[Technology] Paying to publish research findings and a row about the fees

[Harry Wilson's tackle]

I laughed when I saw this, sent to me by Mrs T yesterday (see pasted link below).

‘Too greedy’: mass walkout at global science journal over ‘unethical’ fees | Peer review and scientific publishing | The Guardian

‘Too greedy’: mass walkout at global science journal over ‘unethical’ fees Entire board resigns over actions of academic publisher whose profit margins outstrip even Google and Amazon www.theguardian.com

What a load of cant.

Universities and the academics who run them have created a merit system whereby tenure and promotions are determined by grant income and publication outcome. The journal impact factor (JIF), a number based on how well papers published in a journal are cited by other investigators is the key indicator. The universities and academics know JIF is meaningless (it is an average for all papers published in a journal over 2 years, and we all know that not all examples of something are equal - our game on Monday for example was hardly equal to the previous game, yet both emanated from the same organization), However universities still intimidate junior staff with threats that they must publish in journals with JIFs above ‘x’ or they won’t have a career.

So of course publishers will charge researchers to publish, based on the JIF. They would be mad to not do so. Market forces.

We live in a global capitalist economy. If I think it would help my career to join the Athaneum (rather nice food there, and great for swanky hob-nobbing) I would need to pass their entry requirements and pay the fees. If I want to publish in a well-regarded, high JIF journal then I (or my institution or funders) should pay. Why should my funders such as the MRC or BHF pay? Because they expect me to publish in these journals or they won’t renew my grants.

I’m a member of the Labour party, not a free market zealot. There is something to be said for a different way of publishing. But all the while that individuals and universities demand the kudos of high JIF publications they must pay.

The attitude of my line managers is that if I can’t raise the money to pay for my research and publications then I’m not doing my job. With certain caveats….they are right.

 

[Happy Exile]

That's all interesting and I had no idea. So if I am reading something in a journal how do I know it's the best research there is as opposed to just what someone has been prepared to pay highest to appear in? Or put another way, does this mean the most prestigious journal isn't necessarily where the best research can be found? And isn't it then a self-perpetuating doom-spiral where people are citing research that isn't necessarily the best but because it appears in the most prominent place because someone paid for it to be there?

Earlier this year I had a book published in my field - no academic rigour to it really, but had to jump through endless hoops for the publisher and demonstrate my alleged credibility and a level of professional recognition before they'd even have a conversation with me about publishing. It slightly blows my tiny mind that academic publishing is less strict.

 

[Littlemo]

I laughed when I saw this, sent to me by Mrs T yesterday (see pasted link below).

‘Too greedy’: mass walkout at global science journal over ‘unethical’ fees | Peer review and scientific publishing | The Guardian

‘Too greedy’: mass walkout at global science journal over ‘unethical’ fees Entire board resigns over actions of academic publisher whose profit margins outstrip even Google and Amazon www.theguardian.com

What a load of cant.

Universities and the academics who run them have created a merit system whereby tenure and promotions are determined by grant income and publication outcome. The journal impact factor (JIF), a number based on how well papers published in a journal are cited by other investigators is the key indicator. The universities and academics know JIF is meaningless (it is an average for all papers published in a journal over 2 years, and we all know that not all examples of something are equal - our game on Monday for example was hardly equal to the previous game, yet both emanated from the same organization), However universities still intimidate junior staff with threats that they must publish in journals with JIFs above ‘x’ or they won’t have a career.

So of course publishers will charge researchers to publish, based on the JIF. They would be mad to not do so. Market forces.

We live in a global capitalist economy. If I think it would help my career to join the Athaneum (rather nice food there, and great for swanky hob-nobbing) I would need to pass their entry requirements and pay the fees. If I want to publish in a well-regarded, high JIF journal then I (or my institution or funders) should pay. Why should my funders such as the MRC or BHF pay? Because they expect me to publish in these journals or they won’t renew my grants.

I’m a member of the Labour party, not a free market zealot. There is something to be said for a different way of publishing. But all the while that individuals and universities demand the kudos of high JIF publications they must pay.

The attitude of my line managers is that if I can’t raise the money to pay for my research and publications then I’m not doing my job. With certain caveats….they are right.

