 years to get twice as good.
"Perfect instant translation" requires full AI. Consider the Borges story "Hombre de la esquina rosada". That translates directly to "The Man of the Pink Corner." But the "corner" here means also the corner store, so it's not just the physical corner but also the "hangout for the low-life of the barrio." (I'm quoting from the translator's notes here.) Moreover, the "pink" refers to the pastel hues popular in earlier decades, so the pink implies an older neighborhood "populated by toughs and knife fighters."
If you knew Buenos Aires of 100 years ago then you would infer that directly from the title. But without the context you just know it as "the pink corner."
I've thought that "The Hood of the Pink Corner" would be a better translation, given the double or triple meaning of "hood" to mean "hoodlum", "neighborhood", and more specifically "poor neighborhood." It's not as direct, but it might be better.
How do you explain those nuances to a computer?
Douglas Hofstader gives more examples of the difficulties in natural language to natural language translation. How does one translate a nonsense poem like "Jabberwocky"? One French translation starts "Il brilgue: les tôves lubricilleux. Se gyrent en vrillant dans le guave..."
Or, and I quote from his Gödel, Escher, Bach: Take a look at the first sentence of Dostoevsky's novel 'Crime and Punishment' in Russian and then in a few different English translations... The first sentence employs the street name "S. Pereulok" (as transliterated). What is the meaning of this? A careful reader of Dostoevsky's work who knows Leningrad ... can discover by doing some careful checking of the rest of the geography in the book ... that the street must be 'Stoliaryny Peteulok". .. should we keep the initial so as to reproduce the aura of semi-mystery which appears already in this first sentence of the book? We could get to "S. Lane", ("lane" being the standard translation of "pereulok"). None of the translators took this tack. However, one chose to write "S. Place". [The second] gave the translation as "Stoliarny Place". What about number 3? That's the most interesting of all. The translation says "Carpenter's Lane." And why not, indeed? After all, "stoliar" means "carpenter" and "ny" is an adjectival ending.
He then goes on to argue that perhaps we should just read Dickens instead, since Dickens is, at some level, a "translation" of the Dosteovsky novel.
BTW, Google Translate gives "S. Lane", which none of the three translators used. As that's the obvious and most direct translation, which the humans didn't choose, I assume that we can say it's not the perfect translation.
How would an automatic translation system, without AI, look at "С. Переулок", figure out that it's supposed to be "Столярный Переулок", and translate it correctly?
Just because human brains are not wired to understand exponential growth doesn't mean you can't understand it. I'm talking about human intuition.
I understand the problems inherent in translation. But again, I still feel it's possible to have instant translation within 10 years. The only way to find out is to wait and see. I'll give you a dollar if I'm wrong in 2023.
ha-ha on myself! I read "With DARPA and Google racing to perfect instant translation" as "PERfect" meaning 100% perfect quality, and not "perFECT" as in to smooth out the details.
Well, there goes my PERfect understanding skills.
That's not a bet I want to take. We have "instant translation" now - I've seen demos. It's just rather crappy. What level do you expect that we'll have in 10 years?
My challenge would be to take chapter 12 of Gödel, Escher, Bach and translate it into Swahili then ask the Swahili speaker if it seems stilted or normal. Well, compared to what a human translator would do.
Chapter 12 contains a mixture of English, French, German, and transliterated Russian, but is perfectly understandable to someone who doesn't know the non-English language, because the text describes how to understand the missing parts. It also uses "ASU" as an inversion of "USA", where "ASU" is defined in the text as "Alternative Structure of the Union." Will the translation into Swahili handle that in-text complexity?
I don't think 90% of the population can even understand Godel, Escher, Bach, let alone translate it in any meaningful way. You're asking too much from a computer. Perhaps in 30 years that can happen. I think computers will be able to translate very accurately at a middle-school level within 10 years.
This wasn't a hard chapter of GEB. :) Okay, what about Asterix and Obelisk?
Still, can you quantify what "very accurately" and "middle-school level" means? Do you think in 10 years that machine translation will be able to translate the following two (independent) paragraphs correctly?
Hikaru liked the old English teacher more than the new Scottish one. He still found it hard to understand the replacement's strange dialect as she droned away about the Battle of Agincourt, the Treaties of Aix-la-Chapelle, or how Paris 'deserves a mass', while the boyish Londoner had made far-away European history seem exciting.
Hikaru liked the old English teacher most of all. He moved to London only a year ago, and Mrs. Bane had helped him fall in love with the language. Her 40 years of service had not stunted her enthusiasm one bit, as she taught him and her other charges about the subtleties in Shakespeare, the humor in Carroll, and the epic power of Morrison.
