Ethical and sustainable marketing with AI

 

Hans de Kretser: Good morning, everybody, and thank you for joining us for this webinar on ethical and Sustainable Marketing with AI. My name is Hans. I’m director and founder at HDK, and I’m hosting today’s session.

At the start of the webinar, we asked attendees how they were feeling about AI in marketing. The responses included feeling conflicted, pessimistic, curious and concerned about AI replacing humans or its environmental impact. That gives us a useful starting point for today’s discussion.

To help us explore the topic, I’m delighted to welcome Brooke Gallagher. Brooke helps arts and cultural organisations build practical, confident ways of working with AI. She brings more than 20 years’ experience across arts, culture and technology, including roles at the National Theatre, Tessitura Network and Spektrix.

Through her consultancy, she works with teams on AI training, policies and everyday workflows, with human judgement and creativity at the centre. Brooke, what drew you to working with arts organisations on AI?

Brooke Gallagher: Thank you, Hans, that was a lovely introduction. so I, have Been working for the past, maybe. 18 months, 2 years, was doing a lot of internal work at Spektrix, and some with our clients when I was at Spektrix, and Was just spending a lot of my time interrogating and working with and learning about AI and the possibilities that it brings, and the challenges that many people have already captured in the chat already, which is great.

I think if you are sceptical about AI, that is a good place to be, a good place to start from. And, recently I’ve been doing training, and working on some longer-term projects with organisations like the Barbican, the National Opera and Ballet of the Netherlands. And Bristol Old Vic on a range of different, areas and, things that we’ve been building, so hopefully, as the session goes on, if they’re relevant, then I can share a bit more information about some of those specific things that we’ve been doing.

But I’m really delighted to be here, and I’m glad to see there’s so many people here, and A good amount of scepticism, which I think is a great place to start with.

Hans de Kretser: I also want to introduce Fizz, our digital marketing manager at HDK, who will share some of our experience of working with AI and using it in campaigns. Fizz, do you want to say hello?

Fizz: Hello everybody, nice to see you all this morning. Like Hans said, I wouldn’t consider myself an expert in AI, but it is something that we’ve been exploring, and we’ve been testing within the agency, so hopefully I can give you a few examples of how we’ve done that with clients.

Hans de Kretser: Our first topic is: how can we use AI for arts marketing in a way that’s ethical? We asked attendees where they would most welcome clear guidance on AI, covering accuracy and fact-checking, copyright and creative work, privacy and data, and transparency with audiences. Copyright and creative work came out clearly on top, with privacy and data also ranking highly.

Brooke, does that chime with what you would expect?

Brooke Gallagher: It makes it makes a lot of sense. I think, many organisations that I work with and speak to, the real hard lines are around that creativity and around, copyright and things like that. And actually, I think that is an area where, as an organisation, as an individual, you Make a choice about where you draw those lines, and many organisations are happy to use AI to check assets and check Content, but not to create it, and that’s, a good line in the sand, because AI can be really good at helping to check on those things.

But I think with those things that, around creativity and copyright and critical thinking also, really importantly. I think that’s where we can make a choice how we use the AI, and it’s not it’s not all or nothing, we don’t my the thing that you’ll hear me say over and over again is that If you get energy from something, if you get joy from it, if you feel creative, if it challenges you. Don’t give that to the AI.

Give the AI the stuff that causes you friction, that drains you, that you get that you get no joy from, and I think probably a lot of people got into marketing because they enjoy writing, and they enjoy being creative, and so Hold on to that, and use AI for those things that feel less creative or bring you less joy, like Do you love writing an ad for Meta? Maybe. So if you love it, keep it, but if you if that’s just, a task on your list because it’s quite formulaic and it’s quite factual.

Maybe that’s something that AI, can help you with.

Hans de Kretser: I would agree, and I think I guess it’s obvious that, an arts sector, we would be championing the creative and creative, the copyright and the creative work. It’s something we feel probably very close to, and it’s probably, something that we would want to protect, I would have thought, as well. So I guess that is the obvious one to come out at on top there.

So, in terms of What the what we think about as ethical in AI for arts marketing. Are there some broad topics that we probably all want to be thinking about when we when we think about this? I guess, for example, from HDK’s point of view.

We’re about to review our internal AI usage policy, for the third time, so, it’s it’s probably the thing that we’ll I looked at it the other day, and I felt, well, this last time we wrote it, it felt quite broad. But probably missing some of the detail, what are the topics maybe that we might be thinking around this topic? A subtopic, if you like.