I work in this area and tbh, you are misunderstanding the nature of this. This is not about fees to publish in High JIF journals, the charges are levied against journals regardless of Impact factor and impact factor doesn’t have much to do with this tbh.

The basics are that the costs of publishing are incredible, so high that universities cannot afford to pay them for all their staff. (Usually between £1k to as high as £4K per article) which given that some subject areas publish at high levels, this runs into billions of pounds.

The funders come in because they pay for the research to be done and want the results to be available as Open Access, I.e accessible to anyone with a working computer. They do not require publication in a set journal or one with a particular JIF, they only set that it has to be made open access within (usually) a few months of publication. They do check for this as well.

The main issue with fees are that one, they already charge universities for access to journal content, thus they are “Double dipping” charging us twice for the same thing.

The second is that no one objects to paying fees, only that £3k on avg per article is too high. Nobody but the funders can afford it, and they do very little now to warrant it, apart from organising peer review (which they do not pay a wage for and is done at the time/expense of the academic reviewers) they simply put up the (formatted by the author) paper digitally (as there are fewer print journals anymore) and that’s it. They do not run the journals, they do not edit the journals, all done by academic staff at universities. They only basically host them now on their website and collect a lot of money.

 

[beorhthelm]

probably right. or wrong. bit of a fringe topic? (though i type as someone give a sensible reply )

what have the economic journals done, they must have solved this, right?

 

[Happy Exile]

I work in this area and tbh, you are misunderstanding the nature of this. This is not about fees to publish in High JIF journals, the charges are levied against journals regardless of Impact factor and impact factor doesn’t have much to do with this tbh.

The basics are that the costs of publishing are incredible, so high that universities cannot afford to pay them for all their staff. (Usually between £1k to as high as £4K per article) which given that some subject areas publish at high levels, this runs into billions of pounds.

The funders come in because they pay for the research to be done and want the results to be available as Open Access, I.e accessible to anyone with a working computer. They do not require publication in a set journal or one with a particular JIF, they only set that it has to be made open access within (usually) a few months of publication. They do check for this as well.

The main issue with fees are that one, they already charge universities for access to journal content, thus they are “Double dipping” charging us twice for the same thing.

The second is that no one objects to paying fees, only that £3k on avg per article is too high. Nobody but the funders can afford it, and they do very little now to warrant it, apart from organising peer review (which they do not pay a wage for and is done at the time/expense of the academic reviewers) they simply put up the (formatted by the author) paper digitally (as there are fewer print journals anymore) and that’s it. They do not run the journals, they do not edit the journals, all done by academic staff at universities. They only basically host them now on their website and collect a lot of money.

So I saw an academic at a conference a while ago say to contact her for access to any of her research and she'd send it free of charge, not to pay any kind of paywall fee as neither her or her institution got any of that cash and she owned it not the publications. I'm assuming that's fairly common, and certainly I've contacted a few people over the years and been sent research. How does that fit in the model, or does it further break it?

 

[Littlemo]

So I saw an academic at a conference a while ago say to contact her for access to any of her research and she'd send it free of charge, not to pay any kind of paywall fee as neither her or her institution got any of that cash and she owned it not the publications. I'm assuming that's fairly common, and certainly I've contacted a few people over the years and been sent research. How does that fit in the model, or does it further break it?

That’s usually fine, the copyright rules around the versions of articles and what you can do with them at each stage are complicated but one that is usually permitted is to allow the author to give copies to people for their own use.

 

[Harry Wilson's tackle]

I work in this area and tbh, you are misunderstanding the nature of this. This is not about fees to publish in High JIF journals, the charges are levied against journals regardless of Impact factor and impact factor doesn’t have much to do with this tbh.

The basics are that the costs of publishing are incredible, so high that universities cannot afford to pay them for all their staff. (Usually between £1k to as high as £4K per article) which given that some subject areas publish at high levels, this runs into billions of pounds.

The funders come in because they pay for the research to be done and want the results to be available as Open Access, I.e accessible to anyone with a working computer. They do not require publication in a set journal or one with a particular JIF, they only set that it has to be made open access within (usually) a few months of publication. They do check for this as well.