In the first, "old English teacher" refers to the previous history teacher, who was from London. In the second, it refers to the teacher of English, who is at least 60 years old.
In some Germanic languages, the first is written as something akin to "english teacher" and the second "englishteacher." To translate it correctly, the translator has to read ahead to disambiguate the meaning. Which isn't hard, even for middle-school students. But it is hard for computers.
In other languages, gender is more important. In the first paragraph, it's very likely that the "old English teacher" is a man, and the new Scottish one is a woman. In the second paragraph, the "old English teacher" is a woman.
Is your claim that an automatic translation system will be able to translate these correctly to (say) Swedish or French, within 10 years? I don't see it happening. To do so requires some quite good AI, while much of the progress over the last 10 years has been by rejecting traditional AI and collecting large statistical samples.
BTW, the current Swedish translations from Google translate for these are "Hikaru gillade gamla engelsklärare mer än den nya skotska ett." and "Hikaru gillade gamla engelsklärare mest av allt." Both are wrong. The first should be more like "Hikaru gillade den förra engelsk lärare mer än den nya skotska." and the second "Hikaru gillade den gamla engelsklärare mest av allt.
What about a common "intermediary" language? Instead of writing n^2 translation algorithms, you'd only need 2n: one to translate into the intermediary, and one to translate out. Such a language would have to be incredibly flexible though; my vote is for lojban ;). Done correctly, such a system could also teach the intermediary language to the speaker during translation. Eventually people would start to forgo the translation process and simply speak the common tongue.
I am fairly ignorant on languages. Does Chinese keep up with the fast pace of innovation these days by creating new words for new inventions from the English speaking world?
One of the reasons I think English does make a good exchange language is because it generally has a word for anything that comes along. Also it is the easiest to deal with in terms of building software and existing natural language processing research.
From the article: "People will be fluent in every language. With DARPA and Google racing to perfect instant translation, it won't be long until your cellphone speaks Swahili on your behalf." That was suggested as a plausible prediction for 2022. The comment to which I reply says,
Given the current state of language translating software, I think this is highly unlikely in the next 10 years. Anyone who is bilingual and has used a translation app knows how woefully inadequate it is for actual communication.
The last time a video demonstration of software in development for language translation was posted here on HN,
To someone who spent years learning Chinese as a second language, and then made my living for years as a Chinese-English interpreter, that was pretty impressive.
The economics of the issue is that a machine interpreter just has to be as good as a human interpreter at the same cost. That's a reachable target with today's computer technology. EVERY time I've heard someone else interpreting English or Chinese into the other language, I have heard mistakes, and I am chagrined to remember mistakes that I made over the years. We can't count on error-free machine interpretation between any pair of languages (human language is too ambiguous in many daily life cases for that), but if companies develop tested, validated software solutions for consecutive interpreting (what I usually did, and what is shown in the video) or simultaneous interpreting (the harder kind of interpreting in demand at the United Nations, where even in the best case it is not always done well), then those companies will be able to displace a lot of human professionals who rely on their language ability to make a living.
And relating that comment to a timeline appropriate for today's thread, I think ten years out is about right for a mobile app widely available at reasonable cost that will displace human interpreters for most use cases that now require human interpreters. As I further commented on the the earlier thread 37 days ago,
Right now a lot of interpreters in the United States make a lot of part-time income from gigs that involve suddenly getting telephone calls and joining in to interpret a telephone conversation in two languages. This is often necessary, for example, for physician interviews of patients in emergency rooms or pharmacist consultations with patients buying prescribed drugs (where I last saw a posted notice on how to access such an interpretation service). The IBM Watson project is already targeted at becoming an expert system for medical diagnosis, and patient care markets will surely provide a lot of income for further development of software interpretation between human languages.
It's still good for human beings to spend the time and effort to learn another human language (as so many HN participants have by learning English as a second language). That's a broadening experience and an intellectual delight. But just as riding horses is more a form of recreation these days than a basis for being employed, so too speaking another language will be a declining factor in seeking employment in the next decade.
If there is economic demand for interpreting between a given language pair, there will be great economic incentive to develop software-as-a-service to do interpreting for that language pair. I would expect Japanese-English and Spanish-English and Chinese-English interpreting all to decline as a human paying occupation by ten years from now, even though international communication will surely increase in the same time frame.
I agree that software will soon be able to efficiently interpret between two languages. I take issue with the article stating that "people will be fluent in every language". Having access to an interpreter (software or human) does not make one fluent in a language. Actually being fluent provides a much greater depth of understanding than an interpreter can ever approach. An English-speaking businessman who lives in Japan and relies on an interpreter will not be as efficient at his job as one who becomes fluent in Japanese, due to the enormous subtleties and cultural complexities of language.
