X Close

Humans will defeat the chatbots Amid AI platitudes, mankind's input will still be prized

What will become of writing? Credit: Ding Genhou/VCG via Getty

What will become of writing? Credit: Ding Genhou/VCG via Getty


December 15, 2022   7 mins

I have always written for a living. But most of that wasn’t the fun kind. Before my first essay for UnHerd was published, just over three years ago, I mostly did the other kind of writing: the anonymous stuff that makes up the vast bulk of activities under the broad heading “professional writer”. I’ve written corporate blogs and tweets, bashed out press releases, white papers and website copy, and countless other subtypes of writing for which the commercial world will pay.

It’s dull, and not even very lucrative. Now, to make matters worse, it appears that the robots may be coming for even its meagre wages. ChatGPT, a “large language learning model” optimised for conversation, launched at the end of November, and within a week garnered over a million users. It represents a step-change in AI writing, synthesising immense bodies of information and responding to chat prompts with uncannily clear, coherent and usually fairly accurate paragraphs of reasonably well-written text.

It appears, in other words, that robot “writers” are now so advanced they can match or exceed an averagely competent human writer across a vast range of topics. What, then, does this mean for human toilers in the textual saltmines? There have been breathless reports on how robot writing has the potential to disrupt all manner of writing-related fields, from online search through the kind of “content marketing” I used to do for a living, to academic writing, cheating at schoolwork and news reporting.

Are writers destined to go the way of the artisan textile-makers who starved to death after the inception of the mechanical loom? Perhaps. But this is complicated by the fact that it was the industrialisation of writing that created the role of “professional writer” as we know it — along with much else besides. Now, the digital revolution is on its way to destroying that model of authorship — and with it, driving former denizens of the “world of letters” into new and strange cultural roles.

As the writer Adam Garfinkle has argued, the world of print was, to all intents and purposes, the democratic world of liberal norms. In the UK, literature, high finance, and a great many political norms we now take for granted emerged from the same heady atmosphere in the coffee-houses of 18th-century London. It was this explosion of competing voices, that eventually distilled into ideas of “high” and “low” literary culture, norms of open but (relatively) civil debate on rational terms — and, as literacy spread, the ideas of rationally-based, objective “common knowledge” and mass culture as such.

Print culture in the 19th century saw an astonishing volume of writing produced and devoured for self-improvement, entertainment and political engagement. Industrial workers self-funded and made use of travelling libraries stuffed with classics in translation, exhibitions and museums were popular, and public lectures could spill out into the street. These conditions also produced the “author”, with a capital A, in the sense that those of us over 40 still retain. This figure has two key characteristics: first, a measure of cultural cachet as a delivery mechanism for common culture, and second some means of capturing value directly from this activity.

But the digital revolution has already all but destroyed the old authorship model, even if a small subset of writers still manages to get very rich. De-materialising print served to democratise “authorship”, publishing, and journalism, but by the same token made it far more difficult to get paid. There are a great many bloggers and “content creators” out there; meanwhile very few authors make a living from writing books, and journalist salaries have been stagnant for years. Growing numbers of would-be writers simply head (as I did) straight to PR and communications where the money is somewhat better.

Now, when even routine PR and communications writing work can increasingly be done by robots, we can expect the terrain to mutate still further. We can expect some of the types of work I spent 15 years doing to become fact-checking work instead: a machine, after all, doesn’t know how to judge if its output makes sense. We can glimpse some of the necessity of this in the fact that Stack Overflow, a knowledge-sharing platform for organisational knowledge, has already temporarily banned input from ChatGPT because too many people were posting robot-generated content, and it wasn’t reliable enough.

But even those writers lucky enough to avoid the sense-checking saltmines will struggle to make a living unless they’re already famous. In this context, expect to see patronage making a comeback: a phenomenon that is, in fact, a reversion to the historic norm for creators. The appearance of “slam poet” Amanda Gorman at the inauguration of US President Joe Biden is only one prominent instance of the trend; I can think of a great many interesting and popular new publications, with clear cultural value, which are funded wholly or in part by wealthy philanthropists.

But this won’t be the end of ways that a mechanised acceleration in the volume of content will change the field of letters. For if the print era cultivated habits of long-form reading and thinking that profoundly structured our pre-digital culture and politics, the digital revolution has already taken us from scarcity to excess — and is taking print-era liberal discursive norms with it. In the brave new digital world, “deep literacy” is replaced by an attention economy in which every paragraph has to compete with a trillion others, meaning the incentive is for engagement: thrilling stories, grotesquerie and clickbait.

Think of the change in tone since the New York Times shifted its focus to digital subscriptions, and you’ll see what I mean. Now accelerate that a millionfold with AI-generated text. Anyone clinging to any residual optimism about the norms of objectivity or civil discourse under those conditions hasn’t been paying attention. Nor have those still optimistic about free speech and the marketplace of ideas. For along with an attention economy comes attention politics: the tussle to control what people notice, or who gets a platform, in which some censorship is now inevitable.

We may decry this as “illiberal”, but it’s not wholly unjustified. For while the Twitter Files have revealed just how consequential control of the censorship mechanisms can be, few would deny that “misinformation” is, under digital conditions, genuinely easy to manufacture and spread. Consider, for example, the sophisticated propaganda campaign, powered by bot farms, through which Chinese authorities propagandised Western electorates into demanding their governments introduce China-style Covid lockdowns.