The main issue with fees are that one, they already charge universities for access to journal content, thus they are “Double dipping” charging us twice for the same thing.

The second is that no one objects to paying fees, only that £3k on avg per article is too high. Nobody but the funders can afford it, and they do very little now to warrant it, apart from organising peer review (which they do not pay a wage for and is done at the time/expense of the academic reviewers) they simply put up the (formatted by the author) paper digitally (as there are fewer print journals anymore) and that’s it. They do not run the journals, they do not edit the journals, all done by academic staff at universities. They only basically host them now on their website and collect a lot of money.

Are you a publisher, a researcher or other?

My understanding is that charges levied against the author are primarily dependent on the JIF. If a low JIF journal sets the charge to high the author will go elsewhere.

I understand all about green and gold open access. I am the editor in chief of a journal (coincidentally owned by Elsevier - that has a low JIF and does not levy a page charge, unless gold open access is required).

I just looked up the Lancet. Their fees are not clearly explained in the Instructions and authors are billed after acceptance. They provide a link to open access charges that declare their charge is US$6,800.

I understand all about peer review having written articles on the topic, managed it, participated in it and been subjected to it. Peer review is not done so that the journal can be confident of the value of the work. It is done so that other researchers can be confident in the value of the work.

So ironically the problem with peer review is the academics who expect their work to be peer reviewed but refuse to do any peer review themselves. I have always been happy to referee papers (and grants) for no fee as it is all part of making the system better for me and my peers.

None of this has anything to do with predatory publishers by the way. They are simply criminals. Fraudsters.

Your points are fair but the problem here is a mass flounce by the editorial board of one journal, and the partial way the Grauniad has reported it. Academics have been colluding with the money making aspect of scientific publication ever since Thompson invented the JIF. My senior colleages are slaves to it, in as much as they try to use it to enslave junior colleagues.

Also if as you say in this example the editorial team think the fees charged to authors are too high for this journal, and the fees are unrelated to the JIF, it means the JIF is low and 'who cares anyway'? The JIF may be a duff tool but it is deffo the case that journals with big JIFs make the most income. I have seen smaller journals bundled into a package for uni library subscription (another source of income) but the page charge element is a different matter.

In my area the quality of peer review is poor (done by peer academics) and the literature awash with publication bias and false findings. All of this is driven by the self-imposed enslavement to the JIF by academics themselves. I'd be happy to publish my work by depositing it on a college web page, but no bugger would read it (this is the value of a high JIF journal with a mass circulation and readership). This is fine in some branches of research (maths and engineering) where the numbers of people are interested are small, and even top journals like IEEE have low JIFs.

In biomed where there is money to be spent and made it is a different game, but allowing publishers to make money is down to the academics. Most I know like to keep their head down and their mouth shut and consider me to be a gobshite trouble maker. Which I am. But we need to recognize where the problem lies.

 

[Green Cross Code Man]

Most interesting from the outside looking in. Doesn't exactly fill me with confidence in the system. Maybe we all need to bow to AI as long as AI ends up with better ethical stances than we achieve.

 

[Harry Wilson's tackle]

That's all interesting and I had no idea. So if I am reading something in a journal how do I know it's the best research there is as opposed to just what someone has been prepared to pay highest to appear in? Or put another way, does this mean the most prestigious journal isn't necessarily where the best research can be found? And isn't it then a self-perpetuating doom-spiral where people are citing research that isn't necessarily the best but because it appears in the most prominent place because someone paid for it to be there?

Earlier this year I had a book published in my field - no academic rigour to it really, but had to jump through endless hoops for the publisher and demonstrate my alleged credibility and a level of professional recognition before they'd even have a conversation with me about publishing. It slightly blows my tiny mind that academic publishing is less strict.

