S2 · E02 / 46 min
Dealing with AI Slop at Work
When AI-generated work lands on your desk, who does the thinking? Ibukun and Grace discuss judgment, accountability and better standards for using AI at work.
2 Oct 2026Listen here using Spotify’s player, or open the episode on Spotify.
Player not starting in this browser? Open this episode on Spotify (opens in a new tab)
In this conversation
AI slop can shift work onto colleagues: they have to recover the context, check the claims and decide what the output is supposed to achieve. Ibukun and Grace discuss the cost to professional trust and everyday collaboration.
Communications leaders have a role in setting the standard. The conversation covers being open about AI use, giving people training, setting useful parameters and adapting tools to organisational and African contexts.
Writing remains part of thinking. The hosts explore how to use AI for analysis, research and feedback while retaining your own voice and responsibility for the finished work.
Links discussed in this episode
Ideas to take back to work
- Review AI-assisted work before sharing it, and be ready to explain its reasoning.
- Define what productivity means for the task; measure the usefulness of the result.
- Treat prompting and AI use as skills that need practice, guidance and feedback.
- Give tools clear context, brand guidance and a specific job.
- Keep reading and writing in your own voice, and stand behind the work you sign.
Transcript
Machine-generated transcript. This transcript was generated by AI and may contain errors. It has not been checked word for word. Please use the audio as the authoritative version.
Hello, I'm Ibukun, and I am Grace. We are the hosts of Unmuted Comms, a podcast for comms leaders by comms
leaders. This is a podcast about the leadership, influence, and decisions behind great non-profit
communication. We have candid conversations about what it takes to lead communications in
the non-profits sector. These include things like building strong teams, advising executives,
initiatives, navigating organizational dynamics, earning influence and making comms a more
strategic force for mission and impact. Today's topic is something I think everyone
is now feeling in the workplace might be a bit dicey step on toes quite a bit but I
think it's worth the conversation we are speaking about AI slope what is AI slope I'm sure you have
seen it, you have read it, interacted with it and you can pick out AI, the iteration
of the text, the flow, the structure all of it. And it's something that is so has become
so in every aspect and place that we are working and we are collaborating, we are speaking,
you can even see it in text messages, WhatsApp messages like seriously people are completely
giving over the creativity to AI. And for this episode, let me just start this conversation
with painting a picture and speaking on the work that I receive. I find AI in itself as
a double-edged reality, such as is the case as well for AI Slop. This is because there's
more that is produced, but the more you produce now, there's also more editing. So you'll
find, for example, I will get a lot of content from my colleagues. And I'm so happy to see them
actually producing and feeling confident enough that they have got a tool that helps them to order
their thoughts, put down their work in an organized format. The sad thing here now becomes the fact
that everything looks alike. The tone of the brand is lost. The brand itself, I cannot find it in,
you know, the story, the angles, the whole thing just looks so similar. And so it takes me double
the amount of work it used to in the past to do, to edit, to bring back, you know, in a kind
of perhaps, so we say oxymoron kind of way, the human element that we are trying to get
to. You can imagine humans trained these machines, but now we are also seeking for more humanity
in it. But that is what we're discussing today. And when I think about it, I also see
that, of course, we cannot forget that we lose, yes, I've talked about the brand and
the tone. But then also there's a nuance, which is a big part of it as well.
And then everything starts to sound the same.
So if you remove the logo right now from many, many, mini organizations and many pieces
of content that you're receiving, will literally start to look the same, they're so standardized
and it's boring.
So taking so much time to solve what is apparently supposed to be a productivity solution is not
on. On the other hand, you want to be able to produce enough content with certain timeframes and in an ever-dynamic and audience-driven environment this is key. So I guess as comms professionals we are well on the horns of a dilemma. Should we then be considering the lesser evil? And I pose that to you, Ibukun.
Thanks, Grace. For me, I think sending AI SLOP to people is just plain disrespectful.
Yeah, we're all using AI or at least a lot of us are using AI at work now in different capacities.
But SLOP itself means that what you're pushing out has not been carefully thought through.