Sure, people with a pocket translator will be "fluent" in every language just like people who don't know the difference between "divide by one-half" and "divide in half" but have a calculator app on their phone are now brilliant at math.
These are always fun, and always wrong but of course we can only see what we can see as my grandfather would say.
I suspect that once we can inject data directly into the brain's visual and auditory systems things like telepresence will take on new life. I'd predict that at some point in the future people get cochlear type implants rather than headphones.
I also wonder about the increasing violence of the weather cycles and whether or not that will result in new ways of building cities. Do we start building towns in the midwest that just by staying indoors you can avoid any danger from a tornado? Or housing systems that are immune to floods by either floating up or simply being sealed and ready to be temporarily submerged.
Will communities build power banks? Giant power storage facilities to hold excess renewable power that was generated? Or will that storage be distributed amongst the buildings?
It's a disappointing list for such a grandiose title. Nearly every single thing listed will be achievable in the next 20-30 years. It's reminiscent of the French predictions of life 100 years later, 100 years ago. http://singularityhub.com/2012/10/15/19th-century-french-art...
I would personally like to see some people come up with zany ideas. It's really hard to try and think that far ahead, given how much has changed in the last 100 years.
Computation should be everywhere, along with intelligent "software". Quotes because those entities might not like the phrase "software". I wouldn't.
Beyond that, it's hard to guess. Most predictions don't take culture into account. Technology and culture change and shape each other. Culture may be radically different in 100 years. Kind of like Egyptian culture can be demarcated before and after the rise of Islaam. Longer than 100 years, but gets across the idea.
My gut predictions are about what wont happen. Civilization won't collapse. Global warming will have some change but won't kill all seaside cities. This is because either through market forces (likely) or unified global political action (unlikely), we fix it.
Wars: check. Pandemic plague: no. Space stuff: check.
I general, try not to say what won't happen. Be like Star Trek in the sixties. Propose what seems futuristic to inspire people, and they'll figure out how to do it.
A positive (logically, not ethically) prediction from my side:
In 100 years, the earth will have a smaller population than now (6 Billion), concentrated in clusters of smallish cities (as opposed to mega-cities). Nature will be reconquering much of the world that has been devastated by climate change and the desperate attempts to grow food by starving populations.
Society will be heavily stratified, with a perpetually unemployed lower class living of the state with minimal rights and opportunities. Class distinctions are amplified by biotechnological advances accessible only to those with money and power. Surveillance will be ubiquitous and anonymity will be non-existant for all but the most technologically savvy. Most countries will be governed by some pseudo-democratic system.
Some descendant of humanity, be it genetically enhanced men or our electro-mechanical children, will have colonized asteroids. Moreover, they will be sending out extrasolar expeditions following the probes they have sent out 50 years earlier.
I must be in a bad mood today. Also, I didn't follow your advice.
I think things will be a bit similar to the setting of the 2003 film Code 46 w/ Tim Robbins.
"It is the not-too-distant future. The population is divided between those who live "inside", in high-density cities physically separated from "outside", where a poor underclass live. Access to and travel between the cities is highly restricted, and regulated through the use of health cover documents, known as "papelles" in the global pidgin language of the day (which comprises elements of English, Spanish, French, Arabic, Italian, Farsi and Mandarin).
Most residents of the cities venture outside at night and remain indoors during the day, as direct sunlight has come to be thought of as hazardous to their health. However, many residents inside and outside the cities still work and venture outdoors during the daytime hours (as seen on multiple occasions throughout the film). The form of government appears to be authoritarian in nature and at least somewhat dystopian. Society is regulated by various "codes". The eponymous code of the movie title prohibits "genetically incestuous reproduction", which may occur as a result of the various medical technologies which have become commonplace, such as cloning."
Allow me to make my list of broad predictions that doesn't have to do with gadgets. I expect near-term singularity so my list is pre-singularity, 2040ish going by Ray Kurzweil, could be earlier or later.
1. Massive Population growth, 50-100 billion people on the planet. Not due to natural population growth (which will collapse), but due to cloning and factory production of humans. It seems unimaginable in today's US political climate but there are other countries in the world which do not have the same moral qualms.
2. Incredibly improved education / work assignment systems. Today we go through a K-12 system + many years of tertiary education in order to gain a general education that's only somewhat suited for the tasks we're going to do. The advent of internet based education means we're going to be able to closely target the training with the job in an incredibly efficient way. This does not mean that people will lose general knowledge or become stupider. It means competence and work satisfaction will increase, and people's ability to hop careers will increase, leading to more diversity in people's lives.
