August 5, 2025 • 12 mins
An artificial intelligence expert says the tech will always need human input for marking school exams.
Education Minister Erica Stanford says AI's already partly used to grade assessments, and it could do most of them by 2028.
It's alongside Government plans to replace NCEA with entirely new qualifications.
New Zealand AI Founder, Justin Flitter, told Kerre Woodham the tool should supplement a teacher's process, not become it.
He says the teacher should apply the AI assessment to their lived experience and knowledge of the student.
LISTEN ABOVE
See omnystudio.com/listener for privacy information.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:06):
You're listening to the carry Wood and Morning's podcast from
News Talks, he'd be as.
Speaker 2 (00:12):
We've been discussing.
Speaker 3 (00:13):
Education.
Speaker 2 (00:13):
Minister Erica Stanford says artificial intelligence will be used for
marking in the coming years. She says she also expects
teachers will use it to reduce the workload of lesson
planning there affairs. However, that artificial intelligence won't have cultural
considerations when marking students work, which I think the university
(00:38):
lecturer who texted in probably explained that better than cultural considerations.
I teach at UNI. No way would I use AI
to mark when I mark them learning about my students,
identifying weaknesses and strengths. It influences my subsequent teaching. So
perhaps that explains some teachers concerns about using AI. Founder
(01:01):
of New Zealand AI Justin Flitter joins me now and
a very very good according to you.
Speaker 4 (01:07):
Good morning, nice to be here, nice to chat.
Speaker 2 (01:11):
Thank you now. I mean AI has already been used
in schools to mark. This is just an extension of it.
But the concerns around cultural considerations, the concerns that the
university teacher had about grading that it doesn't if you're
not doing it you don't see what the kids aren't
(01:32):
picking up. Is that something that AI will be able
to do in the future or can do now.
Speaker 5 (01:40):
Again, I think it's around how we construct these AI
systems and how we are feeding in humans.
Speaker 4 (01:46):
And the loop with it. It's not that.
Speaker 5 (01:48):
These AI systems should be grading and reporting back to
students or funomously. They should be a part of a
teacher's process, not the process. So the teachers should be
taking on board the assessment that the AIS tools made
and then bringing in their own lived experience and knowledge
(02:09):
of the students and other things to you know, to
allow AI to augment that process so that they know
it shortens up that mine.
Speaker 4 (02:17):
But it's not the bee or and end of a no.
Speaker 2 (02:20):
And I think, yeah, I think it's probably that people
fear that it is going to take over the job
rather than just to be used as a tool to
do the job better.
Speaker 5 (02:32):
And it's at the moment it's you know, it certainly
has the potential to take on much more of the work,
but I think at the moment in terms of you know,
walk them and teaching these tools and making sure that
we're using them appropriately, and you know, we don't create
more distrust. I mean, New Zealanders has got by father
(02:55):
the lowest levels of trust and AI systems and almost
anywhere in the world. So it's very important that you
know we're all on this journey together and that we're
taking small steps but in the right direction in terms
of building trust and confidence to use these tools to
do time consuming and repetitive work that frees us up
(03:16):
to do higher value work and spending more time face
to face with students.
Speaker 2 (03:21):
Well, that's what I would have thought that that teachers
would be able to build on relationships with the students
in their class to you know, to be able to
inspire kids, to be able to work out what their
passion is in terms of subjects, to be able to
extend the kids and understand that they need more direction,
(03:41):
more more stimulus.
Speaker 5 (03:43):
And looking at those different data points and looking at
those different kind of indicators that we have from you
performance and activity around child.
Speaker 4 (03:52):
We're using the tools.
Speaker 5 (03:53):
To to optimize their learning program to create courses and
webinars and workshops and class programs that are more tailored
and for for the students that they have. And that's
a time consuming task to get more specific, right, Otherwise
(04:14):
it just becomes a generic for everyone. So, you know,
thinking about past course planning and session planning, you know,
there are plenty of great tools out there now that
can automate that process, and that should be throwing up
a ton of time that teachers would have been spending,
you know, after hours at home on the weekends, you know,
(04:37):
doing all that type of stuff.