The AI literally did all the thinking and it's obvious because A,
A, it usually doesn't make sense. And B, when you ask the person who sends it, they can't defend
the work. They are equally confused by what the AI did. And then both of you are now trying
to figure it out. And that's what you're doing when you're sending out stuff. You're not
trying to present something useful. You are just outsourcing that other part of the work
to the other person. And this is 2026, the year of our love. You cannot blame the AI for anything
thing you do. Nobody forced you to use it, the AI certainly didn't use itself or send work.
So if you send me work slap, it's like a slap on the face, if I'm being honest. It gets worse.
You know, I've seen that sometimes as comms department, you send someone work that you've
carefully prepared, someone sends it back with feedbacks or edits and says, and it's obvious
that the person use AI, and you're wondering what exactly is going on here. I've seen this several times.
An even funnier case is now when someone something you've written with AI,
and then someone tells you oh it sounds like AI,
and then sends back their own version which is also written with
AI. Basically the person is saying my slope is better than your slope right?
And it just makes you wonder what exactly are we doing?
I think that's a good question to pose. What exactly are we doing? It almost seems like you're playing a game of who is the worst AI user, but seriously it's a
good point to dive into this discussion and what do we need less of and what we need more of? I think
that's what we'll be unpacking today and how to move into what's wrong with sending AI slope to colleagues.
Well, first of all, we need to take our careers more seriously. Professionally, sending slop or sloppy work, it's the same thing if you ask me. Slop, sloppy work is just saying this is not good. It's not a good look on you.
It gives up the vibe, like you don't take your job seriously. And if you're looking to be taken seriously, let's not get into how comms is already fighting an uphill battle to be taking seriously in the workplace.
AI Slop is actively doing the opposite and working against our interest, especially as a comms department. I might not know for other departments, but as a coms person, I'm speaking from experience.
and speaking from how the trend has been, the conversations that I've been having.
You know many people, many organizations still struggle to see the value of a well-resourced
comms department, right? It's a conversation we've been
having for ages. It's conversation I met when I joined the industry. It was already something
that people were already discussing and have been in the industry for over maybe 11 years now
and it's still something we are discussing. Now imagine that same department we're trying to get more
resourced is sending people clearly AI-generated slope. One day someone is going to wake up and just
say, enough is enough. There is no point having this department if we can just have an intern who
is very good with AI. And so the comms department is literally just some intern or one associate
with an AI subscription. And from what I've heard, this is already happening.
And by the way, this should not even be happening. I think it's unfortunate that organizations are
very quick to replace functions with AI, but that's a conversation that I will happily
sidestep for now. It just requires its own time. And interestingly, I read a poster just the other
day about having a hiring crisis in future because there's going to be this gap where we cannot get
senior people or senior professionals, because they've all been replaced with AI. So I think what
you're saying in terms of, you know, it's happening, yes, but that shouldn't be the direction
that we are taking. So, yes. But I digress. My biggest issue with AI is a loss of individualism,
especially in creativity. And, you now, we are very creative field. Recently, I stumbled upon
on a Reader's Digest condensed book from 1978. And I'm giving you homework here, if you
do not know Reader'S Digest, please go Google it. That was the publication at the height
of its day. And each story in this 1978 issue, it has so much flair. I mean, the play of words,
the expressions, where they paint the picture, you're literally working in the different scenarios
that I've painted before you in the different geographies, it is, I long for a time we can
get back to that.
I don't know if we'll ever get back perhaps, but I betcha it will be a premium dollar for
people who can still work like that.
And so what's considering how to make sure that you're keeping that creative part of
it? That for me is the big thing with AI slope.
I really hate how it standardizes us and we are not created to be standardized, to be
quite honest.