3. Biological immortality or close to it. The same technology that enables cloning will enable replacement body parts for every part of the body. Diseases like Parking-sons will be cured.
4. AI everywhere. Today weak, specific, non-generalized AI run things like warehouses and farms. Tomorrow they'll run transportation, buildings, restaurants and wal-mart, datacenters, factories that build widgets and factories that build factories, write and grade test questions, predict what we'll do and help us do them before we need to ala Google Now. In short AI will be ubiquitous and very very useful if not quite generalized.
5. Huge risk-taking. People are going to experiment, try new and dangerous things, in incredibly exciting ways, if only because we have a huge amount of people.
6. Humongous innovation and economic growth. I'm talking 50-100% annual growth. Today US (and much of the world's) economic growth is so low because of our slowness to embrace new technology and rejection of important technology. This will change, if only because of global competition.
7. Most people will be volunteers in the work of building things, and that work will be done exclusively in virtual realities. Most products will be tested out and deployed first in virtual worlds before being printed into the physical world.
8. Most people will be thought of as "unemployed" by today's standards but they'll feel that they are gainfully employed and deserve oodles of money (and they'll be right).
9. The currency of choice will not be the USD (nor the Yuan nor Euro).
10. I don't know if Mars will be terraformed yet but if it is or we have huge colonies then the population estimate could be at 2x what I said in #1.
What reason/s do you have to believe that "50 - 100 billion people" will be cloned and factory produced? Who would create these humans, and what purpose would they serve, that could not be better served by factory-produced robots?
Also, why do you think (and I'll paraphrase here) "most people will deserve oodles of money"?
I'm (sadly) pessimistic about advances in superlongevity. Even though modern medicine has greatly increased average life expectancy and the number of centenarians, maximum age has not increased. The number of centenarians over 110 has increased ~10x over the past 50 years, but the number older than 114 is pretty flat. There seems to be some biological "brick wall" around age 114.
I talked to a development economist once (well more than once, but the other conversations are not germane) and he claimed that if you ignore infant mortality decreases, populations in Africa (and other poor places) have approximately the same average life expectancy as we do. Kinda summed up modern medicine for me.
The highly expensive care (that is only available in developed countries) only adds a few months, in general. That's why the health cost graphs of most people show a spike in the last year of life. That doesn't mean it shouldn't be provided. That's a really tough call to make. I am saying that it won't have a huge impact on life expectancy.
All that said, people in developing countries "just deal" with a lot of seriously unpleasant (and economically costly) health problems that we'd fix in the U.S. So there are quality-of-life benefits that don't show up in the raw numbers.
Finally, a 50% reduction in across-the-board death rates only buys eight years in life expectancy, because per-year death rates go up exponentially. Another way to look at this is that "only" one year in life expectancy difference actually represents being 9% better at keeping people alive.
For example, life expectancy of people who didn't die violently and survived the first 2 years was about 50 in ancient times. (Including those factors, it was 20-25.) Now it's about 80. That may not seem like a world-changing increase compared to the other improvements, but it actually means that we're 12-16 times better at keeping people alive.
To the contrary, girl babies born since 2000 in the developed world are more likely than not to reach the age of 100, with boys likely to enjoy lifespans almost as long. The article "The Biodemography of Human Ageing" by James Vaupel,
originally published in the journal Nature in 2010, is a good current reference on the subject. A comparison of period life expectancy tables and cohort life expectancy tables for men and women in Britain
helps make the picture more clear. ("Period life expectancy" is what is usually reported for a whole country. But cohort life expectancy provides a better estimate of future lifespans of young people today,
The issue with high longevity is that an across-the-board 50% reduction in the death rate (which was achieved between the 1960s and now) only buys 8 years. If it's only death-rate reduction that's going on, then you only get about 1 year back every 5, which is not bad in absolute terms but disappointing for people who want to live "forever". Additionally, I think there are economic costs associated with death-rate reduction (in lieu of actually slowing or halting aging) that make it unlikely that the 1 year/5 year trend will continue.
In terms of actually slowing down senescence, or the 9%/year increase in death rate, we haven't. We don't have the tools yet. At some point, people will. I'm 50-50 about whether I'll "make the cut" but choose to live as if I won't, for obvious reasons. I am not even guaranteed a day; no one is.
