Home
Sign Up
Jenni Ryall from Other Labs
Oct 19, 2026
Jenni Ryall from Other Labs
15
30
00:00
31:01
1×
SUBSCRIBE FOR UPDATES
Subscribe
Listen on
Apple Podcasts
Apple Podcasts
Spotify
Spotify
YouTube
YouTube
Transcript
0:00
I can work right now at probably 100 times what I could work before. Things that would take me months, I can do in an hour. A lot of those companies in that 20 years didn't fully transform to digital. Yeah.
0:12
And so now you're trying to bring in even more new technology, and they hadn't transformed in the first place.
0:18
It's like the new social media policy, except it's not public, it's, like, everything you share privately- Yeah, yeah, yeah... which is maybe worse.
0:24
What would you build if you started from scratch, but your goal was to inform an audience? This is Not Just Faster Horses, conversations about deploying AI in media companies across product, engineering, and technology.
0:36
Today's guest is Jenny Ryle, who has spent the last two decades inside media shifts that have reshaped the media industry.
0:43
Mashable through the platform era, Tenor in local US news, a stint at Meta during the media bargaining code. Now she runs Other Labs, a advisory helping media companies with AI and what comes next.
0:56
Thanks for joining me, Jenny. Thank you, Ben. There's this big disconnect, 88% of people saying that they're using AI somewhere in the process, but only 6% seeing any sort of bottom-line improvement.
1:06
What is it about media companies that's unique? Yeah, I think the thing with media is that, you know, you're dealing with companies now that are
1:14
decades old, um, in their operating systems and in their tech setup from, you know, clunky CMSes to, um, really old tools that have been, like, built on over time.
1:26
Um, so there's, like, Band-Aid on Band-Aid on Band-Aid when we're thinking about traditional media.
1:31
You know, the newer digital players, I shouldn't say new anymore, from, you know, a decade ago now, um, a lot of them were built on technology, so their, um, ability to change may be a little more fluid, but I still think there's a lot of legacy systems built in there.
1:46
Now, and this happens in every organization, as we have spoken about before, even our own, is that, you know, you're building workflows, you're moving fast, and you don't have time to go back and think about how, um, it might change.
1:58
And so... And media's also been burned by technology over and over from, you know, the pivot to video, um, to the social media era that didn't turn into the business models that were expected.
2:12
And so there's this real tension, obviously, that we all know about between technology companies and the media, and I think that that hurts, and so moving through that to a place of collaboration is hard.
2:25
Um, and you know, I- you mentioned I worked at Mashable, uh, back in 2014, ages ago now, when that was, like, the darling of, um, the social media era, really built on the platforms.
2:37
And you know, we had tools like, um, this cool, like, trending tool that would bring in 500 million URLs across the internet and tell us what people were reading, and it allowed us to act.
2:48
They also built their own CMS, which was the most wonderful CMS I've ever used. Um, and all these tools reduced friction, um, but ultimately their business model didn't sustain what was to come next.
3:01
So now we're in this position where we've gone back to older business models in some way, with subscription now on top. Um, and is that the business model that is setting us up for the future?
3:11
I think that's an open question that we now need to think of, um, with the new technology that's coming in. It's not the first time we've been...
3:18
No [laughs] Like, this seems like digital transformation mark two or mark three now.
3:21
[laughs] It is like, it's funny when you are working in these companies, that it is transformation ultimately, and you know, you're going in and you're looking at the workflows, and AI is a technology that you can implement to help make those workflows more efficient, but it comes down to the same things that it's always come down to, which is often, you know, fear within the organization, which is the people, um, you know, the basics of change management, of understanding where your blockers are and f- Yeah.
3:48
Do you think there's, like, even just inherently in the structure of a lot of, like, I guess, uh, in a, in a larger business you've got these sort of defined silos of how work has always got done, maybe dating back to, like, when the print was the product, not digital as well?
4:01
Yeah.
4:01
Um, and there's definitely a lot of, you know, the technology that everyone knows, which is, like, CMSes that are really clunky, manual tasks like metadata that you're entering in manually, tagging systems that were never really tagging systems and didn't work [laughs] to the way they should.
4:18
But now, um, yeah, using this technology, some of those processes can be simplified.
4:24
So I think, um, that that is a model that was built for traditional media that, um, has become a problem over time, and I, I do think, like, with the digital evolution or transformation or whatever you wanna call it, a lot of those companies in that 20 years didn't fully transform to digital.