Brilliant question. The simply answer is we cannot tell, a priori, what is the best (most correct and relevant) research and how this maps to the JIF (which journal to trust, in other words) any more than we can for a 'newspaper' or TV channel (with some obvious egregious exceptions). We know what is the most complicated and expensive data when we see it - and that's the stuff published in the high JIF journals. Journals advertise their JIF (or 'impact factor) like a badge of honour, but you need an expert in the field to interpret it for you because it is based on citations. In medicine, JIFs can be very high - some Lancet journals (there are many) have JIFs over 70. In my area, 10 is considered very good. In engineering 2 is good, plus engineers don't really care because their sort of work is objective and easily testable. Unequivocal and meaningful biomed research is harder to generate, and to verify.

You have also guessed the existence publication bias, how stuff in high JIF journals is read by more people, trusted by more people and cited by more people (having more impact), but has no guarantee of being more likely to be correct. In fact Nature has one of the highest retraction rates (where authors 'retract' their paper because either they or (more likely) someone else has tried to repeat the work or build directly upon it and it turns out to be bolleaux). This is because Nature published work that 'if true would be of incredible importance'. Because it is important people will try to build on it. But it may not be true and if so this will soon be discovered (one hopes). Me and two others wrote a book on all this stuff a couple of years ago. PM me if you're interest. I warn you it is dull.

So in my game (biomed/drug research) the only test of whether a publication is correct is whether it is reproducible and leads on to something. Ironically, journals with low JIFs attract duller research that is less likely to trigger follow-up, and the work is instantly forgotten. The retraction rate in the journal will be low but that doesn't mean the work published is correct. And don't start me on Chinese 'paper mills'.

Paradoxically peer review at high JIF journals can be bloody onerous. Moon on a stick stuff. 'Maintaining standards of quality and depth' as they put it. And yet the one thing that is repeatedly overlooked is a exploration of whether the work is a fair test (done using randomized and blinded methods). A 'peer' may ask me to recapitulate my animal model data in a second species with trangenic knock out of the putative drug target (techno-bollocks), but won't ask me whether we cheated by doing the study unblinded. Madness. Anyway....

The business of generating and testing and publishing scientific information is, alas, undermined by the usual - it is run and done by human beings.

 

[Littlemo]

Are you a publisher, a researcher or other?

My understanding is that charges levied against the author are primarily dependent on the JIF. If a low JIF journal sets the charge to high the author will go elsewhere.

I understand all about green and gold open access. I am the editor in chief of a journal (coincidentally owned by Elsevier - that has a low JIF and does not levy a page charge, unless gold open access is required).

I just looked up the Lancet. Their fees are not clearly explained in the Instructions and authors are billed after acceptance. They provide a link to open access charges that declare their charge is US$6,800.

I understand all about peer review having written articles on the topic, managed it, participated in it and been subjected to it. Peer review is not done so that the journal can be confident of the value of the work. It is done so that other researchers can be confident in the value of the work.

So ironically the problem with peer review is the academics who expect their work to be peer reviewed but refuse to do any peer review themselves. I have always been happy to referee papers (and grants) for no fee as it is all part of making the system better for me and my peers.

None of this has anything to do with predatory publishers by the way. They are simply criminals. Fraudsters.

Your points are fair but the problem here is a mass flounce by the editorial board of one journal, and the partial way the Grauniad has reported it. Academics have been colluding with the money making aspect of scientific publication ever since Thompson invented the JIF. My senior colleages are slaves to it, in as much as they try to use it to enslave junior colleagues.

Also if as you say in this example the editorial team think the fees charged to authors are too high for this journal, and the fees are unrelated to the JIF, it means the JIF is low and 'who cares anyway'? The JIF may be a duff tool but it is deffo the case that journals with big JIFs make the most income. I have seen smaller journals bundled into a package for uni library subscription (another source of income) but the page charge element is a different matter.

In my area the quality of peer review is poor (done by peer academics) and the literature awash with publication bias and false findings. All of this is driven by the self-imposed enslavement to the JIF by academics themselves. I'd be happy to publish my work by depositing it on a college web page, but no bugger would read it (this is the value of a high JIF journal with a mass circulation and readership). This is fine in some branches of research (maths and engineering) where the numbers of people are interested are small, and even top journals like IEEE have low JIFs.

In biomed where there is money to be spent and made it is a different game, but allowing publishers to make money is down to the academics. Most I know like to keep their head down and their mouth shut and consider me to be a gobshite trouble maker. Which I am. But we need to recognize where the problem lies.