I think that the tools to slow senescence will be developed a few years before it becomes possible to halt and reverse aging outright. There won't be a period where people have 150-year lifespans, followed by 250, then 1000. The first 150-year-old is probably the first 10,000-year-old (because, with 5000 years of technological progress, most of the accidental death causes will be very rare). The world's oldest person will experience life like this: about 110 years of natural aging, followed by some experimental new treatments that keep that person in a senescent state for about 10-15 years, followed by a reversion, probably returning to the health of an average 20-year-old by age 140. This person may be alive now, or may not be born for a couple centuries. I'd bet on "alive now", though.
While I can't find a link, I do recall reading that actuaries have found the theoretical average life time of an "immortal" (that is, no death due to aging, disease, etc) is only 600 years, due to accidents, suicide and/or murder.
Nothing about genetic engineering of humans? I guess that's too unsettling? It seems that there are certain futures that are possible and interesting but people refuse to talk about because they bring up too many disturbing implications.
What about implanting a computer into your brain. An AI that's connected to the `human internet` which record your senses and which you can interact with your thoughts.
There's a trend in human history where as there are more and more of us, we become closer to each other. Where in pre-historic times, different populations only interact through migrations, in ancient times we begin to have trade routes which allow populations to affect each other in terms of years and decades rather than centuries. Development of oceanic travel reduced this to months and years. The telegraph and railway further reduced this to hours and days. Today with the mobile phone and internet, we can interact with each other in terms of seconds and minutes. Will this trend continue into the microseconds, so that in a thousand years time, humankind is a single hive mind-like entity?
"Man considered with himself, for in a way, Man, mentally, was one. He consisted of a trillion, trillion, trillion ageless bodies, each in its place, each resting quiet and incorruptible, each cared for by perfect automatons, equally incorruptible, while the minds of all the bodies freely melted one into the other, indistinguishable."
37 comments
[ 2.1 ms ] story [ 91.8 ms ] threadGiven the current state of language translating software, I think this is highly unlikely in the next 10 years. Anyone who is bilingual and has used a translation app knows how woefully inadequate it is for actual communication. The nuances of language are far too complex, not to mention the importance of non-verbal communication. How can software correctly translate the answer "Well, I'm not sure..." (which could signify hesitation to accept, passive disagreement, a pause for time to think, etc.) without understanding the context of the answer, and the tone of voice and facial expression of the speaker? Not in 10 years...
"Perfect instant translation" requires full AI. Consider the Borges story "Hombre de la esquina rosada". That translates directly to "The Man of the Pink Corner." But the "corner" here means also the corner store, so it's not just the physical corner but also the "hangout for the low-life of the barrio." (I'm quoting from the translator's notes here.) Moreover, the "pink" refers to the pastel hues popular in earlier decades, so the pink implies an older neighborhood "populated by toughs and knife fighters."
If you knew Buenos Aires of 100 years ago then you would infer that directly from the title. But without the context you just know it as "the pink corner."
I've thought that "The Hood of the Pink Corner" would be a better translation, given the double or triple meaning of "hood" to mean "hoodlum", "neighborhood", and more specifically "poor neighborhood." It's not as direct, but it might be better.
How do you explain those nuances to a computer?
Douglas Hofstader gives more examples of the difficulties in natural language to natural language translation. How does one translate a nonsense poem like "Jabberwocky"? One French translation starts "Il brilgue: les tôves lubricilleux. Se gyrent en vrillant dans le guave..."
Or, and I quote from his Gödel, Escher, Bach: Take a look at the first sentence of Dostoevsky's novel 'Crime and Punishment' in Russian and then in a few different English translations... The first sentence employs the street name "S. Pereulok" (as transliterated). What is the meaning of this? A careful reader of Dostoevsky's work who knows Leningrad ... can discover by doing some careful checking of the rest of the geography in the book ... that the street must be 'Stoliaryny Peteulok". .. should we keep the initial so as to reproduce the aura of semi-mystery which appears already in this first sentence of the book? We could get to "S. Lane", ("lane" being the standard translation of "pereulok"). None of the translators took this tack. However, one chose to write "S. Place". [The second] gave the translation as "Stoliarny Place". What about number 3? That's the most interesting of all. The translation says "Carpenter's Lane." And why not, indeed? After all, "stoliar" means "carpenter" and "ny" is an adjectival ending.
He then goes on to argue that perhaps we should just read Dickens instead, since Dickens is, at some level, a "translation" of the Dosteovsky novel.
BTW, Google Translate gives "S. Lane", which none of the three translators used. As that's the obvious and most direct translation, which the humans didn't choose, I assume that we can say it's not the perfect translation.
How would an automatic translation system, without AI, look at "С. Переулок", figure out that it's supposed to be "Столярный Переулок", and translate it correctly?