Speaker 3 (04:38):
And h yeah, as ai that at the does it
have the ability, given the human input, to be able
to grade creative writing?
Speaker 4 (04:54):
Sure that you've got to define what what what is
the definition of creative in us? And and again what
are the what's the criteria that the teachers or the
assesses are looking for to determine what creative means in
the context of that exercise. So that has to be codified,
It has to be written as part of a prompt
(05:16):
or as part of that assessment criteria they would need
to be given. You know, it's no examples of what
good looks up in order to be able to judge
whether or not a student's piece of work meets those
that seasoned criteria. That's awfully pomplex, because.
Speaker 2 (05:38):
So it all depends on the on the human input
before it can grade correctly. Yeah, we're responsible.
Speaker 5 (05:45):
We've got to we've got to teach these assistance we
and we want them to work.
Speaker 4 (05:51):
And you know they don't, they don't just come out
of the box.
Speaker 5 (05:54):
They've got to be changed, just like a teacher, just
like anyone who is going to be assessing that work.
Speaker 4 (05:59):
If you're going to be looking at.
Speaker 5 (06:00):
A reading a student's essay, you know, how do you
know what what good looks like?
Speaker 4 (06:05):
What the standard that needs to be met? And so
it needs to be trained on you know, years and
years of of students work, so understands what level to
expect and then how to be able to you know,
provide feedback or you know, it's an attendment. And this
all has to be transparent, right, we just can't be
a black box. We need to know what data these
(06:27):
tools are trained on. That is determining what you know,
how are they breeding the work and what's the what's
the data set?
Speaker 5 (06:35):
And of course that opens up opportunity for bias and
and all sorts of different things.
Speaker 6 (06:40):
And so it's not it's it's it's a it's it's
it's a complex it's a complex space. It needs you know,
it's good good thoughts.
Speaker 2 (06:51):
Rand it's interesting when you say we don't trust it.
We don't trust humans much either we don't trust the
media or politicians or any It seems so late.
Speaker 4 (07:02):
New Zealand is trust trust humans a lot more than
they trust I mean, it's been count countless reports that
have been troublished over the last year and museums before,
you know, forty first out of forty seven or something,
light at the bottom of a list of countries in
terms of their the level of trust and AI systems.
(07:22):
And you know, I'm unsure how how has managed to
get to this point, but you know every.
Speaker 5 (07:30):
Day where we're trying to rebuild that and and and
show people, you know, what's what's possible today and how
they can be utilizing and leveraging these tools basically and responsibly.
Speaker 2 (07:46):
The only I know that a number of employers are
using AI to sift through applicants and and job seekers
are using AI to judge up their cvs and cover
letters and things. How how do we get to the
point that I just worry that when it comes to
(08:07):
sifting through so you find the appropriate applicants, sometimes people
who aren't who don't have the right background or qualifications
for the job. You just meet them and you know
by the cut of their jib, by the sparkle of
their eyes and the bushiness of their tail, that they
are going to be a good fit for the company
that you that one on one meeting is so important
(08:30):
that you might miss out on that if you're if
you're just you know, if you use a screening process.
Speaker 4 (08:38):
Yeah, but I mean, well we've always had a screening process, right,
There's always some there's always some gates that you have
to get through, and now you have to get through
the AI gates before you get to the human gates,
before you get to meets human. Interesting, Yeah, so you know, yeah,
it's and then you've got the whole generic as people
(09:02):
you know are using checchipoutine or right there a letter
and they're not using it to write in their style
or tone of voice, and you can get a very bland,
you know, very bland cover letter and CV, which means
that everything can you know, from a recruiter or a
hiring manager can all look rather the same, sane And
(09:23):
so you know, thinking about the skills that employers are
looking for, and you know, if you are going to
be using AI tools, how do you use them to
stand out. Yeah.
Speaker 5 (09:35):
And you know, and we know from the recent Cultivate
AI for Work report that employers are looking for people
who can demonstrate critical thinking, people who are not just
going to take the responses of a I tools at
face value. That they are going to blend learned experience,
(09:59):
domain knowledge and AI tools to supercharge the work that
they're So it's a part they're using it as a collaborator.
It's not you know, this is this is how I
this is how I do all my work. It's this
is how I do my work smarter.