I come under other - I am a university librarian that works in Scholarly Communications.

My understanding is that charges levied against the author are primarily dependent on the JIF. If a low JIF journal sets the charge to high the author will go elsewhere.

The APC charges do vary across the journals and yes, the most prestigious ones tend to be more highly priced but I wouldn't say this is exactly true. The authors don't generally pay for papers themselves, the funders pay for it, as it is normally the case that a grant award will pay for it, research intensive institutions receive block grants from many of the funders to pay for papers and will do regardless of the cost. Unfunded authors need to rely on Green Open Access or now, taking advantage of the publisher deals that are now coming in, as few universities can afford to fund publishing. Its also the case that some High JIF journals are fully Open and don't charge at all, The BMJ range is one for example.

The cost of the paper doesn't, as far as I have seen with our academics, really influence where they publish. They usually want to publish in prestigious journals BUT the biggest barrier to that is the volume of submissions those journals get. There are also good reasons to publish in other places, where authors are working with certain groups (Global South etc, industry) then they may want to ensure publication in journals they can access as a priority over JIF, or working across subject areas, subject fit of the work etc. We rarely get any enquiries about the actual price as authors assume the funding will pay for it, and it usually will. Page charges are a different thing, not paid for by the grant, actual robbery especially since they don't print pages anymore. I don't think coloured pixels are anymore expensive than black and white ones.

The JIF may be a duff tool but it is deffo the case that journals with big JIFs make the most income

Again this isn't strictly true. The main income for publishers are the access agreements. They run into high tens (or more) of thousands each year, I can't even say how much exactly as the publishers have secrecy clauses that don't allow it, we aren't allowed to say the increase each year but its a high %. Can't even talk about it with other Uni's. They bundle the journals together (whether you want them or not) so the true cost of each journal is hidden and its not a sole income maker.

My main issue with your focus on JIF is that its not driving things as much as you think it is. The main issue is that academia relies too heavily on publishing papers as its main means of research assessment and valued outputs, the focus is on quantity more than quality. Most of the Uni's I work with are trying to change that, for example where I am, we no longer use JIF in research assessment at all, especially not for things like promotion or performance review. Its a very flawed metric, which doesn't tell you anything about the actual journal article or really anything that valuable about the journal itself. It is hard to change research culture though, a bit like turning an oil tanker, so it will take time to embed. This is why things such as Impact (both academic and Societal) are starting to become more important as well, because you need to have an alternative means to demonstrate the value of your work.

I found it interesting that during the last REF, one Head of Department said that 2 papers the department had in a high JIF journal weren't good enough to submit. I feel that really demonstrates why not to rely on JIF anymore.

 

[Harry Wilson's tackle]

I come under other - I am a university librarian that works in Scholarly Communications.



The APC charges do vary across the journals and yes, the most prestigious ones tend to be more highly priced but I wouldn't say this is exactly true. The authors don't generally pay for papers themselves, the funders pay for it, as it is normally the case that a grant award will pay for it, research intensive institutions receive block grants from many of the funders to pay for papers and will do regardless of the cost. Unfunded authors need to rely on Green Open Access or now, taking advantage of the publisher deals that are now coming in, as few universities can afford to fund publishing. Its also the case that some High JIF journals are fully Open and don't charge at all, The BMJ range is one for example.

The cost of the paper doesn't, as far as I have seen with our academics, really influence where they publish. They usually want to publish in prestigious journals BUT the biggest barrier to that is the volume of submissions those journals get. There are also good reasons to publish in other places, where authors are working with certain groups (Global South etc, industry) then they may want to ensure publication in journals they can access as a priority over JIF, or working across subject areas, subject fit of the work etc. We rarely get any enquiries about the actual price as authors assume the funding will pay for it, and it usually will. Page charges are a different thing, not paid for by the grant, actual robbery especially since they don't print pages anymore. I don't think coloured pixels are anymore expensive than black and white ones.




Again this isn't strictly true. The main income for publishers are the access agreements. They run into high tens (or more) of thousands each year, I can't even say how much exactly as the publishers have secrecy clauses that don't allow it, we aren't allowed to say the increase each year but its a high %. Can't even talk about it with other Uni's. They bundle the journals together (whether you want them or not) so the true cost of each journal is hidden and its not a sole income maker.