We're not. And I think it's even more sensitive when you're in communications, the comms industry,
you know, communications, marketing, advocacy, you just know that how you're more sensitive to it
because you know you just have to have that distinctiveness. The reason why a brand, we all
for example, I'm just recalling some of my master's class where they were teaching us
the definition of a brand. We all use soap, but there's a particular soap you look for when you go to
the market. That's a brand, that's a power brand, just having that identity that separates you from
every other thing. And so what we're trying to avoid is the commoditization, if that's the right
word, of comms, and this is why I'm on my crusade online and always happy that people should stop
using AI for writing. Good writers, I mean, I've seen people who were good writers who have now sort of outsourced their entire profile to just AI and it's really, really distressing. We need to keep those distinctive parts, like you said, the creative parts. We
We have to learn it and we have to keep sharpening it, right? And you need to keep those parts, especially AI writing. You need to
keep your writing away from AI because people who can produce things that people begin to wonder was this AI or was it not? Or it does not sound like anything I've heard. Those people are going to be in demand
and like you said, they're going to be in demand. So I think that people need to start thinking long-term about their careers and not just trying to get a deliverable submitted by the end of the week.
And probably, I'm sure we're gonna get into it, perhaps the other challenge with AI Slope is that people don't necessarily know how to retain nuance in the prompts that they're producing for example.
because, like I said, AI is a tool. It is a
tool and it has a way that needs to be utilized. But who is teaching us
how to utilize it? I guess that's another conversation. But I don't know
what have you had or what are you seeing in terms of
how to put guardrails and how we're using AI? I think that
we're all using AI at least at
company level. A lot of companies are open to it. A lots of organizations are open
But like you pointed out, people are fed up of AI slop, especially things that we have to consume.
We have companies now who are publishing their own AI writing policy because research has shown...
Now this is research now, there's research showing that AI slop is really affecting productivity.
Things that were very excellent before is slowing things down.
instead of speeding things up. For example, there was a company called Vanta. I saw this on my feed
this week. So their general counsel, that's the head of the legal department, he published the
company's AI writing policy. Now, not AI governance holistically, but just AI writing
policy. And what was newsworthy about it was that he wrote it by hand without AI assistance. That's
on. Secondly, normally this type of policy would come from comms or marketing. That's the AI writing part, right? But this came from the legal department and I think everyone should read it by the way. We'll leave it in the episode's description on the website but the interesting part was he called it slop grenades that when you send slop work, he says like sending a grenade that disrupts other people's productivity because they spend more time on
on average, about three hours every week. Research has shown that people spend close to about
three hours each week just trying to fix AI work and making sense of it. I think that's
a big deal. As someone who has been, I'm sure a lot of people have received AI sloppy work,
you just know that this is not a good use of my time. And it's really, really frustrating.
and as a leader is something that my team, people I work with know that I think we are at the place now. We're in a good place now where everyone knows that whatever must come to my desk for review must not sound anything like AI or otherwise we're going to have words, you know?
You have decided to declare war.
Oh yes. Oh, yes. And I think everyone sees and appreciates what I'm trying to do is I'm not just
trying to be difficult. It's, and I'll get to that, right? But it could lead to many things.
It could lead too many things someone put it this way said the social contract with writing
is it should take you longer to write than it takes me to read it. And that's true because
because the amount of effort it takes to communicate, there is a certain amount of
effort. There's skill involved. You think you're a good writer because in our head everything sounds
clear until you start to write then you realize oh I didn't really think this true. It's always
like that if communication was easy we wouldn't have built a profession out of it and this is
part of what we're saying. In that Vanta AI writing policy that was referenced earlier, the way he ended it was beautiful. He said, whatever you're going to put your name on should be something you'll be proud of even years from now. And that's so true. I've looked at some of the things that I wrote with AI. When AI first came out, we were all excited, oh, these things, rights and all that. But I've looking at some
of things that I wrote with AI last year. And to be honest with you, I'm praying that nobody on this
earth is going to read any of it. That's how terrible the writing was. So yeah.
I think we all have those kinds of instances because we hadn't yet known how to use it properly.
