A whole new ballgame in cheating. Introducing ChatGPT

[marshwiggle]

They have just announced "improvements" to ChatGPT. 

https://www.nytimes.com/2023/03/14/technology/openai-gpt4-chatgpt.html


Do these techno-libertarians think much about the potential human costs of their work?

When has technological advancement ever taken into account potential human costs (except to trumpet improvements in efficiency, etc.)?

It's always about "what can we do?" not "what should we do?"

That's often because the ultimate bad outcome of some technological advancement isn't in sight as the many incremental developments along the way happen. Most of the problems are due to unintended consequences, rather than desired goals.
So at each stage of incremental development, there's no obvious moral reason to avoid getting over the current technological hurdle.

Although it's worth pointing out that the historical track record of correctly predicting potential harms is not great. People tend to worry about the wrong things. There's no reason anyone should trust my predictions, but it does seem to me that we often tend to fixate on the dangers of Artificial Intelligence and don't pay enough attention to the ways in which technologies enable and amplify the bad behavior and evil of old fashioned humans. Nobody spent much time worrying about Facebook and YouTube, because they were just platforms for people to display their creativity without traditional gatekeepers. It was Deep Blue and the fact that computers could beat the best humans at chess that got all the attention. We're always afraid of being replaced or undermined by alien technology, but all the bad stuff keeps being the result of people sitting around typing on their phones.

And people choose convenience and comfort over all kinds of other values. The robot apocalypse won't happen because the machines "take over"; it will happen because people invite them to do all of the stuff they don't want to, including (and especially) thinking. (Think "Wall-E rather than "Terminator".)

[Larimar]

They have just announced "improvements" to ChatGPT. 

https://www.nytimes.com/2023/03/14/technology/openai-gpt4-chatgpt.html


Do these techno-libertarians think much about the potential human costs of their work?

Progress is inevitable.  You will be assimilated.  Resistance is futile...

Yeah, really.
As ChatGPT and similar bots "improve", I worry that people will devalue the skill and art of writing more than they already do. It's been a very long time since writing has been a profession one can make a living at, except for journalism. People will wonder, if they don't already, who needs to do or know anything about writing when bots can do it for us? In another thread, someone said they feared that teaching English is a dying profession. What happens when no one but the bots knows how to write anymore? Really, I hope I am dead before we have a world without novels, memoirs, plays, short fiction, and poetry. Or one in which the bots write them and they have no soul.

And people choose convenience and comfort over all kinds of other values. The robot apocalypse won't happen because the machines "take over"; it will happen because people invite them to do all of the stuff they don't want to, including (and especially) thinking. (Think "Wall-E rather than "Terminator".)

I agree with this.


Larimar

[Diogenes]

Keynes thought technology would mean we all could stop working and live happy lives with all our needs met by now.
But don't worry, administration will make sure any time saved from AI will be filled with meetings that could have been AI generated emails.
Any financial benefit will still go to the 1%

[Wahoo Redux]

Progress is inevitable.  You will be assimilated.  Resistance is futile...

I worry that people will devalue the skill and art of writing more than they already do. It's been a very long time since writing has been a profession one can make a living at, except for journalism. People will wonder, if they don't already, who needs to do or know anything about writing when bots can do it for us? In another thread, someone said they feared that teaching English is a dying profession. What happens when no one but the bots knows how to write anymore? Really, I hope I am dead before we have a world without novels, memoirs, plays, short fiction, and poetry. Or one in which the bots write them and they have no soul.

Larimar

I'm the one who worried that teaching English is a slowly dying profession----and it is an overstatement.  But English as a subject, discipline, and profession is under constant attack, often by fellow intellectuals; I'm surprised that anyone would object to mechanistic writing.  Perhaps it is time for the bots to take over.  I'm not sure how exactly it would work, but everyone would have a lot more free time if we could just upload our data and ideas into an app and have it spit out a competent, perfectly-proofread essay or report.  Researching would be easier, writing would be easier, grading would be A LOT easier.