My main issue with your focus on JIF is that its not driving things as much as you think it is. The main issue is that academia relies too heavily on publishing papers as its main means of research assessment and valued outputs, the focus is on quantity more than quality. Most of the Uni's I work with are trying to change that, for example where I am, we no longer use JIF in research assessment at all, especially not for things like promotion or performance review. Its a very flawed metric, which doesn't tell you anything about the actual journal article or really anything that valuable about the journal itself. It is hard to change research culture though, a bit like turning an oil tanker, so it will take time to embed. This is why things such as Impact (both academic and Societal) are starting to become more important as well, because you need to have an alternative means to demonstrate the value of your work.

I found it interesting that during the last REF, one Head of Department said that 2 papers the department had in a high JIF journal weren't good enough to submit. I feel that really demonstrates why not to rely on JIF anymore.

Many thanks for that. I agree with your points except the one about volume versus quality. Where I work the research deans are competitive within the college and in my research division many of us are rejected by the head at ref time, only to be 'picked up' and returned by smaller research divisions. Lower down the ladder the ref return may be more strategic for 'coherence', picking a lower JIF output because it themes better with the whole.

But this is all a racket and expensive. We spend millions employing consultants to help us game the ref return.

And I can guarantee that where I work a professorship requires high grant income and papers in high JIF journals. Several pals of mine throw negative findings in the bin because no high JIF journal will publish a negative finding. Volume counts for nothing.

All the best

 

[Harry Wilson's tackle]

probably right. or wrong. bit of a fringe topic? (though i type as someone give a sensible reply )

what have the economic journals done, they must have solved this, right?

Do you mean journals about the economy or journals that are value for money?

 

[Harry Wilson's tackle]

Most interesting from the outside looking in. Doesn't exactly fill me with confidence in the system. Maybe we all need to bow to AI as long as AI ends up with better ethical stances than we achieve.

Radio 5 suggested that unis are going to be told how to use AI. At my uni we have been told to tell students not to use it (plagiarsm)......

 

[Green Cross Code Man]

Radio 5 suggested that unis are going to be told how to use AI. At my uni we have been told to tell students not to use it (plagiarsm)......

It's certainly a discussion held in educational establishments everywhere I would imagine. Even in my workplace, a primary school, we are working our way around staff using it for marking etc. And I teach computing so it's more than relevant as a subject on it's own. Dangerous and fascinating, we are like moths to a candle and may just be superceding ourselves all too successfully.

 

[Harry Wilson's tackle]

It's certainly a discussion held in educational establishments everywhere I would imagine. Even in my workplace, a primary school, we are working our way around staff using it for marking etc. And I teach computing so it's more than relevant as a subject on it's own. Dangerous and fascinating, we are like moths to a candle and may just be superceding ourselves all too successfully.

It is quite exciting. I love disruptive change.

Personally I haven't yet been mugged by an AI bit of coursework. A benefit of asking questions from within the edge of the spectrum

 

[Green Cross Code Man]

It is quite exciting. I love disruptive change.

Personally I haven't yet been mugged by an IA bit of coursework. A benefit of asking questions from within the edge of the spectrum

According to research, the AI is pretty good at answering more obvious questions ie almost 100% accuracy in multiple choice, but give it something that requires critical thinking and it is as yet quite hopeless. It is also laughably odd with citations whereby AI will simply make up names of scientists for example and give a date for a non-existent paper.

 

[Harry Wilson's tackle]

According to research, the AI is pretty good at answering more obvious questions ie almost 100% accuracy in multiple choice, but give it something that requires critical thinking and it is as yet quite hopeless. It is also laughably odd with citations whereby AI will simply make up names of scientists for example and give a date for a non-existent paper.

Precisely.