In this context, “speech” itself fades in significance, relative to control over the algorithmic parameters of speech. And the fight is on over who sets the rules governing which opinions the algorithms are allowed to index, transmit, learn from – or suppress. The writer James Poulos calls this “catechising the bots”. The necessity of such catechism was learned the hard way, in 2016, when Microsoft released Tay, a chatbot that learned from user input fed in via Twitter. Within a day, mischievous contributors had “taught” Tay to say Nazi things of the kind that recently got even Kanye West cancelled.

In this sense, AI design is rapidly taking on an explicitly moral dimension, as a kind of post-human curator of the “common culture”. One excited tech CEO recently celebrated this as a positive change, saying of Chat-GPT that “It’s just like a human being except it has all the world’s knowledge.”

In the 18th and 19th centuries, then, the Author with a capital A enjoyed status as guardian of and contributor to the common culture. Today, though, we’ve long since ceded that terrain to robots, and to the cyborg theologians who catechise them. No matter how fast you read, there’s no competing with a machine that can instanteously index and synthesise a century’s books; an entire generation has now grown up with Google. Chat-GPT just formalises robots as curators of the consensus view. What’s left, then, for human authors to contribute?

To state the obvious, there are still plenty of things those robots can’t do besides fact-checking — not least because of their catechism. If you want the bland, normative, politically unimpeachable top-line consensus on any given topic, ask Chat-GPT about it and you’ll get three paragraphs of serviceable prose. That’s all well and good for cheating in a school essay, but bland, faceless, normative views aren’t the only kinds of writing for which there’s an appetite.

In the AI age, expect human writers to specialise in the esoteric, tacit, humorous or outright forbidden speech that robots either can’t capture (ChatGPT is reportedly rubbish at jokes), or that “AI bias” teams work hard to scrub from the machine. As one Right-wing anon put it (appropriately colourfully) recently, one way to prove you’re human as a writer in the Chat-GPT age will be to say incredibly racist things.

If I’m right, we can kiss goodbye to what’s left of the print-era connection between “author” and “authority” in the sense of generally-accepted objectivity. But I expect many creators to thrive anyway. It’s just that instead of trying to beat the machine at synthesising the consensus, humans will be prized for their idiosyncratic curation of implicit, emotive, or outright forbidden meanings, amid the robot-generated wasteland of recycled platitudes.

Successful creators in this field may not even be writers in the old sense; at the baroque end of this emerging field, we might place Infowars host Alex Jones, while at the more mainstream-looking end, Elon Musk’s chosen Twitter Files mouthpiece, Bari Weiss. Whatever you think of the narratives either produces, they are less “authors” in the old sense than sensemakers amid the digital noise.

All of this will feed into the political cleavage I described last week: the emerging disagreement over whether it’s wiser to trust the judgement of human individuals, or the inhuman (and catechised) synthetic consensus of the machine. This divide, which I characterised as “Caesarism” vs “Swarmism” is, in practice, not really about despotic governance as such. (The Covid era was, after all, a sharp object lesson in how tyrannical swarmism can be.) Rather, it’s about whether we think there’s anything of value in individual human judgement, despite (or perhaps because of) its mess and bias and partiality.

So the world of post-liberal letters will displace the human element from consensus “common culture”. Instead, some humans will be employed sanity-checking the machine consensus, others will tussle over the right to catechise the bots, and others again will forge high-profile careers curating new kinds of sensemaking. Meanwhile, we’ll see the moral status of human sense-making grow ever more politicised, as elites fight over whether or not all of culture can (or should) simply be automated — however nightmarish the result so far whenever this has been tried.

If something remains distinctively the domain of human creators in this context, it’s not curating common knowledge but the hidden kind: the esoteric, the taboo, the implicit and the mischievous. The machine still can’t meme. May it never learn.


Mary Harrington is a contributing editor at UnHerd.

moveincircles

Join the discussion


Join like minded readers that support our journalism by becoming a paid subscriber


To join the discussion in the comments, become a paid subscriber.

Join like minded readers that support our journalism, read unlimited articles and enjoy other subscriber-only benefits.

Subscribe
Subscribe
Notify of
guest

50 Comments
Most Voted
Newest Oldest
Inline Feedbacks
View all comments
Steve Murray
Steve Murray
1 year ago

I continue to be no less than astounded, not only by Mary’s ability to conglomerate a number of vital culture-tech emergencies (in the sense of having just emerged) but to do so with style.

I’d no idea about her background as a writer, but it’s served her well, even though she almost certainly wouldn’t have realised that whilst going through the motions in her former career. She’s positioned herself, as far as i’m concerned, at the forefront of cultural interpretation simply by bringing to our attention and synthesising trends which might otherwise escape detection.

That’s not to say she’ll always get it right, even if there was such a thing; but that hardly matters. She’s a living example of the human ability she seeks to foretell in this article, always a step or two ahead of the bots.

Last edited 1 year ago by Steve Murray
James Jenkin
James Jenkin
1 year ago
Reply to  Steve Murray

Yes, she is my hero. (Did I just say that out loud?)

Jerry Smith
Jerry Smith
1 year ago
Reply to  Steve Murray

Couldn’t agree more. This article is brilliant, just brilliant. Hail Mary!

Saul Sorenti
Saul Sorenti
1 year ago
Reply to  Steve Murray

She’s not the messiah, she’s a very mischievous girl!

Steve Murray
Steve Murray
1 year ago
Reply to  Saul Sorenti

And all the better for it!

Anthony Michaels
Anthony Michaels
1 year ago
Reply to  Saul Sorenti

She IS the Mesiah! And I should know, I’ve followed a few!