As a former writing center coordinator, I got really, really tired of professors chiding us for their students' grammar mistakes after a half-hour session with an undergraduate tutor. Somehow these people thought we had the magic ability to perform grammar osmosis.  Professors, particularly in business and engineering, were constantly carping that "we weren't teaching students how to write," and at one point someone in the college suggested doing away with freshman comp because it obviously wasn't working (it did not pass).  What if students let an app do all that irksome grammar and syntax for them?

Why should we resist this?

And, honestly, there would be nothing stopping us from writing poems and memoirs even though a computer program might do it much better.  We could kick English and math instruction entirely to the curb.  Sometimes I think medicine would be much better if performed by a bot----imagine how much money we would save.  Imagine kicking insurance to the curb!  Maybe we could kick even more disciplines to the curb and let business take over whatever education we have left?

[lightning]

Keynes thought technology would mean we all could stop working and live happy lives with all our needs met by now.
But don't worry, administration will make sure any time saved from AI will be filled with meetings that could have been AI generated emails.
Any financial benefit will still go to the 1%

Yup. And, any staff positions made redundant by technology and AI, won't result in the re-distribution of budget savings towards the strengthening of remaining programs and addressing other pressing existing needs. Savings will undoubtedly go to something stupid or disappear into the ether. At my university, savings from jobs replaced by AI & tech goes to creating new admin positions, even if new bureaucratic processes have to be created. Most recently, we have a whole new set of compliance, budget oversight, and institutional effectiveness processes that we never had to deal with before, which were created, managed, and implemented, of course, by a new set of administrators that never existed before. Some of the people in the office literally make reports of reports of reports.

[Kron3007]

They have just announced "improvements" to ChatGPT. 

https://www.nytimes.com/2023/03/14/technology/openai-gpt4-chatgpt.html


Do these techno-libertarians think much about the potential human costs of their work?

Progress is inevitable.  You will be assimilated.  Resistance is futile...

Yeah, really.
As ChatGPT and similar bots "improve", I worry that people will devalue the skill and art of writing more than they already do. It's been a very long time since writing has been a profession one can make a living at, except for journalism. People will wonder, if they don't already, who needs to do or know anything about writing when bots can do it for us? In another thread, someone said they feared that teaching English is a dying profession. What happens when no one but the bots knows how to write anymore? Really, I hope I am dead before we have a world without novels, memoirs, plays, short fiction, and poetry. Or one in which the bots write them and they have no soul.

And people choose convenience and comfort over all kinds of other values. The robot apocalypse won't happen because the machines "take over"; it will happen because people invite them to do all of the stuff they don't want to, including (and especially) thinking. (Think "Wall-E rather than "Terminator".)

I agree with this.


Larimar

I imagine this is a similar argument that was made when calculators started to hit the scene.  What happens when no on can do math anymore.  The answer is, we use calculators.

I think you are right that there will be less and less paid writing gigs in the future.  I also agree that our ability to write will suffer, but the quality of writing will likely improve overall.  Even today, chat GPT is probably a better author than the average american (not a high bar). 

So yes, it could spell the end, or at least the decline, of paid writing gigs etc., But it is not going anywhere and will only improve from here. 

As I said, resistance is futile.  All we can do is learn how to adapt and use it.

[Kron3007]

Progress is inevitable.  You will be assimilated.  Resistance is futile...

I worry that people will devalue the skill and art of writing more than they already do. It's been a very long time since writing has been a profession one can make a living at, except for journalism. People will wonder, if they don't already, who needs to do or know anything about writing when bots can do it for us? In another thread, someone said they feared that teaching English is a dying profession. What happens when no one but the bots knows how to write anymore? Really, I hope I am dead before we have a world without novels, memoirs, plays, short fiction, and poetry. Or one in which the bots write them and they have no soul.

Larimar

And, honestly, there would be nothing stopping us from writing poems and memoirs even though a computer program might do it much better.  We could kick English and math instruction entirely to the curb.  Sometimes I think medicine would be much better if performed by a bot----imagine how much money we would save.  Imagine kicking insurance to the curb!  Maybe we could kick even more disciplines to the curb and let business take over whatever education we have left?

AI is most definitely coming for medicine.  I saw some talks on it a couple years ago and I believe it was already better at identifying cancer in imaging systems than humans.  I believe they are already using it, but the conclusions still need human verification.  It is only a matter of time, and to be honestcomputers will be better at many medical tasks than humans.

