8 ms·

The article premise is relatively banal, but the topic the article describes after, is extremely important (IMO). This bits: > Several companies defended thei

The article premise is relatively banal, but the topic the article describes after, is extremely important (IMO). This bits: > Several companies defended their use of AI, telling The Post they use language tools not to replace human writers, but to [...], or to produce content that they otherwise wouldn’t > We published a celebrity profile a month. Now we can do 10,000 a month. describe an undergoing mass-transition of (a large amount of) publishing towards a commodity. Cheap publishing has always existed, but it required somebody to actually write something, therefore, a minimum of creativity/knowledge has always been a requirement. With LLMs, any requirement is gone; publishing can be reduced to sending an input and checking the output. "Interesting" times!

JustSomeNobody 3y ago
For a while now we have had formulaic writing where people write under a successful dead (or almost dead) author's name. This is just using computers to do the same. Now anyone can write the next Borne novel.
clnq 3y ago
It’s amazing to my mind that AI can now help write textbooks on niche topics. I see that as tremendous progress for humanity. Of course, with capitalistic incentives, we will use it for spam and garbage.
- readthenotes1 3y ago
  Of course, without capitalism or the threat of capitalism, we wouldn't have to worry about this sort of thing.
  - clnq 3y ago
    What would be the incentive to pump out garbage? It’s unproductive work that’s sadly rewarded with pay, which pays the rent. Spending resources on things like this is one of the worst aspects of capitalism. Capitalism is just growing capital at any ethical and moral cost.
    - sclarisse 3y ago
      At the small scale, what are the desirable properties of a non capitalist book system? Would all reward and compensation for writing books (physically published or ebook) be dictated by a central authority? Would individual authors be assigned the topics they would be compensated for? Would they be audited for unauthorized use of AI? Would popularity of the books play a role at all, and if so by what means would authors be protected from effort-stealing? If not, why would we believe the right books would be produced and approved?
      - clnq 3y ago
        It does not have to be centrally controlled. You are talking more about planned economy than non-capitalist incentives. Perhaps you jumped to the conclusion that I am talking about the failed Soviet Union experiment when I say that capitalistic incentives harm literature and knowledge sharing? I am not referring to that, at all. Many people already write amazing works for non-capitalist incentives. Most research is done without them and they are often seen as a conflict of interest in producing good quality papers. A lot of textbooks are also written from need by teachers and professors rather than from capitalistic incentives. Many people just want to share their knowledge, this has been the case since before printed press. The main desirable property in a non-capitalistic economy of literature would be that literature would not be seen as means of growing capital. We have had such an approach to literature for a long time in history — a time with remarkably fewer garbage writings than today.
        sclarisse 3y ago
        Okay. That sounds reasonable, with the difficulty of being radically underfunded (especially while locked in competition with the other world.)
      - imtringued 3y ago
        The mechanism is quite simple. Humans get paid enough to not need to write bad quality books in the name of "economic growth" to begin with. A lot of people want free time so they can do uncompensated work that they like doing.
- ska 3y ago
  So far the sweet spot of LLM's seems to be quickly producing mediocre to bad content with factual errors. Assuming you have a talented & knowledgeable person in the loop to improve and fix it, I could see it speeding up the process. But the incentives are already pretty broken here, and I don't know if people will want to do it. The incentives are a bit easier to see if you just churn out bad content across a bunch of areas you don't understand, which is what we are more likely to get, no?
  - clnq 3y ago