Fizz: I suppose just adding to the creative. Conversation that we were just having that Brooke mentioned, in terms of how we’ve balanced that at HDK, I tend to for my clients, I never create images or videos with AI. However, what I have been doing, because I wouldn’t say that I am the world’s best video editor, for example, if I have a really long piece of footage, using AI to help me to find the best way to cut it, so it can fit a certain time, for example.

And that’s something that I’ve been exploring. With a few clients, and building out Making that creative process maybe easier, or to remove the sticking points that we’ve found along the way.

Hans de Kretser: So, I mean, what some of the topics are covered in the poll, I guess. Are there any other things that we might be wanting to think about, just generally, top level, in terms of, when we talk about ethical, what’s ethical here?

Brooke Gallagher: I think that’s really that’s the question, is it? what does ethical mean to you as an individual? What does ethical mean to you at your organisation?

It there’s probably a spectrum of where people fall on different aspects of ethics, so having it with your AI policy, or if you have some charter with guidelines and principles, that can really help your team think about what that means for you in your specific in your specific context, in your specific situation, but I guess maybe one of the things that I think should run through.

All AI work is that the accountability for the output, that sits with you, that sits with the person that produced it, and I think we are seeing, and I’m hearing a lot of stories of Sometimes maybe more junior members of staff who are producing it’s so easy to produce so much stuff now, and we’re all just producing stuff and giving it to other people who aren’t reading it, but they’re producing stuff in response, like. This is this is not the way this is not the way forward. How do we how do we help team members really, identify when what they’re producing is that AI slop?

How do we How do we provide training and support for team members to sometimes, I mean, this happens with me as well, sometimes you can be really led by the AI, and sometimes you forget that you’re the boss, you’re the expert, you’re the one in charge. So Sometimes you just have to remind yourself, oh, I don’t I don’t have to go down this particular path. the accountability is with me, I need to use my critical thinking, I need to use my judgment, and I think, especially with arts marketing.

Maybe different to, I don’t know. Coke marketing, not sure, but, authenticity really matters, and I was saying to the team before we started, you see so many Quick Around the internet, on YouTube, that, well, you can just get AI to learn your voice, and then you’ll never have to write anything again. And I actually I just we probably will get there, but I just don’t think we’re there yet, in terms of, capturing the authentic voice of an arts organisation with principles and values and a desire to make a good impact in the community.

I think that’s really important. I think, ethically, that authenticity needs to sit with humans, and if there are other things you can produce.

That are less consequential, so if there’s a series of ads that are formulaic and very prosaic, or if you’re sending out an email that’s just very factual, you get no joy or energy from it, it’s not for major donors, it’s not for VIPs Maybe it’s, it’s 5 to 6 on Friday, you just need to get that email out to friends and family, because the matinee hasn’t sold tomorrow, and it can be very factual and straightforward. maybe you can use AI for that, and that still sits under the ethical, umbrella, but I do think as an organisation, as an individual, you need to find where your lines are and what ethical really means for you.

Hans de Kretser: One question raised was how we make sure AI use doesn’t entrench sexism and racism that may already be built into the systems it learns from. That brings us to representation and accessibility. Brooke, how should organisations think about that?

Brooke Gallagher: so these systems are built they’ve read all of the internet, and a lot of the internet is white, Western, and male, so There is definitely that, consideration, and I think we what we can do as users of it is to mitigate that, so building into your custom instructions. If anyone here doesn’t have custom instructions built into their preferences in their AI If it’s Copilot, GPT Claude.

If you don’t have custom instructions built in with your preferences and your principles, exactly like you’ve said, Liberty, things like I want you to always be looking out for inherent bias, I want you to challenge anything that comes across as, white Western male. You can build those into your instructions, you can build those into your checks, when you’re checking for works. If you’re setting up somewhat semi-automated checkers, because that’s quite that is quite a good use of if you can set up all your rules.

For your content. Don’t use it to generate, necessarily, but use it to check, and also pass those tools on to more junior members of staff, so they can learn, what, errors to look out for before it lands on the Director of Marketing’s desk. You can build those things into your checking, into your the backbone of your custom instructions.

Then, that’s one way to mitigate it, but I think acknowledging it, knowing that it’s there, and making a choice about how we use it. Is really important.