4:43
Yeah. And so now you're trying to bring in even more new technology, and they hadn't transformed in the first place. Um, so I think that's harder. And then I obviously deal with a lot of, um...
4:54
Well, not obviously, I deal with a lot of, uh, small publishers, like, even down to, like, one and two-person shops through, um, my programs, which is I run the Google AI program, um, previously ran the local news revenue program, um, and ran similar programs at Meta previously.
5:14
And from dealing with these really small publishers, I always find it interesting, um, that they often need to innovate or innovate in a way that, um, it's an add-on, the technology.
5:26
Um, and there's many reasons for that. Obviously they- they're not beholden by, um, old systems or processes, but, um, I do think it is a mindset of just that this can supercharge us, and so moving quickly to that.
5:40
And I think, you know, startups and things have been playing that game for a long time, and I don't think we saw that hit media- Yeah... in the same way.
5:47
I think one example we were talking about before was, like, Gazette- Mm... um, in sort of how they were able to sort of build that product to sort of s- uh, speed up internal workflows and things like that. Yeah.
5:55
So Gazette is a, um, relatively new Australian media startup.
6:00
Um, and they are, um, building a, a local model from a central base, and they worked with us on our Google program to build a product which they called Civique, with a Q, um, and it was...
6:14
Uh, the purpose of that was to, without going to council meetings, could they source that information themselves, and could they turn that into quick-turn stories that could have a human in the loop, an editor review, and then it go out the door?
6:28
And what they found was, um, yes, that was possible by using AI. Um, they had to obviously, um, work with the different forms of how councils were putting out information.
6:40
Some were YouTube videos, some were text, um, emails, um, but they got it in a place where they five times their story output over the 12 weeks they were in the program.
6:50
And I just think about something like that, and if you take that from that really small, um, scale, I think it went from like 1.2 stories to 6.5 stories, so really tiny scale, but what if you think about that model over, you know, the whole of Australia, what does that start to look like?
7:07
Mm. And think of the scale you could get from this technology from using a workflow like that. And that is supporting editorial. That's not ta- It's helping them do more stories than they ever would've been able to.
7:19
Um, and there's a great, well, great, interesting model in the US, 6:AM, which is, um, using,
7:27
um, AI technology in the same sort of way, they're a newsletter company, to work out what markets to go into by running AI newsletters and then deciding where they put the human teams in place.
7:38
Um, and you know, at that scale, I just think it's, it starts to become really interesting, and you've got traditional media that's gonna start competing with that.
7:46
W- what holds them back from being able to move that fast and that agile compared to sort of these smaller players? Yeah, I think, um, although we just wanna move and run, like, it's a moment to sort of
7:58
sit back, look at your business as a whole, figure out where those workflow bottlenecks are. Is it workflows? Is it technology? Is it the people? And then, um, figure out your plan forward from there. Mm.
8:11
And it's really hard for media companies to not be reactive. That's how we're, we're built, we're born- Yeah... we're bred.
8:16
As anyone can now build out a feature or test a new, you know, um, a product that, that ingests, you know, court documents or like- Yeah... whatever else it might be,
8:24
does that old model of decision-making work still, or is, is that sort of up, up in the air as well? I think what we're seeing now is actually these core teams from across an entire organization- Yeah...
8:35
come together to solve a problem, and without that, you lose expertise of the different issues, so it gets stuck at different points. You also, um, don't get the buy, like, buy-in in those environments.
8:48
It's huge that the editorial can have a, have a voice and feel- Mm... heard, and they're the end pro- uh, they're the end user at the end of the day, outside of your audience.
8:56
Um, and I think you and I would've seen a lot of technology that didn't land- Yeah... because the users [laughs] were like, "We won." It was just dropped in, yeah.
9:03
[laughs] The idea for journalists and things to be more hands-on and thinking like product people as well- Mm...
9:08
like everyone in the business, uh, how, how do you sort of manage that change or- Yeah, I think it's actually empowering for a lot of people to be in those conversations.
9:18
I was lucky enough, like, early in my career to be brought into those conversations as an editorial person, and that actually ended up shaping, like, where [laughs] I went with what I've done and why I like, you know, I love collaboration.
9:28
I love technology. Um, and so I think there is, like, a, you know, exposure therapy in a way of, like, letting people see how it's made.