    The incentives are also there for people who produce quality content. I write a blog on niche tech topics and rely on GPT-4 to edit it, fact-check it, grammar-check it, find weaknesses, offer alternative views and so on. It used to take me 16 hours to write an article, now it takes 4. I ask peers to review my articles, and the number of inaccuracies that would get flagged went down from a couple to nearly zero per article. The same applies to textbooks. LLMs are a fantastic tool that can be used for a lot of good. If people use it for spam, that’s a use case problem, not an AI problem. And money is definitely an incentive to abuse this tool. As you say, it’s very easy to imagine how massive quantities of spam earning small revenues per ad view, click, or book purchase can become a sustainable business. Not sure why people don’t see the capitalistic incentives playing a big role in the abuse. They definitely do.
    - ska 3y ago
      I find it interesting . I know a lot of people who have written textbooks, but approximately none of them have made reasonable money at it. So if you could halve the time to produce a good one it would help, even better if you quarter it. But you are still probably losing money on your time relative to other things you could do. So why do it ? Mostly because it (the book) doesn’t exist. If 3 crap versions exist , I don’t know if you would bother .
    - CurleighBraces 3y ago
      Do you genuinely find it good enough to fact check? My experience so far is it will lie horribly, a lot. I still find it useful enough that I subscribe to the pro version, but I would never rely on it to be able to tell the whole truth and nothing but the truth.
- ar_lan 3y ago
  The optimist (which is really my real stance) in me says "AI was always bound to happen, and why shouldn't it? It literally can make so many mundane parts of our lives easier, to let us continue to focus on greater things." Unfortunately, all technology is exploitable - and therefore will be exploited.
- mensetmanusman 3y ago
  I think this can be solved by human organizations tasked with compiling a large range of human content before it gets muddied by low quality LLMs. It would be a good job for cathedral-builder like humans who work on multi-century time scales (like the monks who saved western civilization during the “dark” ages by spending their lives slowly copying and curating texts).
ctvo 3y ago
There's space for a verification system similar to how some countries or regions closely guard their "Made in X" brands. I'll pay extra for a guarantee (somehow, this is your value proposition) that it was written by a human without the help of generative AI.
- euroderf 3y ago
  Or at the very least, fact-checked by a human.
  - ta1243 3y ago
    How would "fact checking" work. Can you establish a train of trust that all such sources were human all the way down, let alone accurate?
    - euroderf 3y ago
      That is a technical challenge for the 21st century. Much more difficult than establishing the provenance of physical goods. But short of that, I was replying to "I'll pay extra for a guarantee (somehow, this is your value proposition) that it was written by a human without the help of generative AI." So yes, I guess that is orthogonal to fact-checking.
    - Kerrick 3y ago
      The Society of Professional Journalists and other news organizations have set a standard that should be the absolute minimum. For example: - https://www.journaliststoolbox.org/2023/05/12/urban_legendsfact-checking/ https://www.journaliststoolbox.org/2023/05/12/urban_legendsf... - https://verificationhandbook.com/ https://verificationhandbook.com/ - https://mediashift.org/2015/02/journalism-professors-should-teach-accuracy-checklists/ https://mediashift.org/2015/02/journalism-professors-should-...
      - medstrom 3y ago
        From the third link: > ... the checklist is the most effective system of preventing errors, so effective that pilots and surgeons use checklists routinely when they fly and operate. (When I had a recent biopsy, I noticed the surgeon and nurses clearly following checklists as they confirmed the site for the procedure before anesthetizing me, checked my identification bracelet and asked my name and date of birth.) Suddenly I'm picturing a new world in which we knowledge workers all have our own personal wikis and carefully go through a checklist for all new info we add to it. Not only "is this AI?" but a general sanity check too.
senko 3y ago
> Cheap publishing has always existed, but it required somebody to actually write something, therefore, a minimum of creativity/knowledge has always been a requirement. Really cheap publishing is churned out in such poor quality that I don't see anyone posessing any creativity or knowledge to speak of wanting to do such mind numbing job. The tabloid articles I sometimes have the misfortune to stumble upon are so devoid of any spark of either, that I genuinely believe that using an AI would improve the quality of it.