For sure, our initial interactions with AI was just something for tears. But that quote
what you said about, because of the amount of effort it takes to communicate, if communication
was easy, we wouldn't have built a profession out of it. That one captured my soul. I think
this is beyond even AI and AI slope, please, we just need to talk about it and show, yeah,
if someone is wondering what's the value for comms, that is why there's a profession around
on it, right? We have to just take a moment and acknowledge that. And I think at this point,
even just as we're thinking about this AI policy that was handwritten, I don't know if people
understand that handwriting is now becoming such a big deal. It fascinates me. It baffles.
The more things change, the more things stay the same. This person hand-wrote, and for us,
like, wow, they did. Can you imagine that? I don't know. But the fact that it was written by a lawyer,
I'm just curious for our audience, should we have a conversation about who is actually best placed to
write the organization's AI policy? Do you think that's something that we should actually sink our
teeth into? But for now, read the link that Ibukun is going to share and then let us know what you
think. So for now, let us get back to some tips on improving how you're using AI at work. You as a comms
leader who is supposed to be setting the pace, being the true north, so to speak, of this whole
thing. Number one is understanding that AI does not replace you. AI is a tool should be hanging
somewhere on everybody's wall. It is not there to replace you, the human being. And I think that
is why we're getting a lot of friction because as opposed to somebody using it as something
that just comes alongside your thinking, you're hanging everything on it, all your ideas,
your creativity, your thinking. Absolutely. Literally just I need to draft an email to my
boss to tell them this drafted hit send. You don't even come back to check if that is the tone,
you don't come back. So you literally and whether it's a report, whether it is a big time policy
for government, it is just wrong on so many levels. So that's number one.
I have a funny, just a very quick example. There was an application, there was competition
that we did recently and people sent in their applications and some of the applications
had, you know that thing that AI does after it has given you the thing, it gives you like
an addendum at the bottom, like if you would like it in another format, if you like it
It was there in the submission. And I was just like, what you just said, just triggered that memory.
And you know, that has actually even happened here in Kenya, National Daily that have been cut out with that line, that line that people don't see.
And it's unfortunate because I have even received emails with that lines.
I've received written. So it is just, please.
What are we doing?
Thank you. What are you doing? You are the human being. You are The Connecting Glue here.
So please, first of all, stop picking AI right for you, everything. At what point are you going to think? And I think AI makes it so convenient. So it's sneakily convenient. I'll just tell AI, I need this. I need another to say sorry. You don't even think about it. We did production thing and because it works on the aspect of making you feel good about yourself. You're very right. I absolutely hate when it does that.
So please, number one, it does not replace you. It's always psyching you up and I'm like, please don't do that.
I've not asked you to do that, all of us have to please analyze this thing.
So it does no replace you and so let's also agree that you shall not let AI write out the entire piece of work for you.
You want to do your, it's great especially for analysis, evidence, data, administrative stuff, stuff like that is absolutely fantastic. Let it do that for you. But when it comes to you actually writing, brainstorm the idea, look at the blind spots. Maybe your argument is weak in some areas and it's better in some errors. That is a good place to start. And then you do the writing.
And when you do the writing, your prompting is very important because even if you're going
to run it through AI to say, for example, you're not a native English speaker and you just
want to make sure that this thing is making sense, you have to tell it, don't change
a theme.
Just let me know where there's weakness in my thinking or my argument and then be the
one to correct it.
Know that there's a reason why there are commas and footsteps and all these rules and whatever
it is for writing.
It's a muscle.
You will lose it. So make sure that you are leading by example. Just as you said, Ibukun, you cannot be the one sending out AI stuff to people and then raising hellfire because people are sending it back to you. So I don't know what you think about all this.