9:36
They will have opinions, especially when they are the end user, as you know from the technical side. So how do you sort of think about, like, changing a, a 20-year-old process?
9:45
Yeah, I think it's definitely changed probably even over the last couple of years.
9:48
I think, like, when we started running these programs, they were really focused on, you know, that small-scale, 12-week experimentation sprint, and then you would look to whether you, um, scale it, you pivot it, or, um, you decide to not move forward, and it was at that point it would turn into full implementation.
10:08
Our, the latest program, which is across Pacific and New Zealand, is actually much more focused on implementation, and it goes for 18 months.
10:17
And I think, like, I don't know how, you know, from a product perspective, I'd love to hear your view, but, like, lengthening that so that we actually get, like, deep into the implementation with the teams doing it has become actually really useful, and the reason for that is, like, this time, you know, you're supporting on the tech s- your company's supporting on the tech side of that.
10:37
We're able to bring in technical support, and that is so... You know, we have this thought that everyone's building with AI, but some people don't wanna be building. Yeah. They wanna be the ideas person.
10:48
They wanna direct it.
10:49
They want to, um, they understand the strategy behind it, but they don't wanna be there on the tools, and so by allowing them to have access to a tech team, they can then start to build in that collaborative way, and then making that decision at the end, like, if it didn't work, it didn't work.
11:02
Move on. Yeah.
11:03
So some of this would be obvious to you 'cause it's product thinking, but I think when you haven't had that expanded across a news business before, um, it's all new ways of thinking that, you know, are still being worked through.
11:14
Yeah, especially if your job was just to sort of fill an A4 page with an article- Yeah... and then they got hand- everything else got handled downstream. Right.
11:20
To now have the ability to influence that whole process, uh, uh, what happens after you finish your, your piece of work, um- Yeah, and still having the audience in mind, which I think is, like, an important thing that we can get to as well, is, like, we can't forget the audience in this, and the, the people closest to the audience are often that editorial team or the audience team or whoever.
11:42
Yeah. And so it's really useful for product in general. Yeah. There's a new line item on the P&L now- Mm... for AI tooling. Yeah, mine's huge. [laughs] Yeah, exactly. Just keeps growing.
11:50
[laughs] Uh, how do, how do you think about measurement? Because it's, it's incredibly tricky in an engineering perspective. Mm, yeah, I think, um, it obviously changes, like, and I see that across the-...
12:00
the different programs. You know, when you're running a revenue program, what are your metrics, um, verse, you know, even an audience program and now an AI program.
12:08
Um, you know, with AI, it's funny, the first program we didn't set the metrics that needed to be tracked, and by, um, this one, which is the fifth cohort, we've decided that actually having a time spent and accuracy metric is, like, essential as the foundation.
12:24
And then if you require other metrics on top from like a product side or, um, what have you, that they can be additional metrics.
12:31
But I think, like, getting consistent with those as a baseline and having people test to those from an editorial or, um, you know, where we need accuracy at 100% [laughs] Yeah... um, that is, like, so critical that...
12:44
And also when you're toi- teaching people about, um, metrics to do with technology or AI for the first time, um, simplifying it is actually really helpful because if it's complex, they get lost in the complexity, and it's really hard to track those metrics week to week.
12:59
Um, so we just, yeah, we start with a baseline, which can be like for an AI pro- for a process pre-AI, the human process, it could just be sitting down and timing how long it takes someone to do a task. Mm. Right?
13:11
Um, really manual, or I'm sure there's a more AI-driven way now. Yeah. Um, and then- Time to publish or something. Yeah. Yeah. And then that's your baseline. Yeah.
13:18
And then, you know, the savings you have through technology end up being, um, your time saved metric that you've tracked over those 12 weeks or 18 months or whatever it is. Yeah.
13:26
How are you thinking about it for the 18 months? What are, what are the sort of things you're expecting to see? 'Cause I imagine it's still developing right now.
13:32
Yeah, it's definitely developing, and this is our first 18 month, but I think... And I think it won't likely be one prototype that six- sticks across the 18 months, maybe it will be.
13:41
It will be things that are built on top and change, and it has, you know, the metrics on the AI side have definitely evolved as, you know, the length of the projects have go- programs have gone on.