I understand the problems inherent in translation. But again, I still feel it's possible to have instant translation within 10 years. The only way to find out is to wait and see. I'll give you a dollar if I'm wrong in 2023.
Well, there goes my PERfect understanding skills.
That's not a bet I want to take. We have "instant translation" now - I've seen demos. It's just rather crappy. What level do you expect that we'll have in 10 years?
My challenge would be to take chapter 12 of Gödel, Escher, Bach and translate it into Swahili then ask the Swahili speaker if it seems stilted or normal. Well, compared to what a human translator would do.
Chapter 12 contains a mixture of English, French, German, and transliterated Russian, but is perfectly understandable to someone who doesn't know the non-English language, because the text describes how to understand the missing parts. It also uses "ASU" as an inversion of "USA", where "ASU" is defined in the text as "Alternative Structure of the Union." Will the translation into Swahili handle that in-text complexity?
Still, can you quantify what "very accurately" and "middle-school level" means? Do you think in 10 years that machine translation will be able to translate the following two (independent) paragraphs correctly?
Hikaru liked the old English teacher more than the new Scottish one. He still found it hard to understand the replacement's strange dialect as she droned away about the Battle of Agincourt, the Treaties of Aix-la-Chapelle, or how Paris 'deserves a mass', while the boyish Londoner had made far-away European history seem exciting.
Hikaru liked the old English teacher most of all. He moved to London only a year ago, and Mrs. Bane had helped him fall in love with the language. Her 40 years of service had not stunted her enthusiasm one bit, as she taught him and her other charges about the subtleties in Shakespeare, the humor in Carroll, and the epic power of Morrison.
In the first, "old English teacher" refers to the previous history teacher, who was from London. In the second, it refers to the teacher of English, who is at least 60 years old.
In some Germanic languages, the first is written as something akin to "english teacher" and the second "englishteacher." To translate it correctly, the translator has to read ahead to disambiguate the meaning. Which isn't hard, even for middle-school students. But it is hard for computers.
In other languages, gender is more important. In the first paragraph, it's very likely that the "old English teacher" is a man, and the new Scottish one is a woman. In the second paragraph, the "old English teacher" is a woman.
Is your claim that an automatic translation system will be able to translate these correctly to (say) Swedish or French, within 10 years? I don't see it happening. To do so requires some quite good AI, while much of the progress over the last 10 years has been by rejecting traditional AI and collecting large statistical samples.
BTW, the current Swedish translations from Google translate for these are "Hikaru gillade gamla engelsklärare mer än den nya skotska ett." and "Hikaru gillade gamla engelsklärare mest av allt." Both are wrong. The first should be more like "Hikaru gillade den förra engelsk lärare mer än den nya skotska." and the second "Hikaru gillade den gamla engelsklärare mest av allt.
One of the reasons I think English does make a good exchange language is because it generally has a word for anything that comes along. Also it is the easiest to deal with in terms of building software and existing natural language processing research.
Given the current state of language translating software, I think this is highly unlikely in the next 10 years. Anyone who is bilingual and has used a translation app knows how woefully inadequate it is for actual communication.
The last time a video demonstration of software in development for language translation was posted here on HN,
http://news.ycombinator.com/item?id=4759620
I commented,
To someone who spent years learning Chinese as a second language, and then made my living for years as a Chinese-English interpreter, that was pretty impressive.
The economics of the issue is that a machine interpreter just has to be as good as a human interpreter at the same cost. That's a reachable target with today's computer technology. EVERY time I've heard someone else interpreting English or Chinese into the other language, I have heard mistakes, and I am chagrined to remember mistakes that I made over the years. We can't count on error-free machine interpretation between any pair of languages (human language is too ambiguous in many daily life cases for that), but if companies develop tested, validated software solutions for consecutive interpreting (what I usually did, and what is shown in the video) or simultaneous interpreting (the harder kind of interpreting in demand at the United Nations, where even in the best case it is not always done well), then those companies will be able to displace a lot of human professionals who rely on their language ability to make a living.
And relating that comment to a timeline appropriate for today's thread, I think ten years out is about right for a mobile app widely available at reasonable cost that will displace human interpreters for most use cases that now require human interpreters. As I further commented on the the earlier thread 37 days ago,
Right now a lot of interpreters in the United States make a lot of part-time income from gigs that involve suddenly getting telephone calls and joining in to interpret a telephone conversation in two languages. This is often necessary, for example, for physician interviews of patients in emergency rooms or pharmacist consultations with patients buying prescribed drugs (where I last saw a posted notice on how to access such an interpretation service). The IBM Watson project is already targeted at becoming an expert system for medical diagnosis, and patient care markets will surely provide a lot of income for further development of software interpretation between human languages.