Hans de Kretser: Another ethical question is whether, by using AI, we’re replacing other people’s work or jobs. For example, if we use AI to proofread, does that put a proofreader out of work? The same question can apply to translators, video editors and other roles.

Brooke, how should we think about that?

Brooke Gallagher: There’s a lot there’s a lot of conversation about this, and there are lots of really different, really striking viewpoints. And I will admit to all of you, I am not a future Futurist, I consume, the thoughts of many futurists, and there are There’s just so much debate out there. There’s not currently any, There’s not currently any statistics showing a huge amount of job loss yet, but then that job loss is always going to track Behind the development of the technology.

There are many people out there that believe that, yes, some jobs will be lost, but different jobs will be created. Some people are very evangelical about the fact that this is going to produce more and different types of work, and if we do short-sighted things, for example. There is some data out there to say that there are fewer positions for junior. coders, junior engineers, for example, and if we if companies stop hiring those junior people, how will it how will we ever get to the senior people.

So, there needs to be consideration about that, and not that short-term thinking, and what the solution to that might be in the long term. It might look different, it might be more, deliberate. education programmes, or training within companies, or, I don’t know what the answer is, but there’s there’s definitely a lot of very smart people who think that jobs will change, like.

I don’t mean to be trite, but, a couple of hundred years ago, it was a job for someone to knock on your window at 4 in the morning to get you up for work. there was a job of someone who brought the ice to you for your ice box. I mean, I don’t really I don’t mean to be trite, but jobs do change over time, and people upskilling, and really focusing on those really human skills, like judgment, strategy, critical thinking, interpersonal relationships.

Why did we get into the arts in the first place? To form, connection with other human beings. I think that’s why many of us, love working in this sector.

So those skills also, really, really important. But there’s so much debate on it, and I would hate to I would hate to place a bet on any of the horses in this race.

Hans de Kretser: The next topic is sustainability: is it a legitimate concern, and what can we do to minimise AI’s carbon footprint? There is understandably a lot of concern about the environmental impact of AI and the infrastructure behind it. Brooke, how do you approach that question?

Brooke Gallagher: I do, and so I think There’s a key difference here between the question of should AI infrastructure be sustainable. The obvious obviously, the answer to that is yes. Should we be putting pressure on the AI labs to make their data centres as efficient as possible?

Yes. And, you’ll see at the moment there is legislation. In some places coming in like New York.

Where they have put a pause moratorium on building new data centres. Putting clauses in for that the AI companies need to be funding or finding their own energy sources, or finding more. More efficient and effective ways of experimenting with running their centres in different ways, yes.

And I very much support that. I’m actually very much in favor of the regulation, that the labs are actually Calling for on themselves, 100%. But that, is very different to, should I not use an AI tool that’s making me more efficient, more effective, because of its environmental cost, because As an individual, even as an organisation, if you look at your The environmental, impact of you Using an AI tool.

It’s got it’s likely to be less than 1%, of your,. The footprint that you have on the environment every day. So if you really want to make a difference with the environment, there are lots of other things you can do.

I will share this calculator, which, again, so this is just A tool. It’s, very difficult to estimate These types of things, but this is a tool that someone called Andy Maisley put together. He is an effective altruist, so and he’s a vegan, and he’s liberal, so he’s not exactly shilling.

For, OpenAI or Anthropic or anyone, and he’s taken some statistics from an, Ecologics, who are an open source Company who are trying their best to put a measurement on what it means to use AI. But for me, as someone so using this calculator, I’m someone who lives in the UK, in a medium home, I drive an average amount, I eat a vegan diet, and I fly sometimes. And I probably use AI more than most people on this, on this call.

That’s 0.14% of my daily carbon footprint. All the caveats, all the caveats, all the caveats in the world about, about estimation and everything, even if that’s out by a factor of 10, there are so many other things that I can be doing to make much more of an impact on the environment as an individual, so It that my AI use is less, in terms of carbon, than having a cup of coffee.

Almonds, which is rough, because I need well, I need to get my protein, right? So it’s just an interesting tool, I think, just to put it in perspective. I don’t We absolutely need to be holding these AI companies to account for everything, and Definitely the poor ways they’ve behaved in the past.

But I think as an individual, as an organisation, there are probably other areas where if you really want to become more sustainable, if you really want to make a change. AI’s the impact of not using AI is negligible compared to other things that you could be doing. Super hot take, I hope I haven’t upset anyone with that, but I think it’s important.