13:51
I do think there is a part in all of this, um, which is, you know, part of the work we do, is sitting with the news organization and understanding what their goals are and how it ladders back up to whether that's the revenue side of the business, obviously, like time saved- Yeah...
14:06
does, and but, like, it's thinking about the contextual side of what are you then doing with that time saved as well that take- stretches it even- Yeah...
14:13
further, or how does that hit your revenue, and then where are you, like, reinvesting that time?
14:18
The challenge with all these AI-enabled or enhanced tools is that you can now build so much more so much faster, but you can also build the wrong things so much faster as well. Yeah.
14:27
In the structure of the program, because it is so focused on one idea, so they go through, um, yeah, this brainstorm and SWOT analysis where they nail down three ideas with attached metrics.
14:39
Then from those three ideas, and their mentor and other labs is working with them the whole time, they decide what project they wanna work on.
14:46
Um, you know, even for the 18 month, whether they pivot it, it still starts at a prototype, and once they've decided that, we help them attach metrics to that, if they've not done that before.
14:57
Um, and so by having a focus and then this mentor that plays that accountability coach role to keep them on track every week, um, and by, you know, we have, um, newsroom packs that they fill out where it's got, like, what their problem is, how they're gonna solve it, what are the metrics, what is the process to get the team organized, who are the key stakeholders.
15:20
So they're all, like, setting in place this plan before they actually move in. Um, and then with their team, they start building this, and they have check-ins every week. So it is...
15:28
The work I do is often very focused on one project.
15:32
If I'm working with larger publishers, it's, yeah, it's really, you know, I, I use a ranking system for prioritization based on, like, the goals of the business of, like, why that idea that, you know, the guy over in the photo team, um, has is actually foundational to the business and where you would, um, put that in your priority chart for l- the leadership team.
15:54
Yeah. Yeah. How are you seeing, or if at all, uh, I guess this new concept of, like, forward deployed engineers, so sort of bringing people out of the old product and engineering team?
16:04
When I'm doing, like, or the team is doing, like, that prioritization framework, like, we are talking to the technical people too. We're understanding it as a whole business.
16:13
Like, I think that is, again, it goes back to collaboration and, you know, those teams across the business working together.
16:20
Like, you know, I have my own business, it's a small team, and I would love to have a technical like you sitting next to me at all times because the things I'd be able to achieve.
16:29
Um, I also think it goes to, like, something I'm realizing more and more, which is about, um, using experience and, and experienced people in the right way in this new ecosystem.
16:41
And, like, what I've realized on the technical side is, like, I'm non-technical, so if I wanna build something, I'm better have a te- having a technical person build it.
16:50
If I wanna do a research report, I'm better having a research designer working with, like, the AI, Claude, whatever you're using.
16:58
Um, or if, um, you know, I'm an editorial person, you need the editorial knowledge in the room.
17:03
And so how I'm seeing, like, these businesses start to set up, that are the smaller ones, is with that layer of experience at the top, the AI underneath it, and then the layer of experience coming back in for, like, that human in the loop or evals or whatever you're doing.
17:17
Um, and so yeah, I think that's really interesting of how, like, the structure of organizations that we're seeing on the small scale, which will end up being the big scale, I imagine, over- Yeah... time look like.
17:28
So there's this wh- whole idea of individual AI where, you know, people in an organization are, like, super, uh, invested in, they're spending their weekends, they're sort of, you know, AI-pilled, uh, in, in learning all this stuff.
17:40
How do you manage that enthusiasm- [laughs]...
17:42
without sort of breaking the rest of the business so you don't have your, you know, uh, editor-in-chief out vibe coding a new web application that breaks all your security systems or whatever else?
17:56
I don't think there's an easy answer, but I think you should find those people.
17:59
Yeah And you should bring those people in, and you should take advantage of their excitement, um, for this new technology because they are going to be so much further than probably anyone else in your business.
18:11
They're the... You've essentially spotted an early adopter, right? Yeah. And those people you can learn so much from, um, and probably wanna share that information 'cause they're excited.
18:21
Yeah, and it goes back to what you were saying just before about, like, there's a lot of really useful information about how they're interacting, and, I mean, from our side, the way I think of that is, like, they've built, like, the cardboard cutout of what they want.
18:32
Mm-hmm. And now you take it to a builder to go build it. Yeah, and I do think, like, we need to be really careful with, like, what you're saying. You know, I'm definitely or have been AI-pilled.