Steve Murray
Steve Murray
1 year ago
Reply to  Saul Sorenti

And all the better for it!

Anthony Michaels
Anthony Michaels
1 year ago
Reply to  Saul Sorenti

She IS the Mesiah! And I should know, I’ve followed a few!

Jonas Moze
Jonas Moze
1 year ago
Reply to  Steve Murray

”Humans will defeat the chatbots”
Mary’s headline – and I did not feel she demonstrated her thesis – lots of fun and pretty writing, but all I got was a that a few human writers will remain, as a few artisan weavers are out there making Craft/Art decorative things to sell – but for fabric we turn to the machine made.

Steve Murray
Steve Murray
1 year ago
Reply to  Jonas Moze

I doubt it was Mary’s headline.
Quite often, it seems the Unherd editorial team will produce an headline to grab attention, and sometimes change the headline whilst the topic is still “live”.

Steve Murray
Steve Murray
1 year ago
Reply to  Jonas Moze

I doubt it was Mary’s headline.
Quite often, it seems the Unherd editorial team will produce an headline to grab attention, and sometimes change the headline whilst the topic is still “live”.

James Jenkin
James Jenkin
1 year ago
Reply to  Steve Murray

Yes, she is my hero. (Did I just say that out loud?)

Jerry Smith
Jerry Smith
1 year ago
Reply to  Steve Murray

Couldn’t agree more. This article is brilliant, just brilliant. Hail Mary!

Saul Sorenti
Saul Sorenti
1 year ago
Reply to  Steve Murray

She’s not the messiah, she’s a very mischievous girl!

Jonas Moze
Jonas Moze
1 year ago
Reply to  Steve Murray

”Humans will defeat the chatbots”
Mary’s headline – and I did not feel she demonstrated her thesis – lots of fun and pretty writing, but all I got was a that a few human writers will remain, as a few artisan weavers are out there making Craft/Art decorative things to sell – but for fabric we turn to the machine made.

Steve Murray
Steve Murray
1 year ago

I continue to be no less than astounded, not only by Mary’s ability to conglomerate a number of vital culture-tech emergencies (in the sense of having just emerged) but to do so with style.

I’d no idea about her background as a writer, but it’s served her well, even though she almost certainly wouldn’t have realised that whilst going through the motions in her former career. She’s positioned herself, as far as i’m concerned, at the forefront of cultural interpretation simply by bringing to our attention and synthesising trends which might otherwise escape detection.

That’s not to say she’ll always get it right, even if there was such a thing; but that hardly matters. She’s a living example of the human ability she seeks to foretell in this article, always a step or two ahead of the bots.

Last edited 1 year ago by Steve Murray
Peter Johnson
Peter Johnson
1 year ago

“ If you want the bland, normative, politically unimpeachable top-line consensus on any given topic, ask Chat-GPT about it and you’ll get three paragraphs of serviceable prose.” Or just read consensus narrative served up daily by the entire mainstream media.

Philip Burrell
Philip Burrell
1 year ago
Reply to  Peter Johnson

I struggle to see this supposed consensus when I go on the BBC website each morning to look at the headlines from all the daily newspapers!

David Simpson
David Simpson
1 year ago
Reply to  Peter Johnson

Which is why Chat-GPT is really a threat to all the hacks playing the game in the MSM. They should be very afraid – something that can write exactly what they write, only grammatically correct and properly spelt.

Meanwhile all us sceptics out in the dark will continue to be ignored, but to happily engage with each other. I see a new heaven, and a new dawn.

Jonas Moze
Jonas Moze
1 year ago
Reply to  David Simpson

I await the Drop Down Menue soon to come where we pick the theme and style

Victorian Steady

Modernist Cool

Youthfull Tic-Toc

Trendy Liberal

Postmodernist Neo-Marxist

and

Rightwing Loon

ml holton
ml holton
1 year ago
Reply to  Jonas Moze

… It’s already happening!

Historian Neil Ferguson used a demo on the last ‘Goodfellows’ podcast, asking for an opinion ‘in the style of’ himself. It was rather startlingly, albeit somewhat generic, in that, it can only build on what’s available, not on what he is thinking now …

I’ve seen another example on Twitter ‘in the style of’ Yeats. The context was contemporary, but the style was definitely ‘Yeats’.

Kind of depressing, actually.

Ted Ditchburn
Ted Ditchburn
1 year ago
Reply to  ml holton

I have spent years watching people look for corners where us humans can find some value in what we do… the first I can remember was people will always want the feel of a newspaper (and the serendipity of the curated ‘read’)… that one didn’t wear well.
As a press photographer on a far off planet a long long time ago I remember noticing that the most memorable images of the Gulf war were, for the first time video snippets, and the stills, from robot cameras on US planes (now drones) or in the missiles themselves, blowing up whatever they were aimed at.
Thousands of photographers including many hundreds with various bits of the military produced virtually nothing.

Until it was over anyway, on that road to Basra.
These days it’s domestic cctv, dashcams, and mobile phones.
as the author said, some photographers can earn a living ‘within the new system’ but they are fewer every year and the living is always more precarious every year.

ml holton
ml holton
1 year ago
Reply to  Ted Ditchburn

… it will be interesting to see how this radically transforms our current system of education … As example, a neuro-link chip implant enhanced with AI & internet access, at birth, could place any human into a pre-ordained slot for optimum civil functionality. School, per se, could become redundant.

ml holton
ml holton
1 year ago
Reply to  Ted Ditchburn

… it will be interesting to see how this radically transforms our current system of education … As example, a neuro-link chip implant enhanced with AI & internet access, at birth, could place any human into a pre-ordained slot for optimum civil functionality. School, per se, could become redundant.