Important, but into perspective.

Hans de Kretser: Thank you, Brooke. One point worth highlighting is that AI is now built into many everyday tools by default, so we may be using it more than we realise. I think and, someone else said, you have much more of an impact on the environment if you are a meat eater, and if you use an AI tool.

And thank you, Dean. And that resonated with me, because we were literally having a conversation, last week with, a colleague. Who’s a meat eater, and was making a very convincing argument that sometimes that is maybe over overstated.

So, I think what we find is, trying to navigate this, we don’t have a black and white,. Guidance on what to do here, and I guess what it becomes more personal and trying to find the ways that you can contribute, and, I suspect that, it’s very it’s going to be harder and harder to resist Using AI, as your colleagues are using it around you, or your competitors. So it’s I think that’s going to be really what a real one to challenge.

So, yeah.

Fizz: Maybe I’ll just add a, another point of view, because Many of you, I think, when AI first came out, I was really anti-using it, actually. I really didn’t want to Touch it, because of the environmental impact that I believed it would cause, but it’s really interesting to hear Brooke Talk about it in that way, and I know that near the HDK office, actually. There was a data centre.

That’s being built, on Brick Lane, not specifically for AI, but it’s still quite a substantial data centre, that lots of people have been protesting, so I think that Was it Maya mentioned the impact on local people as well? We have to build these data centres. There’s lots of discussion, to be had around that.

But I really value a different viewpoint that Brooke’s bought, I suppose, and to put it into context is really useful.

Hans de Kretser: I get I get the point. I guess there’s also some people might be thinking, well, yes, I individually Won’t make an impact just by using AI, but if everyone’s doing it, what is that cumulative effect? So I think, some people will be thinking about that, and Hoping that they’re, if they’re making a choice in one direction, other people will be doing the same thing, and that together we can have that impact as well.

But ultimately, I’m very curious to see how, as time goes on, people will be able to avoid using AI, because as was said, it’s being built in more and more to the tools that you’re already using. I have to manage the company’s accounts, and Xero has AI built into it, and there’ll be all these sorts of tools that I’m using that are, adopting AI as part of what they’re offering. So, the challenge is definitely there about how you can resist it, and I think, I think ultimately what you’re saying, I think what I always felt before as well.

But I. Brooke was saying, we need to hold these companies to account, absolutely. I think I was a bit more, maybe, naive when I was thinking about this.

I was thinking, well, they will sort it out. Maybe they won’t sort it out without the pressure from us. But I was always I’ve always been of that opinion that, we AI will help find the solutions to some of these environmental issues that we’re struggling with.

But clearly, I think we have to have a drive to encourage people to push that way as well. Okay, so let’s move on to our third discussion, today. And, this one is about. “Are there any AI tools or workflows that could be particularly useful for anyone marketing a cultural organisation or event?” So, are there any AI tools or workflows that could be particularly useful for anyone marketing a cultural organisation or event?

So I think, yes, let’s get something going in the chat again. So what’s one marketing task you’d like AI to help with, or one you’re already using it for? So this is a chance To share the ways that you’re using, if you are using, I know not everyone is using AI, but if you are, how are you using it, or how would you like to be using AI to help you with your marketing work?

So Ellie jumps in with editing and improving product captions. And, Katie, says we use Tessitura, and there is AI coming in their analytics, which can help forecasting. Improve SEO sorry, it’s hard to keep things, because they’re coming in so fast.

Improve SEO readability and increase reach. As a challenge to come up with possible pushback arguments against our recommendation ideas, so using it for critical thinking and pushing back on ideas. Transcripts from audio and video editing.

We have experimented with AI to help optimize our website pages so they appear in AI search results, and this is something we talked about in a webinar a few months ago. Digital marketing reporting, data gathering analysis, yes, and we talked about how Google Analytics has got some, really helpful AI, tools there. Adding subtitles, we’ve checked, after, but was always very efficient to get that started, I presume.

So patterns, analysis, diagnostics, alt images for, alt text images, for websites, again, something that we were talking about, in a webinar a few months ago. Taking complex information from research that is outside my experience and turning it into engaging messaging for comms channels, yes, that’s something that I’m I will often do. So those are still coming in.

I wonder, Brooke, whether you’ve got some, maybe,, examples that go into, maybe, a case study in a bit more detail that illustrates how maybe people that you’ve come across are using AI these days.