[marshwiggle]

I'm the one who worried that teaching English is a slowly dying profession----and it is an overstatement.  But English as a subject, discipline, and profession is under constant attack, often by fellow intellectuals; I'm surprised that anyone would object to mechanistic writing.  Perhaps it is time for the bots to take over.  I'm not sure how exactly it would work, but everyone would have a lot more free time if we could just upload our data and ideas into an app and have it spit out a competent, perfectly-proofread essay or report.  Researching would be easier, writing would be easier, grading would be A LOT easier.

As a former writing center coordinator, I got really, really tired of professors chiding us for their students' grammar mistakes after a half-hour session with an undergraduate tutor. Somehow these people thought we had the magic ability to perform grammar osmosis.  Professors, particularly in business and engineering, were constantly carping that "we weren't teaching students how to write," and at one point someone in the college suggested doing away with freshman comp because it obviously wasn't working (it did not pass).  What if students let an app do all that irksome grammar and syntax for them?

Remember when spell-checkers were a novelty? Who complains about students using them now? (Heck, who doesn't TELL students to spell-check their work?)

Remember when grammar checkers were a novelty? Most complaints were that the suggestions weren't very good. 

If ChatGPT (or its progeny) sits in the background like spelling and grammar tools that already exist, but suggests more holistic changes, I don't think it will be long until it becomes an expectation. (One warning to students who think this will solve their problems: Right now, grading students' content can be difficult when their writing is bad. How much easier will it be to see bad arguments and logic when they are expressed clearly and concisely?)

[Istiblennius]

We're working on updating our academic honesty statement for lab writeups to accommodate AI... One of the key messages we are trying to convey is that for describing a specific experiment, AI kind of sucks and probably will earn students a poor grade.

If any of my colleagues are on this thread, I'm outing myself, but oh well. My moniker is probably already outing me to those who know me well :)

A note on the use of Artificial Intelligence Writing tools - some students have explored using AI tools to develop their lab writeups and have discovered that the AI tools are not particularly effective at clearly articulating the work they did in the unique circumstances of their experiments. Also, the AI tools do not tend to cite effectively, which has led to students earning zeros for plagiarism. If you choose to use an AI tool to help you draft your writeup, you should consider that a starting idea generation or brainstorming tool and a first step in building your writeup. Make sure that you go through and adjust to your own words so you can very clearly describe what you did and how you understand the concepts. You should also ensure you are providing effective references for your writing.

[Caracal]

I'm the one who worried that teaching English is a slowly dying profession----and it is an overstatement.  But English as a subject, discipline, and profession is under constant attack, often by fellow intellectuals; I'm surprised that anyone would object to mechanistic writing.  Perhaps it is time for the bots to take over.  I'm not sure how exactly it would work, but everyone would have a lot more free time if we could just upload our data and ideas into an app and have it spit out a competent, perfectly-proofread essay or report.  Researching would be easier, writing would be easier, grading would be A LOT easier.

As a former writing center coordinator, I got really, really tired of professors chiding us for their students' grammar mistakes after a half-hour session with an undergraduate tutor. Somehow these people thought we had the magic ability to perform grammar osmosis.  Professors, particularly in business and engineering, were constantly carping that "we weren't teaching students how to write," and at one point someone in the college suggested doing away with freshman comp because it obviously wasn't working (it did not pass).  What if students let an app do all that irksome grammar and syntax for them?

Remember when spell-checkers were a novelty? Who complains about students using them now? (Heck, who doesn't TELL students to spell-check their work?)

Remember when grammar checkers were a novelty? Most complaints were that the suggestions weren't very good. 

If ChatGPT (or its progeny) sits in the background like spelling and grammar tools that already exist, but suggests more holistic changes, I don't think it will be long until it becomes an expectation. (One warning to students who think this will solve their problems: Right now, grading students' content can be difficult when their writing is bad. How much easier will it be to see bad arguments and logic when they are expressed clearly and concisely?)