It's still good for human beings to spend the time and effort to learn another human language (as so many HN participants have by learning English as a second language). That's a broadening experience and an intellectual delight. But just as riding horses is more a form of recreation these days than a basis for being employed, so too speaking another language will be a declining factor in seeking employment in the next decade.
If there is economic demand for interpreting between a given language pair, there will be great economic incentive to develop software-as-a-service to do interpreting for that language pair. I would expect Japanese-English and Spanish-English and Chinese-English interpreting all to decline as a human paying occupation by ten years from now, even though international communication will surely increase in the same time frame.
I suspect that once we can inject data directly into the brain's visual and auditory systems things like telepresence will take on new life. I'd predict that at some point in the future people get cochlear type implants rather than headphones.
I also wonder about the increasing violence of the weather cycles and whether or not that will result in new ways of building cities. Do we start building towns in the midwest that just by staying indoors you can avoid any danger from a tornado? Or housing systems that are immune to floods by either floating up or simply being sealed and ready to be temporarily submerged.
Will communities build power banks? Giant power storage facilities to hold excess renewable power that was generated? Or will that storage be distributed amongst the buildings?
I would personally like to see some people come up with zany ideas. It's really hard to try and think that far ahead, given how much has changed in the last 100 years.
Computation should be everywhere, along with intelligent "software". Quotes because those entities might not like the phrase "software". I wouldn't.
Beyond that, it's hard to guess. Most predictions don't take culture into account. Technology and culture change and shape each other. Culture may be radically different in 100 years. Kind of like Egyptian culture can be demarcated before and after the rise of Islaam. Longer than 100 years, but gets across the idea.
My gut predictions are about what wont happen. Civilization won't collapse. Global warming will have some change but won't kill all seaside cities. This is because either through market forces (likely) or unified global political action (unlikely), we fix it.
Wars: check. Pandemic plague: no. Space stuff: check.
I general, try not to say what won't happen. Be like Star Trek in the sixties. Propose what seems futuristic to inspire people, and they'll figure out how to do it.
A positive (logically, not ethically) prediction from my side:
In 100 years, the earth will have a smaller population than now (6 Billion), concentrated in clusters of smallish cities (as opposed to mega-cities). Nature will be reconquering much of the world that has been devastated by climate change and the desperate attempts to grow food by starving populations.
Society will be heavily stratified, with a perpetually unemployed lower class living of the state with minimal rights and opportunities. Class distinctions are amplified by biotechnological advances accessible only to those with money and power. Surveillance will be ubiquitous and anonymity will be non-existant for all but the most technologically savvy. Most countries will be governed by some pseudo-democratic system.
Some descendant of humanity, be it genetically enhanced men or our electro-mechanical children, will have colonized asteroids. Moreover, they will be sending out extrasolar expeditions following the probes they have sent out 50 years earlier.
I must be in a bad mood today. Also, I didn't follow your advice.
"It is the not-too-distant future. The population is divided between those who live "inside", in high-density cities physically separated from "outside", where a poor underclass live. Access to and travel between the cities is highly restricted, and regulated through the use of health cover documents, known as "papelles" in the global pidgin language of the day (which comprises elements of English, Spanish, French, Arabic, Italian, Farsi and Mandarin).
Most residents of the cities venture outside at night and remain indoors during the day, as direct sunlight has come to be thought of as hazardous to their health. However, many residents inside and outside the cities still work and venture outdoors during the daytime hours (as seen on multiple occasions throughout the film). The form of government appears to be authoritarian in nature and at least somewhat dystopian. Society is regulated by various "codes". The eponymous code of the movie title prohibits "genetically incestuous reproduction", which may occur as a result of the various medical technologies which have become commonplace, such as cloning."
1. Massive Population growth, 50-100 billion people on the planet. Not due to natural population growth (which will collapse), but due to cloning and factory production of humans. It seems unimaginable in today's US political climate but there are other countries in the world which do not have the same moral qualms.
2. Incredibly improved education / work assignment systems. Today we go through a K-12 system + many years of tertiary education in order to gain a general education that's only somewhat suited for the tasks we're going to do. The advent of internet based education means we're going to be able to closely target the training with the job in an incredibly efficient way. This does not mean that people will lose general knowledge or become stupider. It means competence and work satisfaction will increase, and people's ability to hop careers will increase, leading to more diversity in people's lives.
3. Biological immortality or close to it. The same technology that enables cloning will enable replacement body parts for every part of the body. Diseases like Parking-sons will be cured.