Speaker 4 (10:15):
Yeah.
Speaker 2 (10:16):
Yeah, absolutely.
Speaker 4 (10:18):
I think where we're worry as people who you know,
are still saying, oh, we shouldn't be using this, or
if you're not allowed to use it, or you know,
it's cheating or something like that. It's like you're just
not clearing students for workforce that they're about to enter into.
A few are blocking them from using A tools. Yeah,
(10:39):
you know, it's like saying don't use the internet.
Speaker 2 (10:42):
Yeah, absolutely, or calculators. I think there are even some
people said okay, just.
Speaker 4 (10:48):
One final ilevant.
Speaker 2 (10:51):
A textasy is how best would a seventy six year
young retired woman learn about AI as a tool she
would like to.
Speaker 5 (10:58):
Use losee Oh what a fantastic love this.
Speaker 4 (11:06):
Experiment. To jump onto chat chipet. You can use the
free version or if you can, a for a plus
version and start experimenting, start asking the questions, look more,
look for a free course in your area. There might
be people that are dying free AI witch shops. Otherwise
(11:31):
there'll be YouTube videos and those sorts of things. So
to start experimenting, ask tea what you can do. There's
the right because you can't ask Excel what you could
do with itself, but you can ask chachipt how I
can use it in my daily life. And so start
(11:52):
having a conversation with it and start touched up and
curious around. Just start start start using it.
Speaker 5 (12:02):
It's the best way of doing it, asking Fisians and
being curious brilliant.
Speaker 2 (12:06):
Thank you so much for your time. Justin Flint, the
founder of New Zealand AI. There's also workshops that Justin's
company offers, some paid, some but there's some information on
there as well.
Speaker 1 (12:18):
For more from Kerry Wood and Mornings, listen live to
News Talks at b from nine am weekdays, or follow
the podcast on iHeartRadio.
You're listening to the carry Wood and Morning's podcast from
News Talks, he'd be as.
Speaker 2 (00:12):
We've been discussing.
Speaker 3 (00:13):
Education.
Speaker 2 (00:13):
Minister Erica Stanford says artificial intelligence will be used for
marking in the coming years. She says she also expects
teachers will use it to reduce the workload of lesson
planning there affairs. However, that artificial intelligence won't have cultural
considerations when marking students work, which I think the university
(00:38):
lecturer who texted in probably explained that better than cultural considerations.
I teach at UNI. No way would I use AI
to mark when I mark them learning about my students,
identifying weaknesses and strengths. It influences my subsequent teaching. So
perhaps that explains some teachers concerns about using AI. Founder
(01:01):
of New Zealand AI Justin Flitter joins me now and
a very very good according to you.
Speaker 4 (01:07):
Good morning, nice to be here, nice to chat.
Speaker 2 (01:11):
Thank you now. I mean AI has already been used
in schools to mark. This is just an extension of it.
But the concerns around cultural considerations, the concerns that the
university teacher had about grading that it doesn't if you're
not doing it you don't see what the kids aren't
(01:32):
picking up. Is that something that AI will be able
to do in the future or can do now.
Speaker 5 (01:40):
Again, I think it's around how we construct these AI
systems and how we are feeding in humans.
Speaker 4 (01:46):
And the loop with it. It's not that.
Speaker 5 (01:48):
These AI systems should be grading and reporting back to
students or funomously. They should be a part of a
teacher's process, not the process. So the teachers should be
taking on board the assessment that the AIS tools made
and then bringing in their own lived experience and knowledge
(02:09):
of the students and other things to you know, to
allow AI to augment that process so that they know
it shortens up that mine.
Speaker 4 (02:17):
But it's not the bee or and end of a no.
Speaker 2 (02:20):
And I think, yeah, I think it's probably that people
fear that it is going to take over the job
rather than just to be used as a tool to
do the job better.
Speaker 5 (02:32):
And it's at the moment it's you know, it certainly
has the potential to take on much more of the work,
but I think at the moment in terms of you know,
walk them and teaching these tools and making sure that
we're using them appropriately, and you know, we don't create
more distrust. I mean, New Zealanders has got by father
(02:55):
the lowest levels of trust and AI systems and almost
anywhere in the world. So it's very important that you
know we're all on this journey together and that we're
taking small steps but in the right direction in terms
of building trust and confidence to use these tools to
do time consuming and repetitive work that frees us up
(03:16):
to do higher value work and spending more time face
to face with students.