Anyway, it's just, we could go on and on. But I think something that I think helps is to acknowledge
that we're all using AI. I think there's still some, it' a sensitive topic in some areas and in
some industries. So I think if people are using AI in the organization, we should be open about it.
like, oh, it's not a problem. Like for example, my team, they know I use AI. I love AI. It's really,
really great for the things I use it for. So, and I welcome that. They even, I want to advise them
sometimes. I try to give them better prompts on how to get to the results that would probably help
them because I've noticed. And that's the first thing I think people should first do is to acknowledge
knowledge that we are using AI, because once we are acknowledging that we're using AI then
we can talk about it. We can ask for help with an actual advice because a lot of people I
was really shocked that people still struggle to do prompts well you know and AI there's
also an aspect of it that is you know garbage in garbage out it's what is how you prompt
there's a skill to prompting and it's just being open to how other people are using
in it, right? And then you learn, you pick up one or two things and then you also develop
your own style. So once we can acknowledge that we're using AI, then we can start being
more strategic on advice and tips on how to use it, sharing different tools or even software
that we feel, oh, this is more suited to doing this and this is better suited to
doing this. That's one. Then two is that using AI is a skill, and I think that's also something
that we don't discuss as well as we should. There was another post by a professor from
Emory University that went viral, and his message was just basically using AI
is a scale. That was what he was saying. It's almost like what we have been saying throughout
about this episode that he's tired of receiving slob. He's frustrated, he's angry, you could
tell it in the post, right? And his conclusion was that this thing is a tool and it requires
some skill. The same way you won't just go to Photoshop and just put nonsense together
and send it out. That's how AI is. There's a skill to using it and people need to be trained.
people need to be open to being trained don't just assume oh it's AI we're all using it I can use it no there is a skill to it we should allow people with the expertise to teach us and to show us better ways of getting better results right um the goal is not speed at all costs it seems like in a bid to show that we are AI compliant almost everyone is now using AI and nobody's applying the judgment that take it takes to know that
that should this kind of message even be, have a whiff of AI at all? I mean, people are sending birthday messages with AI. It's, I'm like, it's just, you just wonder is the human race about to go extinct because it does not make any sense. Why are you sending me a birthday message that you used AI to write? You might, I could go to Chalchipita and ask you to send me a
birthday message, if that was what I wanted.
I want you with all your grammatical errors.
If it's too bad, just send the voice notes instead.
But AI, no.
So yeah, so that's really my thinking around that.
It's really great, like you said, about research, transcribing,
synthesis, et cetera, but that means
you now have more time to dedicate to the writing itself, and thinking through the idea.
Absolutely. And I think that is the level of understanding that may be lacking in terms
of how does AI make me more productive? It is not about me just churning out content for
the sake of it. It is actually like any other. And you rightly said, it is a skill. The same
when you go and learn how to drive, and you have to give this car the right prompt for it
to move from point A to point B. Literally, it's just the same way that you have
to learn howto prompt these LLMs, and then capture that and ensure that because, you know,
for example, even like us in Africa, you realize our nuances and contexts are not always captured
correctly, if at all. So there's that aspect and limitation as well. And if we are not careful
to understand and to perhaps, and that is something as communicators we should really
be thinking what does productivity mean for me here? Is it just churning out content?
Is it actually being able to go through, for example, large data sets and then using
that for my report and how do I then teach others to do that? I think that is important
so that you're not just telling people oh now you can use AI for your LinkedIn post.
you can use. Just go ye forth as the organization, we're all colleagues here and we are trying
to get more visibility. I mean, it's just going to, it becomes a train wreck. So I think
it's important, just like you said, having, and I love that the professor who you quote
there actually mentions it that, hey, you know what? AI, use is a skill, prompting is
a skill. Human judgment is still integral to all this coming together, right? So that
is something that we cannot overemphasize. And being that you have to lead this charge,
then you have be proficient, ask the hard questions. Why I'm using AI? How I'm
using it? What do you want it to do? So even as you're developing the ideas and bring in
different voices from the organization, and you've got something that reflects how then
you bring solutions to your work as opposed to adding bottlenecks to what you're doing.
Something else that I want to add on to that is that, and by the way, yes, write using
your own words.
That imperfect birthday message that you've mentioned, it is true.
I absolutely as well hate all those, and just wait, Christmas is around the corner.
You've seen that in me.
that are looking random and confused, I don't know, just put it down. At least then I know
that is a heartfelt message as opposed to some manufactured thing, right? When you are talking
about prompts as well, it's important that you, especially in giving parameters, I found
It will be greater success when I give the AI parameters to work within, not just review
for me. It will rewrite, review and everything, give it parameters, what do you want it to review?