18:42
I'm trying to claw my way out of it, um, a little bit and actually run a business, but, um, I will say there is this efficiency gap that's sort of happening, or a trap I should say, is that, you know, you can work...
18:57
I can work right now at probably 100 times what I could work before. I'm...
19:02
Yeah, like, you know, things that would take me months, I can do in an hour, and so I just am stacking those things [laughs] that you could do, or I was. Um, and- It's a bit of psychosis almost.
19:12
[laughs] Like, you just get into a trance. The flow. Yeah, yeah. It's like a writing flow for me. Yeah, yeah, exactly. It's, like, psychological. Um, and yeah, so through that, when you're working so much faster...
19:23
Like, it's fine for me 'cause it's my own b- business. Yeah.
19:25
But if you're within an organization, I think, like, leaders need to be aware of this, and that you start putting out at, like, x times capacity, but then there's just, like, being aware of it and understanding that give those people time to breathe, and pull them off for strategic time, um, because the people I'm seeing that are deep in AI are often, like, strategy people or people that like to problem solve and think things through, so they've actually got a lot of diverse talent.
19:52
So I do think, like, being aware of that, using those people to pull them out and help with strategy is useful.
19:59
And then just from, like, back to your original question of, like, how I would handle that in a, or how other labs handles it in a organization or in a market, is, like, spending time on that discovery.
20:11
So, you know, doing a, a user needs survey of, like, where, where your either people or where the newsrooms are at, um, on the... with their AI capability. Are they using, like, accounts that are personal accounts?
20:27
Ad hoc use is, is huge in every organization. Knowing what tools they're using- Yeah... and then un- getting insight from them on why they're using those tools, and then thinking, do you bring them into the business?
20:38
We can no longer just stick in this, like- Yeah, the amount of horror stories of, like, people uploading to free ChatGPT accounts, you know. Yeah. Confidential documents from someone's- No...
20:46
company drive to do a quick analysis- Yeah... and then realizing, well, you've sort of just trained a model on your own internal data. [laughs] Because people don't...
20:54
Your team probably don't think of the risks as much as, like, the leaders do.
20:57
So I think it's, like, setting a policy for your organization at the get-go, making that really clear on how that, um, you know, affects the people, and communicating it well, and working out where your line in the sand is- Mm...
21:10
'cause it's gonna be different for everyone.
21:12
And then, um, within that, providing the tooling that people need to add to your business in this really accelerated way if, you know, that's part of their role or what- Yeah... you want them to be doing.
21:23
Yeah, there's a really weird amount of compliance and sort of privacy stuff I'm sure is gonna explode really soon- Yeah...
21:30
where people are using one Claude or ChatGPT, whatever it is, account for a mix of their personal and- Mm... company stuff, and then when they leave that company, who own it?
21:39
'Cause once upon a time, it was like, well, you leave the business, they just get your email account and your, your folders- Oh, yeah... off Google Drive. Where now it's in your brain on your laptop. [laughs] Y- yeah.
21:46
And, and, like, whose laptop is that? Whose Claude account is that? Do you... Have you, like, uploaded your blood work to... out of curiosity, and now- No... the, the business that you were employed by- Mm...
21:55
i- is the custodian if they paid for it and whatever else. Uh, and all of a sudden, that's all m- you can't just, like, go in and delete the chats about, you know, when I got it to work on my family tree or something.
22:06
And now, like- [laughs]... you've got all this, and- All my health problems- Yeah... even worse. Yeah. [laughs] It's so new.
22:10
Like, uh, one of the CTOs, Nick, was saying that they, yeah, they go in when someone leaves, and they basically go into that person's account and then get Claude to then write, like, a, a handover brief- [laughs]...
22:20
almost out of that person's Claude account, and then they load that into somewhere else- Awesome... before they shut it down. But it's...
22:25
But if people haven't got a policy and they're not trained to be like, "Do not put in, like, your personal stuff because- Yeah...
22:30
this is a company Claude account," then anything in there is essentially should be viewed as discoverable- Yeah... by your- It's fascinating... organization.
22:38
Um- It's like the new social media policy, except it's not public. It's, like, everything you share privately- Yeah. Yeah, yeah... which is maybe worse.
22:44
Yeah, but yeah, as you said, like, policy and structure, and, like, actually thinking through these, like, big questions- Yeah... as an organization- Yeah... are, like, it's critical to work out, like- Yeah...