Ted Ditchburn
Ted Ditchburn
1 year ago
Reply to  ml holton

I have spent years watching people look for corners where us humans can find some value in what we do… the first I can remember was people will always want the feel of a newspaper (and the serendipity of the curated ‘read’)… that one didn’t wear well.
As a press photographer on a far off planet a long long time ago I remember noticing that the most memorable images of the Gulf war were, for the first time video snippets, and the stills, from robot cameras on US planes (now drones) or in the missiles themselves, blowing up whatever they were aimed at.
Thousands of photographers including many hundreds with various bits of the military produced virtually nothing.

Until it was over anyway, on that road to Basra.
These days it’s domestic cctv, dashcams, and mobile phones.
as the author said, some photographers can earn a living ‘within the new system’ but they are fewer every year and the living is always more precarious every year.

ml holton
ml holton
1 year ago
Reply to  Jonas Moze

… It’s already happening!

Historian Neil Ferguson used a demo on the last ‘Goodfellows’ podcast, asking for an opinion ‘in the style of’ himself. It was rather startlingly, albeit somewhat generic, in that, it can only build on what’s available, not on what he is thinking now …

I’ve seen another example on Twitter ‘in the style of’ Yeats. The context was contemporary, but the style was definitely ‘Yeats’.

Kind of depressing, actually.

Jonas Moze
Jonas Moze
1 year ago
Reply to  David Simpson

I await the Drop Down Menue soon to come where we pick the theme and style

Victorian Steady

Modernist Cool

Youthfull Tic-Toc

Trendy Liberal

Postmodernist Neo-Marxist

and

Rightwing Loon

Philip Burrell
Philip Burrell
1 year ago
Reply to  Peter Johnson

I struggle to see this supposed consensus when I go on the BBC website each morning to look at the headlines from all the daily newspapers!

David Simpson
David Simpson
1 year ago
Reply to  Peter Johnson

Which is why Chat-GPT is really a threat to all the hacks playing the game in the MSM. They should be very afraid – something that can write exactly what they write, only grammatically correct and properly spelt.

Meanwhile all us sceptics out in the dark will continue to be ignored, but to happily engage with each other. I see a new heaven, and a new dawn.

Peter Johnson
Peter Johnson
1 year ago

“ If you want the bland, normative, politically unimpeachable top-line consensus on any given topic, ask Chat-GPT about it and you’ll get three paragraphs of serviceable prose.” Or just read consensus narrative served up daily by the entire mainstream media.

Graham Willis
Graham Willis
1 year ago

It is grand scale mimicry, uncomprehending. I am sure it will produce convincing French philosophy. It would be interesting if someone could get a PhD with it.

James Jenkin
James Jenkin
1 year ago
Reply to  Graham Willis

That is a great question. You should try. (When you ask for X number of words it seems to have a breakdown.)

James Jenkin
James Jenkin
1 year ago
Reply to  Graham Willis

That is a great question. You should try. (When you ask for X number of words it seems to have a breakdown.)

Graham Willis
Graham Willis
1 year ago

It is grand scale mimicry, uncomprehending. I am sure it will produce convincing French philosophy. It would be interesting if someone could get a PhD with it.

R Wright
R Wright
1 year ago

The idea of a racism Voight-Kampff test needing to be used to check one is human amid a miasma of artificial political correctness might be the funniest yet most terrifying thing I have ever heard. A superb essay as ever.

R Wright
R Wright
1 year ago

The idea of a racism Voight-Kampff test needing to be used to check one is human amid a miasma of artificial political correctness might be the funniest yet most terrifying thing I have ever heard. A superb essay as ever.

Robert Hochbaum
Robert Hochbaum
1 year ago

“It’s just like a human being except it has all the world’s knowledge.”
This is just the latest statement from a tech world leader that has made me, over time, come to distrust the whole culture they have inspired and built. ChatGPTs ability to rapidly assemble huge quantities of data, along with its ability to select an appropriate style guide for assembling that data into 750 words is not thinking. It’s not creating. I’m not even sure all that data out there is actually knowledge. Increasingly, its just data. Is the data out there that shows that many people around the world believe the world is flat knowledge?
Perhaps that statement was taken out of context but I doubt it. These are the same people who casually call the physical world live in ‘meat space’. That phrase creeps me out. No, ChatGPT is most definitely NOT just like a human being.

David Baker
David Baker
1 year ago

Agreed on all points. The culture of the tech world is creepy, and trending hard towards anti-human. And despite the impacts a super fast indexing machine might create (which is, at the core, what all computers are), there is no thinking going on.

Thinking requires perspective and “aboutness,” and humans think through our embodied perspective in the world. While not a in-depth analysis, the Chinese room thought experiment is a great primer on the difference between thinking and the algorithmic sorting done by AI.

David Baker
David Baker
1 year ago

Agreed on all points. The culture of the tech world is creepy, and trending hard towards anti-human. And despite the impacts a super fast indexing machine might create (which is, at the core, what all computers are), there is no thinking going on.

Thinking requires perspective and “aboutness,” and humans think through our embodied perspective in the world. While not a in-depth analysis, the Chinese room thought experiment is a great primer on the difference between thinking and the algorithmic sorting done by AI.