- colecut 3y ago
  tabloid articles are shit. how do you improve the quality of that
  - squarefoot 3y ago
    No way to do that, shit will always be shit, but they can alter the perception of those articles by creating flocks of fake virtual personalities that praise them at command, then return to their daily chit chat to gain subscribers to be used at the next campaign. I feel that will soon become daily routine in politics and product advertising.
  - visarga 3y ago
    RLHF your AI until you like the output?
throwuwu 3y ago
Clearly you’ve never heard of Chuck Tingle.
ResearchCode 3y ago
They could already translate those articles in 100 languages with machine translation. Those articles were bad and not widely read.
visarga 3y ago
> publishing can be reduced to sending an input and checking the output Checking the output takes a considerable amount of effort if you want it to be factual. Don't trivialise it, it's what is left for humans to do when AI gets to work. For example coding is not much faster with chatGPT than manually, most of the time is spend debugging anyway, not writing code.
Findeton 3y ago
But, I'm sure the people who wrote the book mentioned in the article can rest easy: creativity, knowledge, and originality are still valuable, and LLMs are not (yet) able to reproduce those.
- visarga 3y ago
  I think the source of creativity and knowledge is the world. Without access to the world we can't do science. LLMs are like brains in a jar, but when they get tools they can integrate external signals and actually become creative and even discover knowledge - like [1] and AlphaGo's move 37 (it has access to the go table, which is the "world" of Go) But this "access to the world" needs massive scale, we are billions of humans all experiencing the world, AI needs probably a similar number of agents doing massive trials and search. AlphaGo sure did need lots of self-play games, not just a couple. AlphaTensor learned to improve matrix multiplication by the same method. Biological evolution produced us the same way. It's an open-ended exploration problem. A closed system can't do it, and the scale needed makes it expensive. [1] - Evolution through Large Models - https://arxiv.org/abs/2206.08896 https://arxiv.org/abs/2206.08896
SoftTalker 3y ago
Just as email caused an explosion in low-value business communication. Before email there were memos, which required a person to dictate the memo to a secretary (usually), then the secretary type it up, then the person to proofread/correct it, then the secretary to retype the final memo, send it to duplicating, and then send the duplicates to the mailroom for distribution to recipients. Now because email makes all this effortless, we spend half our day (if not more) reading and responding to email, whereas before we might get a few memos a week at most. AI will hopefully drown in its own output, as there will be too much of it for humans to handle.
- mym1990 3y ago
  Email also caused an explosion in medium and high value communication, some in business and some in personal life. Pre-texting I had a great time keeping in touch with friends around the country. "which required a person to dictate the memo to a secretary (usually), then the secretary type it up, then the person to proofread/correct it, then the secretary to retype the final memo, send it to duplicating, and then send the duplicates to the mailroom for distribution to recipients." ...how is this an optimal process? I mean should we go back to riding horses around town as well, since cars probably increase the amount of low-value travel?
  - SoftTalker 3y ago
    Didn't say that the old process was optimal but the cost of it meant that there were built-in natural limits. As the GP said, removing the time and effort required often results in production of a lot of worthless output. And yes cars have undoubtedly increased the amount of low-value travel. I don't think we need to go back to horses but we might want to better price in the externalites.
andrewjl 3y ago
Makes me wonder who is going to be reading all this, though I guess costs mean that producing really long-tail (in terms of readership numbers) publications becomes profitable?
- robotresearcher 3y ago
  The next gen AIs will be reading all this in their training data, which is a very weird situation.
- tuatoru 3y ago
  Is someone a celebrity if only two people ever read the book about them? Can anyone be a (traditional) celebrity in all this noise? Somehow I don't think this is what was meant by the "15 minutes of fame" quote misattributed to Andy Warhol. But it's what we're going to get.