I never use MCQs. Nor SAQs. I'm a lateral thinking essay man. This is university not f***ing colouring in. I have done a lot of external examining around UK unis, and the SAQs do my head in. First rule - map the Q to an idea A, then mark down for omission and error. Rocket science? Seems to be

 

[beorhthelm]

According to research, the AI is pretty good at answering more obvious questions ie almost 100% accuracy in multiple choice, but give it something that requires critical thinking and it is as yet quite hopeless. It is also laughably odd with citations whereby AI will simply make up names of scientists for example and give a date for a non-existent paper.

saw something recently from team working with GPT that ~20% results are "hallucinations", stuff its just made up.

to be fair i thought this gives it more air of intelligence, to err is human and all that.

 

[Green Cross Code Man]

saw something recently from team working with GPT that ~20% results are "hallucinations", stuff its just made up.

to be fair i thought this gives it more air of intelligence, to err is human and all that.

I can well believe that. One thing I found a few weeks back, was that if I quoted lyrics from the Canadian band Rush, it would invariably give a lovely polite answer and then wrongly attribute them to someone else. For example, 'Who wrote the lyric "When the men who hold high places, must be the ones to start"?' It would say something like 'This lyric was written in 1968 by Jimi Hendrix'. I'd respond 'Jimi Hendrix DID NOT write that lyric, who did?' and it would apologise and then very definitely say it was written by Roger Waters, Bob Dylan and so on and so on. When at last I suggested the correct attribution, it would again apologise and correctly give information as to the time and album it came from. Odd. Chat GPT obviously had the information but as you say 'hallucinated' the answer. I wonder if that is a natural evolution - i.e. that it is trying somehow to produce a less deterministic and more random way of thinking as presumably it needs that kind of output in order to learn to seem more human and therefore reach a level that can seem inspired?

 

[Happy Exile]

I can well believe that. One thing I found a few weeks back, was that if I quoted lyrics from the Canadian band Rush, it would invariably give a lovely polite answer and then wrongly attribute them to someone else. For example, 'Who wrote the lyric "When the men who hold high places, must be the ones to start"?' It would say something like 'This lyric was written in 1968 by Jimi Hendrix'. I'd respond 'Jimi Hendrix DID NOT write that lyric, who did?' and it would apologise and then very definitely say it was written by Roger Waters, Bob Dylan and so on and so on. When at last I suggested the correct attribution, it would again apologise and correctly give information as to the time and album it came from. Odd. Chat GPT obviously had the information but as you say 'hallucinated' the answer. I wonder if that is a natural evolution - i.e. that it is trying somehow to produce a less deterministic and more random way of thinking as presumably it needs that kind of output in order to learn to seem more human and therefore reach a level that can seem inspired?

William Gibson said that the future is already here, just not evenly distributed, and that seems true of AI. ChatGPT is already a bit "old hat" and limited compared to some of the AI that's already in use (not speculative, but in use already). That said, I use ChatGPT a few times a week to simplify tasks and a software engineer mate reckons his team are 75% more productive using it. They have it generate baseline code which they then edit, or they paste in their code and have it debug what they've done. Both much quicker than humans alone.

I mostly use it to summarise documents - yesterday for example I was reviewing a number of case studies where huge long narrative blocks of text essentially hid what people were trying to say. ChatGPT pulled out the key points in no time and gave me enough to work out if I wanted to look deeper. I use it on my own writing too. I was asked to submit 200 words on a topic for a trade magazine. Bashed out a first draft of just under 300, Asked ChatGPT to edit it down to 200 and correct grammar and make recommendations for clarity...and I only made a handful of amends myself after from the output so it retained my "voice".

In my use-case that's the strength of ChatGPT - it's like a very fast, very capable assistant augmenting what I do not replacing it. I wouldn't use it for searching for things, but AI is being integrated into searching. At the moment you have to go to the site to use it but it's being seamlessly integrated into our day-to-day...Bing is about to be heavily changed with AI features, Google too...in a couple of years we'll look at internet search as it is now as something very quaint and basic.

We're barely scratching the surface though, and some elements of what I do for a living will 100% be replaced by AI soon. IBM announced last week a recruitment freeze on 30% of roles because they think AI will replace them in the next 5 years - while most analysts think new jobs will be created (many jobs people do now didn't exist pre-Internet or smart phone for example) there's still a prediction of a net job loss overall globally. Interesting times.