22:53
where do you stand on that? Yeah. What... Do you need a handover note? Do you, you ban it, or do you make it that what you put in is company data? Yeah. I, like... Yeah. There's some fascinating questions around- Yeah...
23:04
all this sort of stuff. Flipping, flipping to, like, the output, I guess, of media and things like that, adaptations of media are transforming sort of... and putting pressure on legacy players- Yeah...
23:13
how they've sort of worked, which, you know, the, the model was simple. It was page views- Mm... fed by Google, and monetized by, by, you know, programmatic or direct sold ads. Mm. And that was a sort of flywheel. Yeah.
23:24
How AI is putting a crunch on that and, and how people are trying to pivot. Yeah, I think, like... And it's not just AI. I think there's been, like, pressure on this for probably since that decade ago- Yeah...
23:37
um, with all the digital publishers. I think AI has just accelerated or just- Yeah... it's really put the pressure on.
23:41
But the reason that those organizations outside of social were able to get so much reach, um, was because they were removing friction with the audience. Um, and so, you know, you could reach them on Facebook.
23:52
It was, uh, also there was a lot of push, but there were, there was some pull going on.
23:56
But, um, and then over time, I think that disconnect has just continued, of there's this audience that needs information- As quickly as they can get it, and it needs to be factual, and it needs to be high quality- Mm...
24:12
if you have editorial standards that you believe in. Um, but now the friction to get that information is, is so high, and it's, um, you know, especially if you're thinking of a hard paywall. [laughs] Do you...
24:27
I mean, you and I have worked in multiple media companies. Like, how many ads and pop-ups to, like, look at a iPhone case, top 10 iPhone cases for the- Oh, yeah.
24:34
[laughs] And, and like, like, like these bigger players, that's because they've got s- commercial people, they're like, "Well, that's what we can exploit." Yeah.
24:40
They exploit in a completely legitimate sense, but, like, there's an opportunity, let's game it for SEO. We'll, we'll put out 15 articles about iPhone cases- Yeah... um, where maybe
24:49
creators, uh, and more identity-driven s- personality-driven stuff, uh, is more authentic and hasn't dipped into those games as much. They've had, like, a core narrative. Well, yeah, it's starting to get there, right?
24:59
Yeah. But, like, I think they had connection and a lack of jargon that people were drawn to. Mm.
25:03
Um, I think they were kinda like the new digital media where, like, um, same with podcasts and things like that, but I think, like, when you think about AI summaries and why a person is getting information that way, it's because they just want the information.
25:18
Yeah. And, like, you know, if they wanted to know what time the AFL's on, they would have to Google that, look through, which, like, people are hitting with SEO. Like, publishers- Yeah...
25:27
are hitting, but then you've got to read through a story to get to, like, what ti- and maybe it's not even there. Yeah.
25:32
And so, like, yeah, this friction got created over time by, like, bad ads or pay- now paywalls or- Well, that was the incentive, like the- Yeah... incentive structures And so, um, I think if you actually...
25:44
We're, we're in the business of serving the audience- Yeah... from an editorial point of view, and along the way, I think- Not, not shareholder value. Yeah. Yeah.
25:51
Along the way, I think there was a big disconnect by doing that, and that is why we've found ourselves in this situation.
25:57
So when I think about, like, what comes next, like, building a newsroom of what would you build if you started from scratch, but your goal was to inform an audience? Yeah.
26:04
And I think it looks completely different to- Mm... some of what we've seen today, and that is just an evolution over time.
26:10
Um, but I think we're at this, like, reckoning point now where we can't ignore what the audience wants anymore.
26:16
We can't ignore the technology and how far it's moved anymore, and, um, we need to really sit down and think what is valuable about a news business today. Yeah.
26:26
Um, and you know, from my perspective, it would be, like, primary sourced information, original content, um, real time facts, um, and- And trust. Like, that's, like- Yeah... that's, that's, that- Human verification...
26:38
that's, that's the ultimate thing that a big brand has that maybe a creator is more susceptible to- Yeah... uh, is... And, and definitely AI, is like, "Is this trustworthy? Can I verify it and validate it?"
26:47
Exactly, and that's gonna become more important, and you know, you're seeing... I'm seeing content people get hired for evaluations at tech companies. Like, they need that editorial knowledge.