Is it flow?
Is it structure? Is it argument?
Is it idea?
What do you wanted to retain in terms of the nuance and the context?
All these things, you have to think about it before getting to prompting it.
And then another thing that we've been exploring, and one that we are probably going
going to inform how, for example, we use AI in the organization. You have your brand guidelines.
Oftentimes you will have a section of it, not necessarily all of it but I know for some
organizations it's all of this available on the website so it's you know for your suppliers and
all these things anybody can come in and see what are your brand guide brand guidelines take that
document and feed it into AI and then you're able to tell it I want to you know design an infographic
for example, and I want to show maybe the relationship between part X and part Y, reference this
brand guideline, you know, and be able to meet. So it can literally keep itself within certain
spaces and parameters to work through and give you something that is on brand, taking into consideration
all the different things that are in the document. That I have found to have really, and we keep
exploring with it to see. The challenge has been some teams will come and say, okay, so can I just
to read that and then I tell it to write for me a document based on all the, and now
that opens another hole.
So this instance is that no, to this point you can only use this methodology, please,
and not to create images.
I don't want to get, today we're talking about AI stuff, I don' t want to use your AI
images for your own team staff.
Oh my God, the images are the worst, oh God.
Please not AI images, not event posters, don't just right now we are being so no you want your infographics yes you want to find that we have a chat stuff stuff like that and then let us please teach before you release it to the world. So that also yeah we are giving just enough room for experimentation without um without you know saying too many no-no's but also being very cognizant that my goodness the more you give the more people want to explore so up to that point
I know we have brand guidelines and I know AI can do for you many things, but we don't
want it to get to that level. But yeah, so that is one thing, that is something that we are
exploring in the organization. I don't know what that looks like for you, Ibukun.
I think that that's such a good idea. I think, you know, I've been hearing a lot of people
talking about multiplayer AI, that it's team members all working with the same AI and then
and the AI is interacting in real time, like an actual team member.
So that sort of links into what you just said.
I think having a guide, that kind of guide somewhere
that the AI, all our AIs can easily access,
not just sitting on one person's desk
but it's in a central place.
Anybody in the organization using AI can easily
access or maybe even forced to access
before anything is published.
I think that would be very good.
It solves a lot of problems. I think it's a very good idea. Another idea is maybe we can create a free web page online that people could point their AI to before they publish anything, you know? So that thing sort of sweeps for brand violations. That's also another idea that, you could explain just, you saying it just sort of triggered a flow of inspiration because these things are
tools, right? They need as much guidance as possible. So I think that kind of website would
really be helpful. Another idea that in our organization, something that we're experimenting
with is creating skills for people who produce anything that has to be written. So a skill
is basically like a model, a packaged reusable set of instructions for an AI to use. It tells
the AI how to do something, or how you want it to do
something. You know because basically these are large language models LLMs so they
are a bit like generalists they know a bit of everything and so they don't unless
you prompt them properly they will not go as deep as possible into that particular thing
so you have to be very specific um so those are what skills are for if anyone who is very familiar
with AI this year would have would this would not be news to them but for a lot of people
are still not, they're still very familiar with only chat GPT. There is a lot that has happened in the AI space right now. So even if you're going to use AI, you might as well even just learn it very well, right? Instead of just giving us the low level sloppy version. That's another one. I think if everyone in the team are using these kinds of skills made by the comms department, theoretically speaking, you can improve everyone's output or at least upgrade it beyond just the regular slop.
but it will not eliminate the sloppiness of the AI. I don't know why, it's just the way the machines have been trained but it does not fix it. However there is always a market improvement if those skills are used so that's really what do you think?