JustSomeNobody 3y ago

For a while now we have had formulaic writing where people write under a successful dead (or almost dead) author's name. This is just using computers to do the same. Now anyone can write the next Borne novel.

clnq 3y ago

It’s amazing to my mind that AI can now help write textbooks on niche topics. I see that as tremendous progress for humanity. Of course, with capitalistic incentives, we will use it for spam and garbage.

readthenotes1 3y ago
Of course, without capitalism or the threat of capitalism, we wouldn't have to worry about this sort of thing.
- clnq 3y ago
  What would be the incentive to pump out garbage? It’s unproductive work that’s sadly rewarded with pay, which pays the rent. Spending resources on things like this is one of the worst aspects of capitalism. Capitalism is just growing capital at any ethical and moral cost.
  - sclarisse 3y ago
    At the small scale, what are the desirable properties of a non capitalist book system? Would all reward and compensation for writing books (physically published or ebook) be dictated by a central authority? Would individual authors be assigned the topics they would be compensated for? Would they be audited for unauthorized use of AI? Would popularity of the books play a role at all, and if so by what means would authors be protected from effort-stealing? If not, why would we believe the right books would be produced and approved?
    - clnq 3y ago
      It does not have to be centrally controlled. You are talking more about planned economy than non-capitalist incentives. Perhaps you jumped to the conclusion that I am talking about the failed Soviet Union experiment when I say that capitalistic incentives harm literature and knowledge sharing? I am not referring to that, at all. Many people already write amazing works for non-capitalist incentives. Most research is done without them and they are often seen as a conflict of interest in producing good quality papers. A lot of textbooks are also written from need by teachers and professors rather than from capitalistic incentives. Many people just want to share their knowledge, this has been the case since before printed press. The main desirable property in a non-capitalistic economy of literature would be that literature would not be seen as means of growing capital. We have had such an approach to literature for a long time in history — a time with remarkably fewer garbage writings than today.
      - sclarisse 3y ago
        Okay. That sounds reasonable, with the difficulty of being radically underfunded (especially while locked in competition with the other world.)
    - imtringued 3y ago
      The mechanism is quite simple. Humans get paid enough to not need to write bad quality books in the name of "economic growth" to begin with. A lot of people want free time so they can do uncompensated work that they like doing.
ska 3y ago
So far the sweet spot of LLM's seems to be quickly producing mediocre to bad content with factual errors. Assuming you have a talented & knowledgeable person in the loop to improve and fix it, I could see it speeding up the process. But the incentives are already pretty broken here, and I don't know if people will want to do it. The incentives are a bit easier to see if you just churn out bad content across a bunch of areas you don't understand, which is what we are more likely to get, no?
- clnq 3y ago
  The incentives are also there for people who produce quality content. I write a blog on niche tech topics and rely on GPT-4 to edit it, fact-check it, grammar-check it, find weaknesses, offer alternative views and so on. It used to take me 16 hours to write an article, now it takes 4. I ask peers to review my articles, and the number of inaccuracies that would get flagged went down from a couple to nearly zero per article. The same applies to textbooks. LLMs are a fantastic tool that can be used for a lot of good. If people use it for spam, that’s a use case problem, not an AI problem. And money is definitely an incentive to abuse this tool. As you say, it’s very easy to imagine how massive quantities of spam earning small revenues per ad view, click, or book purchase can become a sustainable business. Not sure why people don’t see the capitalistic incentives playing a big role in the abuse. They definitely do.
  - ska 3y ago
    I find it interesting . I know a lot of people who have written textbooks, but approximately none of them have made reasonable money at it. So if you could halve the time to produce a good one it would help, even better if you quarter it. But you are still probably losing money on your time relative to other things you could do. So why do it ? Mostly because it (the book) doesn’t exist. If 3 crap versions exist , I don’t know if you would bother .
  - CurleighBraces 3y ago
    Do you genuinely find it good enough to fact check? My experience so far is it will lie horribly, a lot. I still find it useful enough that I subscribe to the pro version, but I would never rely on it to be able to tell the whole truth and nothing but the truth.
ar_lan 3y ago
The optimist (which is really my real stance) in me says "AI was always bound to happen, and why shouldn't it? It literally can make so many mundane parts of our lives easier, to let us continue to focus on greater things." Unfortunately, all technology is exploitable - and therefore will be exploited.
mensetmanusman 3y ago
I think this can be solved by human organizations tasked with compiling a large range of human content before it gets muddied by low quality LLMs. It would be a good job for cathedral-builder like humans who work on multi-century time scales (like the monks who saved western civilization during the “dark” ages by spending their lives slowly copying and curating texts).

ctvo 3y ago

There's space for a verification system similar to how some countries or regions closely guard their "Made in X" brands. I'll pay extra for a guarantee (somehow, this is your value proposition) that it was written by a human without the help of generative AI.

euroderf 3y ago
Or at the very least, fact-checked by a human.
- ta1243 3y ago
  How would "fact checking" work. Can you establish a train of trust that all such sources were human all the way down, let alone accurate?
  - euroderf 3y ago
    That is a technical challenge for the 21st century. Much more difficult than establishing the provenance of physical goods. But short of that, I was replying to "I'll pay extra for a guarantee (somehow, this is your value proposition) that it was written by a human without the help of generative AI." So yes, I guess that is orthogonal to fact-checking.
  - Kerrick 3y ago
    The Society of Professional Journalists and other news organizations have set a standard that should be the absolute minimum. For example: - https://www.journaliststoolbox.org/2023/05/12/urban_legendsfact-checking/ https://www.journaliststoolbox.org/2023/05/12/urban_legendsf... - https://verificationhandbook.com/ https://verificationhandbook.com/ - https://mediashift.org/2015/02/journalism-professors-should-teach-accuracy-checklists/ https://mediashift.org/2015/02/journalism-professors-should-...
    - medstrom 3y ago
      From the third link: > ... the checklist is the most effective system of preventing errors, so effective that pilots and surgeons use checklists routinely when they fly and operate. (When I had a recent biopsy, I noticed the surgeon and nurses clearly following checklists as they confirmed the site for the procedure before anesthetizing me, checked my identification bracelet and asked my name and date of birth.) Suddenly I'm picturing a new world in which we knowledge workers all have our own personal wikis and carefully go through a checklist for all new info we add to it. Not only "is this AI?" but a general sanity check too.