Robert Hochbaum
Robert Hochbaum
1 year ago

“It’s just like a human being except it has all the world’s knowledge.”
This is just the latest statement from a tech world leader that has made me, over time, come to distrust the whole culture they have inspired and built. ChatGPTs ability to rapidly assemble huge quantities of data, along with its ability to select an appropriate style guide for assembling that data into 750 words is not thinking. It’s not creating. I’m not even sure all that data out there is actually knowledge. Increasingly, its just data. Is the data out there that shows that many people around the world believe the world is flat knowledge?
Perhaps that statement was taken out of context but I doubt it. These are the same people who casually call the physical world live in ‘meat space’. That phrase creeps me out. No, ChatGPT is most definitely NOT just like a human being.

Xaven Taner
Xaven Taner
1 year ago

Thought provoking as always. I wonder though whether it would be possible to distinguish a human Author (clinging onto their capitalisation) who deals in alternative sensemaking – the esoteric, the taboo, the implicit and the mischievous – from the legions trying to keep their head above water in the attention economy, whose content is progressively pushed towards the extreme, high arousal content that the platforms reward? Isn’t the truth that in the end, everyone is pushed into making clickbait trash?

Xaven Taner
Xaven Taner
1 year ago

Thought provoking as always. I wonder though whether it would be possible to distinguish a human Author (clinging onto their capitalisation) who deals in alternative sensemaking – the esoteric, the taboo, the implicit and the mischievous – from the legions trying to keep their head above water in the attention economy, whose content is progressively pushed towards the extreme, high arousal content that the platforms reward? Isn’t the truth that in the end, everyone is pushed into making clickbait trash?

Lesley van Reenen
Lesley van Reenen
1 year ago

This is the amateur bungling version of what really awaits humanity. Focus on this and you will lose sight of what really is happening. AI is a horse that has already bolted from the stables.
I watched a compelling interview done on AI with Rogan and Musk done in 2018 and Musk by that stage had already had high level talks with Obama and the US government about the dangers of AI. Musk by 2018 already felt that the horse had bolted and has a fatalistic approach to AI…. It is going to happen, AI is way superior and humans are going to have to try to mitigate the risk and engage in the process or be doomed.
This is the long form fascinating interview…. Clips are also available. I watched over a few days a few years ago. Prescient. I needed to check… he does talk about Neuralink.
https://www.youtube.com/watch?v=ycPr5-27vSI

Lesley van Reenen
Lesley van Reenen
1 year ago

This is the amateur bungling version of what really awaits humanity. Focus on this and you will lose sight of what really is happening. AI is a horse that has already bolted from the stables.
I watched a compelling interview done on AI with Rogan and Musk done in 2018 and Musk by that stage had already had high level talks with Obama and the US government about the dangers of AI. Musk by 2018 already felt that the horse had bolted and has a fatalistic approach to AI…. It is going to happen, AI is way superior and humans are going to have to try to mitigate the risk and engage in the process or be doomed.
This is the long form fascinating interview…. Clips are also available. I watched over a few days a few years ago. Prescient. I needed to check… he does talk about Neuralink.
https://www.youtube.com/watch?v=ycPr5-27vSI

Guy Aston
Guy Aston
1 year ago

ChatGPT is coffee table writing. It certainly failed to excite me, but I suspect it is early days for this sort of thing. Will it ever write a novel? I doubt it as that needs a Caesar’s mind.

Guy Aston
Guy Aston
1 year ago

ChatGPT is coffee table writing. It certainly failed to excite me, but I suspect it is early days for this sort of thing. Will it ever write a novel? I doubt it as that needs a Caesar’s mind.

Josef O
Josef O
1 year ago

I still wait to see how this ChatGPT exactly performs. At first sight it seems perfect for a kind of Orwell’s “New Speak”.

Josef O
Josef O
1 year ago

I still wait to see how this ChatGPT exactly performs. At first sight it seems perfect for a kind of Orwell’s “New Speak”.

Timothy Corwen
Timothy Corwen
1 year ago

Good analysis by Mary Harrington, once again. This should not be unexpected: Our worth as persons has been standardised and monetised for decades. We chase after competitive validated-entitlement worth (for attention, consideration, approval, support, and having an influence) and ignore greater-value worth (dedication to a domain of deeper value). To paraphrase John Donne: ‘We run to algorithms, and algorithms meet us as fast’.
Timothy Corwen (author, The Worth of a Person)

Timothy Corwen
Timothy Corwen
1 year ago

Good analysis by Mary Harrington, once again. This should not be unexpected: Our worth as persons has been standardised and monetised for decades. We chase after competitive validated-entitlement worth (for attention, consideration, approval, support, and having an influence) and ignore greater-value worth (dedication to a domain of deeper value). To paraphrase John Donne: ‘We run to algorithms, and algorithms meet us as fast’.
Timothy Corwen (author, The Worth of a Person)

Prashant Kotak
Prashant Kotak
1 year ago

But seriously can we not see, we are already retreating into ever decreasing circles of what distinguishes humans from machine intelligence, as each bastion of human-only activity falls by the wayside one by one? So where’s the last stand?