Speaker 2 (03:21):
Well, that's what I would have thought that that teachers
would be able to build on relationships with the students
in their class to you know, to be able to
inspire kids, to be able to work out what their
passion is in terms of subjects, to be able to
extend the kids and understand that they need more direction,
(03:41):
more more stimulus.
Speaker 5 (03:43):
And looking at those different data points and looking at
those different kind of indicators that we have from you
performance and activity around child.
Speaker 4 (03:52):
We're using the tools.
Speaker 5 (03:53):
To to optimize their learning program to create courses and
webinars and workshops and class programs that are more tailored
and for for the students that they have. And that's
a time consuming task to get more specific, right, Otherwise
(04:14):
it just becomes a generic for everyone. So, you know,
thinking about past course planning and session planning, you know,
there are plenty of great tools out there now that
can automate that process, and that should be throwing up
a ton of time that teachers would have been spending,
you know, after hours at home on the weekends, you know,
(04:37):
doing all that type of stuff.
Speaker 3 (04:38):
And h yeah, as ai that at the does it
have the ability, given the human input, to be able
to grade creative writing?
Speaker 4 (04:54):
Sure that you've got to define what what what is
the definition of creative in us? And and again what
are the what's the criteria that the teachers or the
assesses are looking for to determine what creative means in
the context of that exercise. So that has to be codified,
It has to be written as part of a prompt
(05:16):
or as part of that assessment criteria they would need
to be given. You know, it's no examples of what
good looks up in order to be able to judge
whether or not a student's piece of work meets those
that seasoned criteria. That's awfully pomplex, because.
Speaker 2 (05:38):
So it all depends on the on the human input
before it can grade correctly. Yeah, we're responsible.
Speaker 5 (05:45):
We've got to we've got to teach these assistance we
and we want them to work.
Speaker 4 (05:51):
And you know they don't, they don't just come out
of the box.
Speaker 5 (05:54):
They've got to be changed, just like a teacher, just
like anyone who is going to be assessing that work.
Speaker 4 (05:59):
If you're going to be looking at.
Speaker 5 (06:00):
A reading a student's essay, you know, how do you
know what what good looks like?
Speaker 4 (06:05):
What the standard that needs to be met? And so
it needs to be trained on you know, years and
years of of students work, so understands what level to
expect and then how to be able to you know,
provide feedback or you know, it's an attendment. And this
all has to be transparent, right, we just can't be
a black box. We need to know what data these
(06:27):
tools are trained on. That is determining what you know,
how are they breeding the work and what's the what's
the data set?
Speaker 5 (06:35):
And of course that opens up opportunity for bias and
and all sorts of different things.
Speaker 6 (06:40):
And so it's not it's it's it's a it's it's
it's a complex it's a complex space. It needs you know,
it's good good thoughts.
Speaker 2 (06:51):
Rand it's interesting when you say we don't trust it.
We don't trust humans much either we don't trust the
media or politicians or any It seems so late.
Speaker 4 (07:02):
New Zealand is trust trust humans a lot more than
they trust I mean, it's been count countless reports that
have been troublished over the last year and museums before,
you know, forty first out of forty seven or something,
light at the bottom of a list of countries in
terms of their the level of trust and AI systems.
(07:22):
And you know, I'm unsure how how has managed to
get to this point, but you know every.
Speaker 5 (07:30):
Day where we're trying to rebuild that and and and
show people, you know, what's what's possible today and how
they can be utilizing and leveraging these tools basically and responsibly.
Speaker 2 (07:46):
The only I know that a number of employers are
using AI to sift through applicants and and job seekers
are using AI to judge up their cvs and cover
letters and things. How how do we get to the
point that I just worry that when it comes to
(08:07):
sifting through so you find the appropriate applicants, sometimes people
who aren't who don't have the right background or qualifications
for the job. You just meet them and you know
by the cut of their jib, by the sparkle of
their eyes and the bushiness of their tail, that they
are going to be a good fit for the company
that you that one on one meeting is so important
(08:30):
that you might miss out on that if you're if
you're just you know, if you use a screening process.