I don't think that's likely to happen. The mechanics of writing can't actually be separated out from content. Revising isn't just about fixing up some messy prose, it's about figuring out what you are actually trying to say. ChatGPT can't make an argument on it's own-it sort of specializes in non arguments, and if it does create one that's only because it finds stuff to copy. You don't want to be taking advice on revising your work from something that can't make arguments and can't figure out what they mean.

I get the impression some people seem very impressed with the fact that CHATGPT can produce competent sounding prose, but it's actually similar to a lot of the stuff I get from students. It is reasonably polished, it sounds like it is saying something, but it doesn't go anywhere, it doesn't make any argument and it has no real point.

[Caracal]

Progress is inevitable.  You will be assimilated.  Resistance is futile...

I worry that people will devalue the skill and art of writing more than they already do. It's been a very long time since writing has been a profession one can make a living at, except for journalism. People will wonder, if they don't already, who needs to do or know anything about writing when bots can do it for us? In another thread, someone said they feared that teaching English is a dying profession. What happens when no one but the bots knows how to write anymore? Really, I hope I am dead before we have a world without novels, memoirs, plays, short fiction, and poetry. Or one in which the bots write them and they have no soul.

Larimar

And, honestly, there would be nothing stopping us from writing poems and memoirs even though a computer program might do it much better.  We could kick English and math instruction entirely to the curb.  Sometimes I think medicine would be much better if performed by a bot----imagine how much money we would save.  Imagine kicking insurance to the curb!  Maybe we could kick even more disciplines to the curb and let business take over whatever education we have left?

AI is most definitely coming for medicine.  I saw some talks on it a couple years ago and I believe it was already better at identifying cancer in imaging systems than humans.  I believe they are already using it, but the conclusions still need human verification.  It is only a matter of time, and to be honestcomputers will be better at many medical tasks than humans.

It is more and more widely used, but it isn't doing away with humans. It's good at identifying and flagging common findings. That can be really useful for radiologists because it saves them time and it also can reduce mistakes by finding things they might have missed. However, you still need a person to go actually look at the thing it's flagged and evaluate them. It also isn't particularly good at identifying rarer conditions.

As an example,I use the census a fair amount for my work so I've gotten pretty good at navigating ancestry. They've long had a thing where they give you lists of suggestions in the sidebar of other records that might be relevant. It used to be I ignored them because they were rarely useful. A few years ago, I realized that the system had gotten much better and now I make use of it a lot. It identifies  documents that would have taken me lots of time and searching to find, and some things that I might never have found. It's quite useful. However, it does produce a fair number of false positives, some of which look promising until you actually go look at the document and compare it to the information you have. I also can't rely on it to find everything. It's pretty good when everything makes sense, but when something is a little off or confusing, it misses stuff that I can track down. I suspect AI with imaging is pretty similar. You do need the human with expertise and that's not likely to change.

[Wahoo Redux]

I don' t think this has been posted yet:

Atlantic: The College Essay Is Dead: Nobody is prepared for how AI will transform academia.

The AI essay referenced in the article

[Caracal]

I don' t think this has been posted yet:

Atlantic: The College Essay Is Dead: Nobody is prepared for how AI will transform academia.

The AI essay referenced in the article

Wait, that's "graduate level?" Also, man is that a terrible prompt. And this, "Ultimately, we need to understand the interactions among learning styles and environmental and personal factors, and how these shape how we learn and the kinds of learning we experience" is not an argument. How do we need to understand them? What do we need to understand? It's so vague that there's no real counterargument possible. What would you say "we don't need to understand those things?"

[marshwiggle]

I don' t think this has been posted yet:

Atlantic: The College Essay Is Dead: Nobody is prepared for how AI will transform academia.

The AI essay referenced in the article

Wait, that's "graduate level?" Also, man is that a terrible prompt. And this, "Ultimately, we need to understand the interactions among learning styles and environmental and personal factors, and how these shape how we learn and the kinds of learning we experience" is not an argument. How do we need to understand them? What do we need to understand? It's so vague that there's no real counterargument possible. What would you say "we don't need to understand those things?"

Postmodernism has probably lowered peoples' standards for what counts as an"argument" that all kinds of drivel will pass.

[waterboy]

I just tried Google's Barb - that was pretty slick, though when I asked about myself (couldn't resist), it got my PhD wrong. Most everything else was pretty spot on.