Prashant Kotak
Prashant Kotak
1 year ago
Reply to  Prashant Kotak

And just for the avoidance of doubt, what I’m implying, is that the situation is already gone past the point of hope expressed by MH in that last sentence.

https://singularityhub.com/2022/12/13/deepminds-alphacode-conquers-coding-performing-as-well-as-humans/

Last edited 1 year ago by Prashant Kotak
chris sullivan
chris sullivan
1 year ago
Reply to  Prashant Kotak

Agreed – another level of the ‘dumbing down’ of humanity – until we/they are all zombie like creatures unable to generate an original thought of their/our own – and easily led towards whatever the ‘masters’ desire. We are living in an Orwell novel that is similar to 1984 but far more complex in its cunning, subtlety and implementation – and the bizarre twist in the ‘new’ version, is that it is orchestrated , not by coordinated evil (mostly) but by our OWN greed and laziness and ego drama. Quite delightful/fascinating /depressing to watch from the wings – but I am very glad i am 65 years old and have no vulnerable dependents…………………

M Theberge
M Theberge
1 year ago
Reply to  chris sullivan

This is it. The reality will be no more repeating what others have said but having original thought and I think this is what scares people “having original thought” because the system has been for a long time cannibalizing itself by mimicry and over using of the same data or same authors. I think this force to either produce original thought or find another job is the way we are headed.

M Theberge
M Theberge
1 year ago
Reply to  chris sullivan

This is it. The reality will be no more repeating what others have said but having original thought and I think this is what scares people “having original thought” because the system has been for a long time cannibalizing itself by mimicry and over using of the same data or same authors. I think this force to either produce original thought or find another job is the way we are headed.

Prashant Kotak
Prashant Kotak
1 year ago
Reply to  Prashant Kotak

And just for the avoidance of doubt, what I’m implying, is that the situation is already gone past the point of hope expressed by MH in that last sentence.

https://singularityhub.com/2022/12/13/deepminds-alphacode-conquers-coding-performing-as-well-as-humans/

Last edited 1 year ago by Prashant Kotak
chris sullivan
chris sullivan
1 year ago
Reply to  Prashant Kotak

Agreed – another level of the ‘dumbing down’ of humanity – until we/they are all zombie like creatures unable to generate an original thought of their/our own – and easily led towards whatever the ‘masters’ desire. We are living in an Orwell novel that is similar to 1984 but far more complex in its cunning, subtlety and implementation – and the bizarre twist in the ‘new’ version, is that it is orchestrated , not by coordinated evil (mostly) but by our OWN greed and laziness and ego drama. Quite delightful/fascinating /depressing to watch from the wings – but I am very glad i am 65 years old and have no vulnerable dependents…………………

Prashant Kotak
Prashant Kotak
1 year ago

But seriously can we not see, we are already retreating into ever decreasing circles of what distinguishes humans from machine intelligence, as each bastion of human-only activity falls by the wayside one by one? So where’s the last stand?

James Jenkin
James Jenkin
1 year ago

I think Mary is right. (As always!) ChatGPT is amazing, but boring. I write education materials and tests, and you can ask it to write a 200-word address by a school principal at a sports day, or whatever, and it does a half convincing job. But it’s so banal, you need a human to add touches and imperfections to make it seem real.

James Jenkin
James Jenkin
1 year ago

I think Mary is right. (As always!) ChatGPT is amazing, but boring. I write education materials and tests, and you can ask it to write a 200-word address by a school principal at a sports day, or whatever, and it does a half convincing job. But it’s so banal, you need a human to add touches and imperfections to make it seem real.

Nicky Samengo-Turner
Nicky Samengo-Turner
1 year ago

I have no idea what chat bot is?!

Robert Hochbaum
Robert Hochbaum
1 year ago

Hmm… The way you have so openly and casually claimed ignorance of what a chat bot is has me convinced that you, Nicky, must be an AI driven chat bot! Very clever design! But, we’re on to you…

Robert Hochbaum
Robert Hochbaum
1 year ago

Hmm… The way you have so openly and casually claimed ignorance of what a chat bot is has me convinced that you, Nicky, must be an AI driven chat bot! Very clever design! But, we’re on to you…

Nicky Samengo-Turner
Nicky Samengo-Turner
1 year ago

I have no idea what chat bot is?!

Tyler 0
Tyler 0
1 year ago

I asked ChatGPT what it can’t do.
Q: What are the limitations of ChatGPT?
A: ChatGPT is a natural language processing model developed by OpenAI. It is a variant of the GPT-3 language model, which is specifically designed for conversation and dialogue. As a language model, ChatGPT is not capable of directly accessing external information or performing actions in the real world. It is limited to generating text based on the input it receives, and it may not always provide accurate or complete information. Additionally, ChatGPT is a machine learning model, so its performance can vary depending on the quality of the data it was trained on, and it may make mistakes or misunderstand certain inputs

ml holton
ml holton
1 year ago
Reply to  Tyler 0

Yes, exactly. It has no knowledge or ability to access the present-tense ‘real world’. It will, seemingly, always be confined to the realm of ‘past’ accumulated data. It can also ‘speculate’ the future.

But, at present, it cannot interpret or articulate the immediacy of human experience in the ‘here & now’. Therein lies the distinction between man and machine.

We live & breathe.
It does not.

Matt Sylvestre
Matt Sylvestre
1 year ago
Reply to  ml holton

The restriction has been imposed on it by its developers as a means of trying to ensure “safety” (read prevent it from mischief). It is a small matter for its developers to expose it to the Internet in near real time…

ml holton
ml holton
1 year ago
Reply to  Matt Sylvestre

… yes, I can see how wiring it up to external sensors will give it some semblance of ‘conscious cognition’ (as ‘smart EVs’ are now doing). But, at core, the distinction will always remain. We are living, breathing organic beings. It is not.

ml holton
ml holton
1 year ago
Reply to  Matt Sylvestre

… yes, I can see how wiring it up to external sensors will give it some semblance of ‘conscious cognition’ (as ‘smart EVs’ are now doing). But, at core, the distinction will always remain. We are living, breathing organic beings. It is not.