Speaker 4 (08:38):
Yeah, but I mean, well we've always had a screening process, right,
There's always some there's always some gates that you have
to get through, and now you have to get through
the AI gates before you get to the human gates,
before you get to meets human. Interesting, Yeah, so you know, yeah,
it's and then you've got the whole generic as people
(09:02):
you know are using checchipoutine or right there a letter
and they're not using it to write in their style
or tone of voice, and you can get a very bland,
you know, very bland cover letter and CV, which means
that everything can you know, from a recruiter or a
hiring manager can all look rather the same, sane And
(09:23):
so you know, thinking about the skills that employers are
looking for, and you know, if you are going to
be using AI tools, how do you use them to
stand out. Yeah.
Speaker 5 (09:35):
And you know, and we know from the recent Cultivate
AI for Work report that employers are looking for people
who can demonstrate critical thinking, people who are not just
going to take the responses of a I tools at
face value. That they are going to blend learned experience,
(09:59):
domain knowledge and AI tools to supercharge the work that
they're So it's a part they're using it as a collaborator.
It's not you know, this is this is how I
this is how I do all my work. It's this
is how I do my work smarter.
Speaker 4 (10:15):
Yeah.
Speaker 2 (10:16):
Yeah, absolutely.
Speaker 4 (10:18):
I think where we're worry as people who you know,
are still saying, oh, we shouldn't be using this, or
if you're not allowed to use it, or you know,
it's cheating or something like that. It's like you're just
not clearing students for workforce that they're about to enter into.
A few are blocking them from using A tools. Yeah,
(10:39):
you know, it's like saying don't use the internet.
Speaker 2 (10:42):
Yeah, absolutely, or calculators. I think there are even some
people said okay, just.
Speaker 4 (10:48):
One final ilevant.
Speaker 2 (10:51):
A textasy is how best would a seventy six year
young retired woman learn about AI as a tool she
would like to.
Speaker 5 (10:58):
Use losee Oh what a fantastic love this.
Speaker 4 (11:06):
Experiment. To jump onto chat chipet. You can use the
free version or if you can, a for a plus
version and start experimenting, start asking the questions, look more,
look for a free course in your area. There might
be people that are dying free AI witch shops. Otherwise
(11:31):
there'll be YouTube videos and those sorts of things. So
to start experimenting, ask tea what you can do. There's
the right because you can't ask Excel what you could
do with itself, but you can ask chachipt how I
can use it in my daily life. And so start
(11:52):
having a conversation with it and start touched up and
curious around. Just start start start using it.
Speaker 5 (12:02):
It's the best way of doing it, asking Fisians and
being curious brilliant.
Speaker 2 (12:06):
Thank you so much for your time. Justin Flint, the
founder of New Zealand AI. There's also workshops that Justin's
company offers, some paid, some but there's some information on
there as well.
Speaker 1 (12:18):
For more from Kerry Wood and Mornings, listen live to
News Talks at b from nine am weekdays, or follow
the podcast on iHeartRadio.
Popular Podcasts
On Purpose with Jay Shetty
I’m Jay Shetty host of On Purpose the worlds #1 Mental Health podcast and I’m so grateful you found us. I started this podcast 5 years ago to invite you into conversations and workshops that are designed to help make you happier, healthier and more healed. I believe that when you (yes you) feel seen, heard and understood you’re able to deal with relationship struggles, work challenges and life’s ups and downs with more ease and grace. I interview experts, celebrities, thought leaders and athletes so that we can grow our mindset, build better habits and uncover a side of them we’ve never seen before. New episodes every Monday and Friday. Your support means the world to me and I don’t take it for granted — click the follow button and leave a review to help us spread the love with On Purpose. I can’t wait for you to listen to your first or 500th episode!
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Special Summer Offer: Exclusively on Apple Podcasts, try our Dateline Premium subscription completely free for one month! With Dateline Premium, you get every episode ad-free plus exclusive bonus content.
24/7 News: The Latest
The latest news in 4 minutes updated every hour, every day.