I like that. Personally, I didn't even know about the skill and that's something I'm looking
forward actually to explore because it would take out a lot of, you know sometimes you put down all
these documents and yes you put them on a shared platform for people to access but they're not
necessarily using it but if you have something that you can easily and quickly guide someone
and say okay here's what it is, that one I think is something that we'd be happy to utilize where
were able to say, this is how speakers' organizations are. They'll give even examples and really train the model. I think that is one thing that I hear a lot of people who are researching, who are working very closely with AI and doing all kinds of things with it. Train the model to be, I don't want to say it's not to be like you, but to really perform at a higher level than your regular models, because it understands your voice, it understands the words that you use.
your flow, even your humor. It just understands things about you that makes it that much easier.
You can say, okay, fine, utilize this and I want you to be able to build this and the other for me.
And like you say, it's never 100%. I have seen even others that have been marketed on,
I don't know, platforms like LinkedIn and purchased this because now we've gotten it right.
But when you go and look at it, when you're doing a testing period,
but it has not eliminated this thing 100%. So even as you're here marketing 90-day guarantee,
whatever, we are not yet there. That's why I say human judgment cannot be replaced. It must, it is the differentiating factor here and must be used alongside AI.
That's one person who said you can use a slop generating machine to fix the slop generating
machine. Exactly. It's just how it is. So I was going to say this earlier, but I think there's something people aren't aware of. I also learned about it, I think maybe last year and that's this phenomenon, it's called the IKEA effect. The IKEA, you know, the company, IKEA where you buy things and you put it together, do it yourself, furniture. There's this
phenomenon, it's called the IKEA effect. People tend to overvalue something because they helped create
it. I remember the first time I used AI to do something, to write, to generate an image,
et cetera, or to even do a very complex calculation. I was so thrilled. Like, oh, wow. And that's
the IKEA effects. You become very sentimental about it. You've become very invested that,
that, oh this was something I helped create and I had minimal impact. It's just the wonder
of technology right? It's normal it's just like first time you see an iPhone being used
it was magical part of the problem is that it also helps makes you blindsided to criticism
or free feedback you stop being open to it and the thing about being a good writer or
a good commsperson is the ability to receive feedback no matter how good your writing
business, one of the things they train into you is that you need an editor. And a lot
of people now where the problem is that a lot of people are not familiar with this process
or even willing to do that. That's why we have people online right now posting slop and
they're proud of it. Or you hear people saying things like AI generated this in one shot.
I didn't do anything. And they're saying it with pride, so much pride. It's the IKEA effect
and you're like, this thing is not really good. Whoever, you know, is close to you should have told
you that this thing isn't really good, it would have been better if you just did something else,
right? So things like that, that's one of the things that we are seeing right now and it's
really weird but I know it's a phase, it's always like that with new technology but that's what
you said earlier about the personal responsibility, you need to, it is exciting but you need
to know that there's also a brand or a personality or a responsibility aspect of it. What you post,
what you create, what you put out under your name is it reflects your own judgment and it makes us
realize that okay I don't think this person if this person says the comms person what they are selling
or their premium or their value proposition is the judgment they bring you're not really displaying
that judgment in your own personal branding so how am I so sure that you're going to do it with my
my brand or my company and I think that's a realization many people forget that you have
to lead by example and even as I mean as we've been speaking about this I think one two challenges
that I would probably pause for put forward for all of us is number one run a session on how not
to use AI if you're looking for accountability just building on what you just say right now I
I mean, that is a session on how not to use AI. Organize that for the organization because
it also keeps you in check. You'll be able to pick your AI, stop, you're able to
pick out what people say, this is not it. And then a challenge, be part of the drafting
of the AI policy. Offer your services and say, let me come in and write for it because
I think at that level, it becomes very difficult for you to sit back and say oh my goodness,
I don't know why these people are doing one, two, three things and you're doing the same
thing. Once you put yourself out like that in those two ways, training and drafting that
document, I assure you, even you, your judgment is going to come in very heavily about how
to use AI, how not to use the AI and everything else about AI's law, for sure.
Sure. Like, for sûr. Of course, it won't mean you many friends when you turn into the AI
I've had some people, you know, enter my comment section and being a bit negative, which is fine.
I really, really love it. You know, let's have the debate, let us have the conversations.
But sometimes you have to be the bad guy to save people from themselves. It's just the truth.