senko 3y ago

> Cheap publishing has always existed, but it required somebody to actually write something, therefore, a minimum of creativity/knowledge has always been a requirement. Really cheap publishing is churned out in such poor quality that I don't see anyone posessing any creativity or knowledge to speak of wanting to do such mind numbing job. The tabloid articles I sometimes have the misfortune to stumble upon are so devoid of any spark of either, that I genuinely believe that using an AI would improve the quality of it.

colecut 3y ago
tabloid articles are shit. how do you improve the quality of that
- squarefoot 3y ago
  No way to do that, shit will always be shit, but they can alter the perception of those articles by creating flocks of fake virtual personalities that praise them at command, then return to their daily chit chat to gain subscribers to be used at the next campaign. I feel that will soon become daily routine in politics and product advertising.
- visarga 3y ago
  RLHF your AI until you like the output?

throwuwu 3y ago

Clearly you’ve never heard of Chuck Tingle.

ResearchCode 3y ago

They could already translate those articles in 100 languages with machine translation. Those articles were bad and not widely read.

visarga 3y ago

> publishing can be reduced to sending an input and checking the output Checking the output takes a considerable amount of effort if you want it to be factual. Don't trivialise it, it's what is left for humans to do when AI gets to work. For example coding is not much faster with chatGPT than manually, most of the time is spend debugging anyway, not writing code.

Findeton 3y ago

But, I'm sure the people who wrote the book mentioned in the article can rest easy: creativity, knowledge, and originality are still valuable, and LLMs are not (yet) able to reproduce those.

visarga 3y ago
I think the source of creativity and knowledge is the world. Without access to the world we can't do science. LLMs are like brains in a jar, but when they get tools they can integrate external signals and actually become creative and even discover knowledge - like [1] and AlphaGo's move 37 (it has access to the go table, which is the "world" of Go) But this "access to the world" needs massive scale, we are billions of humans all experiencing the world, AI needs probably a similar number of agents doing massive trials and search. AlphaGo sure did need lots of self-play games, not just a couple. AlphaTensor learned to improve matrix multiplication by the same method. Biological evolution produced us the same way. It's an open-ended exploration problem. A closed system can't do it, and the scale needed makes it expensive. [1] - Evolution through Large Models - https://arxiv.org/abs/2206.08896 https://arxiv.org/abs/2206.08896

SoftTalker 3y ago

Just as email caused an explosion in low-value business communication. Before email there were memos, which required a person to dictate the memo to a secretary (usually), then the secretary type it up, then the person to proofread/correct it, then the secretary to retype the final memo, send it to duplicating, and then send the duplicates to the mailroom for distribution to recipients. Now because email makes all this effortless, we spend half our day (if not more) reading and responding to email, whereas before we might get a few memos a week at most. AI will hopefully drown in its own output, as there will be too much of it for humans to handle.

mym1990 3y ago
Email also caused an explosion in medium and high value communication, some in business and some in personal life. Pre-texting I had a great time keeping in touch with friends around the country. "which required a person to dictate the memo to a secretary (usually), then the secretary type it up, then the person to proofread/correct it, then the secretary to retype the final memo, send it to duplicating, and then send the duplicates to the mailroom for distribution to recipients." ...how is this an optimal process? I mean should we go back to riding horses around town as well, since cars probably increase the amount of low-value travel?
- SoftTalker 3y ago
  Didn't say that the old process was optimal but the cost of it meant that there were built-in natural limits. As the GP said, removing the time and effort required often results in production of a lot of worthless output. And yes cars have undoubtedly increased the amount of low-value travel. I don't think we need to go back to horses but we might want to better price in the externalites.

andrewjl 3y ago

Makes me wonder who is going to be reading all this, though I guess costs mean that producing really long-tail (in terms of readership numbers) publications becomes profitable?

robotresearcher 3y ago
The next gen AIs will be reading all this in their training data, which is a very weird situation.
tuatoru 3y ago
Is someone a celebrity if only two people ever read the book about them? Can anyone be a (traditional) celebrity in all this noise? Somehow I don't think this is what was meant by the "15 minutes of fame" quote misattributed to Andy Warhol. But it's what we're going to get.