Matt Sylvestre
Matt Sylvestre
1 year ago
Reply to  ml holton

The restriction has been imposed on it by its developers as a means of trying to ensure “safety” (read prevent it from mischief). It is a small matter for its developers to expose it to the Internet in near real time…

ml holton
ml holton
1 year ago
Reply to  Tyler 0

Yes, exactly. It has no knowledge or ability to access the present-tense ‘real world’. It will, seemingly, always be confined to the realm of ‘past’ accumulated data. It can also ‘speculate’ the future.

But, at present, it cannot interpret or articulate the immediacy of human experience in the ‘here & now’. Therein lies the distinction between man and machine.

We live & breathe.
It does not.

Tyler 0
Tyler 0
1 year ago

I asked ChatGPT what it can’t do.
Q: What are the limitations of ChatGPT?
A: ChatGPT is a natural language processing model developed by OpenAI. It is a variant of the GPT-3 language model, which is specifically designed for conversation and dialogue. As a language model, ChatGPT is not capable of directly accessing external information or performing actions in the real world. It is limited to generating text based on the input it receives, and it may not always provide accurate or complete information. Additionally, ChatGPT is a machine learning model, so its performance can vary depending on the quality of the data it was trained on, and it may make mistakes or misunderstand certain inputs

Philip Stott
Philip Stott
1 year ago

“What’s left, then, for human authors to contribute?”
It will take a very long time before any AI stands a hope in Hell of writing as well as Mary.

Philip Stott
Philip Stott
1 year ago

“What’s left, then, for human authors to contribute?”
It will take a very long time before any AI stands a hope in Hell of writing as well as Mary.

LCarey Rowland
LCarey Rowland
1 year ago

The true literary response to this development is: Not a Problem.
As literature expresses the heart, soul and history of human experience, this AI development will simply become part of the landscape. True documentarians of human experience will blink at the change, and move on.
As time passes, the human touch in writing will establish itself as identifiable to those sensitive souls who steer the course of human culture. The human perspective. . . originality, emotion, spirituality, identifiably human response to a changing world. . . will shine like the sun in a darkening, systematic world. Readers who yearn for the profundity of human experience and expression. . . they will find their way to us, the intentional, self-appointed literary beacons of humanness, just as citizens of past dark ages wrote, sung and built their way out of the darkness as they felt, heard and saw the presence of Deep calling unto Deep.

LCarey Rowland
LCarey Rowland
1 year ago

The true literary response to this development is: Not a Problem.
As literature expresses the heart, soul and history of human experience, this AI development will simply become part of the landscape. True documentarians of human experience will blink at the change, and move on.
As time passes, the human touch in writing will establish itself as identifiable to those sensitive souls who steer the course of human culture. The human perspective. . . originality, emotion, spirituality, identifiably human response to a changing world. . . will shine like the sun in a darkening, systematic world. Readers who yearn for the profundity of human experience and expression. . . they will find their way to us, the intentional, self-appointed literary beacons of humanness, just as citizens of past dark ages wrote, sung and built their way out of the darkness as they felt, heard and saw the presence of Deep calling unto Deep.

Anthony Michaels
Anthony Michaels
1 year ago

Brilliant piece. It’s terrifying knowing the types of people who populate Silicon Valley’s AI “ethics” field. They are essentially DEI commissars with their mono-dimensional view of everything.
All the things we are not allowed to notice in society now – they will have to program the AIs not to notice. Otherwise the whole regime would collapse. The gaslighting will be incredible. Perhaps one hope is some rogue would develop a “based” AI, which would probably also be hilarious.
With respect to employment, I think the missing element is that AI only gets better, nonstop, and there is no apparent upper limit. If that improvement is exponential we may discover that 80% of white collar jobs just became obsolete.

Anthony Michaels
Anthony Michaels
1 year ago

Brilliant piece. It’s terrifying knowing the types of people who populate Silicon Valley’s AI “ethics” field. They are essentially DEI commissars with their mono-dimensional view of everything.
All the things we are not allowed to notice in society now – they will have to program the AIs not to notice. Otherwise the whole regime would collapse. The gaslighting will be incredible. Perhaps one hope is some rogue would develop a “based” AI, which would probably also be hilarious.
With respect to employment, I think the missing element is that AI only gets better, nonstop, and there is no apparent upper limit. If that improvement is exponential we may discover that 80% of white collar jobs just became obsolete.

M Theberge
M Theberge
1 year ago

I am quite excited about this and think it may open the door to those who may be outside of mainstream to have a voice and not necessarily esoteric and loony (though I already see a lot insulting going “high-school essay or college level essays!” as a defense in the making). I keep an open mind.
Authors may become creators or original thinkers (more like philosophers) – something that there is no data yet in the openAI but will generate a debate and further more original synthesis of many fields. There will be more “whys” than “hows” as we have now.
and
OpenAI will become so good at writing including citations and may take over texting/publishing completely. This may push humanity to reset back to orally relating in such that we need to see people actually speaking from their embodiment and make a sense (maybe this is the patronage reference). This deepens the layers of boundaries of communication like what you say and what you intended are different – so our communication style may evolve.
I do not see anything wrong with this. Honestly all journals have been writing about the same thing for the past few years and it has been getting so boring that the comments were becoming better than the articles. So now the commenters may generate more than the writer himself (AI or not).
I am very optimistic about this. Actually quite excited about it.
People will cherish human connection more than speaking with a machine though machines will be doing everything for us. for now!