26:57
That editorial knowledge is very valuable- Yeah... um, for high quality, like for a high quality ecosystem. Yeah. Mm.
27:04
Uh, turning to where it's gone wrong, uh, or, or the sort of trip ups, uh, [laughs] where- This is the fun bit... there's, there's a lot of these to pick from. Where are you seeing people get it the most wrong?
27:15
Um- Wrong. [laughs] Look, I'm not a, I'm not a designer, but I do have an eye for design. Yes. And don't look at my website, because that was d- I'm, I'm saying that was done a year ago, like it was pre-current AI. Yeah.
27:26
Um, because it is AI slop from Gemini. Um, but I would say now with Claude Design coming out, the amount of AI slop that no one seems to realize, AI UX slop- Yeah... that no one realizes, is, like, bothering me so much.
27:41
If I see, like, any more of those, like, pill buttons or the ticker, these are all on my site- Yeah... or the ticker or, like, the little green dot or the, the capped eyebrow- Yeah, yeah. Like, it needs to stop.
27:52
It's amazing how quickly that time from it sort of flooding the market to, like, on the nose. Yeah. [laughs] 'Cause AI writing took a while, like ChatGPT-isms- Yeah, yeah, yeah...
28:00
and Claude-isms, still highly annoying, but that took, that took a year or two- I also think-... to get to that point... the general public, or I, I think there's a mass of people that aren't even at that point yet.
28:09
Yeah. They haven't realized, and you know, for people that are maybe looking at this stuff every day- Yeah... it was quicker for m- for me to realize, 'cause I was in, uh, uh, Claude Design.
28:17
Well, we're sort of chronically online as well- I- [laughs]...
28:18
so we see this stuff a lot, and I think I was telling you, like the other day on, like, the Harvey Bay community page, some young person doing an anonymous post about, like, r- you know, "Young people like us don't look at your ChatGPT-generated flyers and won't support your business if you can't even pay a designer one hour to build you a better flyer."
28:34
[laughs] "Because they all just look the same." Yeah, that's fascinating, and I do think, like, one thing that sort of... You know, it gets on the nose, and then it starts to affect reputation of a business. Yeah.
28:43
And if I'm seeing, you know, that kind of website, and then I really have to work out quickly, like, do they have clients? Do they have an audience? Like, is this- Is it a synthetic business? Yeah. It is... Yeah.
28:54
Is this real, or is it just a front? And, um, so that's sort of, yeah, one thing I'm seeing at the moment. Yeah, es- especially if you're trading on value and taste and credibility- Yeah...
29:04
and all these things, and you've got something that looks synthetic essentially. Yeah. Like, it's a GMO website or a GMO flyer. Like, it's, it's not organic, uh, handcrafted, you know? No.
29:14
It doesn't have to necessarily be handcrafted, but that, that you've put in attention, you've put in effort, you've bothered about what you're putting out to the world. Yeah.
29:21
Not just something that looks like everything else, a derivative of all mankind. Yeah.
29:25
I think we think, and I definitely thought this when I was early using AI, that, like, I was being so creative, and I was adding my creative flair to it. Yeah.
29:33
But in fact, even if you're doing that, it's still often UX slop at the end. Yeah. And so, like, it's... I learnt the hard lesson that, you know, putting in your style guide for editorial or your design system- Mm...
29:46
as, like, a separate object- Yeah... that you put into the AI system first- It d- definitely de-slops it. Uh, yeah. Yeah, it helps in a way. Yeah.
29:54
It will still be a vac there, but yeah, it's just paying attention to these things so that you don't get stung.
29:59
To finish today, what, what are, are you guys most excited about where this is going for having worked in legacy media, challenges of, of legacy media, um,
30:09
from an audience perspective, and just sort of all the shifting patterns? Mm. What's motivating you the most getting out of bed and, and doing this stuff? This is, like, quite a moment for a tech optimist, [laughs]
30:21
which I'm self-declared.
30:22
There are real issues, and there are real fears, and there are real, um, things we're dealing with in media from, you know, copyright to monetization that aren't easy problems to solve, and I'm not gonna sit here and say I have the answers to them, and there's a lot of smart people working in all these different fields.
30:38
But I think collaboration as an industry with technology, um, which feels like the other sometimes, um, is critical to get through this, and working out what these new problems are and starting to look ahead rather than, like, being fearful and staying in the past.