4. AI everywhere. Today weak, specific, non-generalized AI run things like warehouses and farms. Tomorrow they'll run transportation, buildings, restaurants and wal-mart, datacenters, factories that build widgets and factories that build factories, write and grade test questions, predict what we'll do and help us do them before we need to ala Google Now. In short AI will be ubiquitous and very very useful if not quite generalized.
5. Huge risk-taking. People are going to experiment, try new and dangerous things, in incredibly exciting ways, if only because we have a huge amount of people.
6. Humongous innovation and economic growth. I'm talking 50-100% annual growth. Today US (and much of the world's) economic growth is so low because of our slowness to embrace new technology and rejection of important technology. This will change, if only because of global competition.
7. Most people will be volunteers in the work of building things, and that work will be done exclusively in virtual realities. Most products will be tested out and deployed first in virtual worlds before being printed into the physical world.
8. Most people will be thought of as "unemployed" by today's standards but they'll feel that they are gainfully employed and deserve oodles of money (and they'll be right).
9. The currency of choice will not be the USD (nor the Yuan nor Euro).
10. I don't know if Mars will be terraformed yet but if it is or we have huge colonies then the population estimate could be at 2x what I said in #1.
Also, why do you think (and I'll paraphrase here) "most people will deserve oodles of money"?
The following site has some interesting charts about centenarians and mortality: http://www.grg.org/calment.html
All that said, people in developing countries "just deal" with a lot of seriously unpleasant (and economically costly) health problems that we'd fix in the U.S. So there are quality-of-life benefits that don't show up in the raw numbers.
Finally, a 50% reduction in across-the-board death rates only buys eight years in life expectancy, because per-year death rates go up exponentially. Another way to look at this is that "only" one year in life expectancy difference actually represents being 9% better at keeping people alive.
For example, life expectancy of people who didn't die violently and survived the first 2 years was about 50 in ancient times. (Including those factors, it was 20-25.) Now it's about 80. That may not seem like a world-changing increase compared to the other improvements, but it actually means that we're 12-16 times better at keeping people alive.
http://www.demographic-challenge.com/files/downloads/2eb51e2...
originally published in the journal Nature in 2010, is a good current reference on the subject. A comparison of period life expectancy tables and cohort life expectancy tables for men and women in Britain
http://www.ons.gov.uk/ons/rel/lifetables/period-and-cohort-l...
helps make the picture more clear. ("Period life expectancy" is what is usually reported for a whole country. But cohort life expectancy provides a better estimate of future lifespans of young people today,
http://www.time.com/time/specials/packages/article/0,28804,1...
and is still steadily on the rise around the world.)
Life expectancy at age 40, at age 60, and at even higher ages is still rising throughout the developed countries of the world.
http://www.scientificamerican.com/article.cfm?id=longevity-w...
In terms of actually slowing down senescence, or the 9%/year increase in death rate, we haven't. We don't have the tools yet. At some point, people will. I'm 50-50 about whether I'll "make the cut" but choose to live as if I won't, for obvious reasons. I am not even guaranteed a day; no one is.
I think that the tools to slow senescence will be developed a few years before it becomes possible to halt and reverse aging outright. There won't be a period where people have 150-year lifespans, followed by 250, then 1000. The first 150-year-old is probably the first 10,000-year-old (because, with 5000 years of technological progress, most of the accidental death causes will be very rare). The world's oldest person will experience life like this: about 110 years of natural aging, followed by some experimental new treatments that keep that person in a senescent state for about 10-15 years, followed by a reversion, probably returning to the health of an average 20-year-old by age 140. This person may be alive now, or may not be born for a couple centuries. I'd bet on "alive now", though.
There's a trend in human history where as there are more and more of us, we become closer to each other. Where in pre-historic times, different populations only interact through migrations, in ancient times we begin to have trade routes which allow populations to affect each other in terms of years and decades rather than centuries. Development of oceanic travel reduced this to months and years. The telegraph and railway further reduced this to hours and days. Today with the mobile phone and internet, we can interact with each other in terms of seconds and minutes. Will this trend continue into the microseconds, so that in a thousand years time, humankind is a single hive mind-like entity?
"Man considered with himself, for in a way, Man, mentally, was one. He consisted of a trillion, trillion, trillion ageless bodies, each in its place, each resting quiet and incorruptible, each cared for by perfect automatons, equally incorruptible, while the minds of all the bodies freely melted one into the other, indistinguishable."
http://filer.case.edu/dts8/thelastq.htm
I almost want to write my own list of predictions now.