Right? Yes. So, and I don't mind. I don' mind being that person. I'm very fine with it.
as long as we get the desired result, no problem. Another thing with AI is the verbosity. Oh my God, I don't know why it's always so long. I have to tell it, can you just be a succinct? Don't tell me 99 or I want a yes or no, please. Oh, my God. That thing irritates me. It just goes on and on like, and what you need is probably just one paragraph and it goes and gives you a word.
I am used to, right? Exactly. The truth of the matter is that you can make anything done
by AI better just by making a shot. Just tell it, just give me the points. I don't need
the history of this thing. I do not want to know what people ask. Just give me what I've
asked you, you know? Exactly! So, it's really funny that, and I don' know, maybe AI is picking
making it up from us, because you know there are people within organizations where they
posture with very long emails or very long documents. You say oh let's do this and then
they send you a very long report that they know nobody is going to be able to read because of
just how long it is maybe that's what the AI too is trying to do I don't know but um but badly quote
George Orwell he said you know never use a long explanation where a short one will do that's
part of the comms training that we're talking about. I think that's even something to feed my
my AI agent and just tell it, please, can you refer to this particular quote severally?
Please. Because at some point I even get confused.
I'm like, what did I even ask you?
You've given me so many scenarios and so many this, that, the other.
And I'm thinking, no, my original question is even lost on me, completely lost on
me. So I understand that is one of my absolute pet peeves over the above AI slope is just the
verbiosity of that tool.
I hit it with a passion.
and yes, you have to be the one to come and let it know. No, just in a very short concise precise piece don't even try to take it out and about but I know also there is the issues around how AI then kind of makes you right so to speak maybe maybe you want to comment about that yes we talk about writing, writing styles
of, even, I don't know, personally one of the things that I experienced was that in trying
to write, I started sounding a bit like AI. And I think that's from engaging with it a
lot and reading what it gives you. It's really, really disturbing. And because it makes sense
because as a writer, you are what you read. Most of us develop our style of writing from
books, like you said, Reader's Digest, novels, Stephen King and all that. So now that we are
consuming a lot of AI work, it's just going to seep into the way you write. And the problem
is it primes your mind in that dry, very formulaic AI. It's not this, it is that. It is this or
that matters, the load-bearing. It's so formulaic you can tell and so sometimes just doing the first drafts from scratch I think it's just a good idea because eventually this is just normal. The way you get better as a writer is to read really good writing especially people who can write better than you. So if we are spending a lot of time at work reading AI stuff either from our own AI use
or what people are sending us, then we need to know that this is a career hazard and we
need to start some of those mitigation steps. I think that's really just something else
that occurred to me sometime this year. Reading that thing that people don't like
to do and they do not know how valuable it is but in this conversation today we've talked
about AI slope and given practical tips of how you can mitigate what we are already seeing.
Of course, it's a learning space for all of us as we continue to find better ways to utilize
this tool.
And I just want to encourage you as you are listening and you are finalizing and doing
all this that you need to do in terms of AI slope, please take this conversation, see how
it applies to your current place of work and then say no to AI slope. Completely ban
it. Have like, I really have got that no hunting kind of thing there. Please just AI slope
with a no sticker, put it everywhere. And then agree to be held accountable. I think
that it's a space of vulnerability, but it's
a place for us to be able to show our value where we can say, you know what, I'll lead
by example. This is the direction that I think we should take. So yeah, I think
I hope that this conversation has at least sparked something
in you. And if you think otherwise, think differently, please
by all means, would love to engage in that conversation. PMs are open.
Yes, please. Yes, they are open, we'd love to. Or invite us to your podcast
so we could come discuss it further.
Absolutely.
I would love to hear from you and your feedback.
So thanks, Ibukun.
It's great speaking with you and hearing myself think, so to speak.
Of course.
All right. Thank you.
Take care.
Thank you for joining us for this episode of Unmitted Conversations in Comms.
We'll see you in the next one.
Hosted by Ibukun Taiwo and Grace Thuo
Music: Uplifting Bass by Lily J, via Mixkit