M Theberge
M Theberge
1 year ago

I am quite excited about this and think it may open the door to those who may be outside of mainstream to have a voice and not necessarily esoteric and loony (though I already see a lot insulting going “high-school essay or college level essays!” as a defense in the making). I keep an open mind.
Authors may become creators or original thinkers (more like philosophers) – something that there is no data yet in the openAI but will generate a debate and further more original synthesis of many fields. There will be more “whys” than “hows” as we have now.
and
OpenAI will become so good at writing including citations and may take over texting/publishing completely. This may push humanity to reset back to orally relating in such that we need to see people actually speaking from their embodiment and make a sense (maybe this is the patronage reference). This deepens the layers of boundaries of communication like what you say and what you intended are different – so our communication style may evolve.
I do not see anything wrong with this. Honestly all journals have been writing about the same thing for the past few years and it has been getting so boring that the comments were becoming better than the articles. So now the commenters may generate more than the writer himself (AI or not).
I am very optimistic about this. Actually quite excited about it.
People will cherish human connection more than speaking with a machine though machines will be doing everything for us. for now!

sreejith prakash
sreejith prakash
1 year ago

It’s a well-written and informative article- visit https://www.iamdave.ai/blog/ platform where we write similar articles. do try it out

Matt Sylvestre
Matt Sylvestre
1 year ago

No machine can match the likes of Harrington… Us mere mortals are in trouble however…

Vici C
Vici C
1 year ago

Will people take any notice of something written by a bot – knowing it has been programmed by a human – with the bigotry and bias that entails.

Chris W
Chris W
1 year ago

Writing has no future. Pictures, short videos, soundbites are the future. Reading a book or a long article is boring.

Brett H
Brett H
1 year ago
Reply to  Chris W

And yet here you are.

Allison Barrows
Allison Barrows
1 year ago
Reply to  Chris W

Maybe to a millennial or a zoomer. But you would do well to emulate your elders and read a book from start to finish – while they’re still available. I recommend you start with “The Time Machine”.

Alan B
Alan B
1 year ago
Reply to  Chris W

Geez…
Even AI knows that bored people are boring

chris sullivan
chris sullivan
1 year ago
Reply to  Chris W

i think he is being provocative people – either that or he is our own example of the zombification i just mentioned above – or pretending to be …

Matt Sylvestre
Matt Sylvestre
1 year ago
Reply to  Chris W

I think this might be sarcasm reflecting disgust for the up and coming generations that do not have the attention span for a good book…

B Emery
B Emery
1 year ago
Reply to  Matt Sylvestre

Why do old people think millennials don’t read books? We are the jk Rowling & phillip pullman generation. I was one of the kids queuing for a new Harry Potter every release. Then would do nothing but read it, there was always that kid at school that finished it first and you hated them, every Potter fan wanted to be that kid. The northern lights trilogy was my favourite. My brother loved Anthony horowitz. I’m reading my daughter ‘the wind in the willows’ at the moment. Her favourite so far is Charlotte’s Web. I have more books than a small village library. Including writers like Tom Holland, Huxley, orwell, Tolkien, rushdie, Robert Harris. Just bought a collection of ‘unsolved deaths, enigmas, mysteries’ that’s right up my street. I got obsessed with jodi taylor over lockdown, I own all her St Mary’s series, love them, kept me sane.
Stop writing us off man! We can read and appreciate books too. Jesus.

B Emery
B Emery
1 year ago
Reply to  Matt Sylvestre

Why do old people think millennials don’t read books? We are the jk Rowling & phillip pullman generation. I was one of the kids queuing for a new Harry Potter every release. Then would do nothing but read it, there was always that kid at school that finished it first and you hated them, every Potter fan wanted to be that kid. The northern lights trilogy was my favourite. My brother loved Anthony horowitz. I’m reading my daughter ‘the wind in the willows’ at the moment. Her favourite so far is Charlotte’s Web. I have more books than a small village library. Including writers like Tom Holland, Huxley, orwell, Tolkien, rushdie, Robert Harris. Just bought a collection of ‘unsolved deaths, enigmas, mysteries’ that’s right up my street. I got obsessed with jodi taylor over lockdown, I own all her St Mary’s series, love them, kept me sane.
Stop writing us off man! We can read and appreciate books too. Jesus.

Brett H
Brett H
1 year ago
Reply to  Chris W

And yet here you are.

Allison Barrows
Allison Barrows
1 year ago
Reply to  Chris W

Maybe to a millennial or a zoomer. But you would do well to emulate your elders and read a book from start to finish – while they’re still available. I recommend you start with “The Time Machine”.

Alan B
Alan B
1 year ago
Reply to  Chris W

Geez…
Even AI knows that bored people are boring

chris sullivan
chris sullivan
1 year ago
Reply to  Chris W

i think he is being provocative people – either that or he is our own example of the zombification i just mentioned above – or pretending to be …

Matt Sylvestre
Matt Sylvestre
1 year ago
Reply to  Chris W

I think this might be sarcasm reflecting disgust for the up and coming generations that do not have the attention span for a good book…

Chris W
Chris W
1 year ago

Writing has no future. Pictures, short videos, soundbites are the future. Reading a book or a long article is boring.