Brooke Gallagher: Sure. I’ve turned my video off because

Brooke Gallagher: So, because of the issues, I’m so sorry, you don’t get to see my smiling little face. But so I’ve worked with quite a so I quite often find that the marketing teams are actually some of the most Proactive teams in terms of adopting, because the use cases are pretty vast and pretty wide. So I think, AI is really good at things and these have come in the comments, but analysing, researching, checking.

Create ideas generation, like we are obviously all creative individuals, but the AIs have read all of the ideas, so I don’t have all of the ideas of the world in my head, but the AIs do, so they can be quite good at, coming up with interesting ideas, especially if you’re feeling tired at the end. End of a long day of meetings, and you need to produce, it’s fun to collaborate with them, but obviously, again, you’re the boss, you’re the expert, and really good for automating as well.

So, some of the things that I’ve seen Obviously, you’ve probably got some great dashboards that you have access to based on your whatever systems you’re using, but what AI is really good at doing is taking a lot of disparate data sets, and that could include Reporting that has lots of numbers, that could include feedback that is natural language. it’s good at passing. Different types of information and pulling themes together.

It’s really good at identifying through lines and connections that may be As humans, we cannot pass that amount or that variation in data, so good at finding through lines, really good at looking at, audience behavior and Finding connections between lookalike customers or lookalike shows that, that you didn’t you didn’t think were there, so maybe you have a production of Carmen, and you think, well, if they’ve come to Carmen, they’ll come to Tristan and Isolde.

But maybe the AI picks up on the fact that Carmen, that’s quite saucy, that’s quite exciting, maybe they’d also want to come to rent, for example. those kinds of things is actually really good at pulling out, connections that we haven’t seen. In terms of real life.

Use cases that I have come across and worked with. There is an opera house in Paris, and they use Claude, and they’ve used it to do that thing that I mentioned earlier. They’ve built a lot of skills in Claude, which are looking for very specific Erin.

There’s typos, brand, violations in the work, so they create all of the work, the humans create the work, the posters, the copy The content, but before a piece, an asset, like a new poster, reaches the desk of the marketing director, they run it through their Are they using the right hex code?

So it looks for all of those types of things, so they’re not they’ve really reduced the friction around Those This feedback loops, and getting caught in those feedback loops, and it frees up the more senior members of the team to really look at it strategically, rather than saying, that’s not how you spell Rakamananoff, for example, or it’s not on in July, it’s on in June. So that’s, a really simple one out in the wild. I’m I’ve also, there’s a commercial theater company in London who’ve built a couple of things.

They use ChatGPT. And, they do have a, a non-consequential email builder, which you just drop the name of the show and what tickets you’re trying to shift. And it will create an email, really simple, really straightforward.

A non-consequential email that’s going out to friendly people just to say, look, these tickets are available. They have also built an in-house HTML previewer in ChatGPT, so that means that Because if you use ChatGPT Business, or if you use Copilot, that falls under Microsoft and OpenAI.

Enterprise data protections, they don’t train on your data, but what that means is for that organisation, they have got their HTML checker, and all of Everything they’re putting into that checker is now staying in-house as protected by Well, as protected as any system is, let’s be very honest, any, all systems are penetrable in some way, shape, or form. But in terms of, as strong as being in a Word doc in an Excel file somewhere, so they’ve kept that in-house.

Another really interesting one that I’m working on at the moment, it’s more in the fundra Phrasing aside, but there is an organisation in the Netherlands who they have very big premiere nights, so opening night, and they’re There are seat maps for those opening nights, lots of VIPs, lots of donors, lots of press. 150 to 200 guests at every one. That takes days and days for a human to seat, so we’re looking at, I don’t think AI will ever get that perfect, because who sits next to who and where in the auditorium?

There’s a lot of human judgment that goes into that, but if we could get it to do a first pass, and it can do 75% accuracy, then that’s at least reducing the friction and enabling those fundraisers to spend their time fundraising, rather than Administration, around a seat map. And then also a couple of other, more general examples, which you might find useful. Amgueddfa Cymru, Museum Wales, they use Tessitura, and they’ve built a in-house, agent, where if someone’s got a question about how to pull a report, or how to use different functionality.

In Tessitura, they can ask that Copilot agent, and it will look at their internal documentation first. Before it then if it can’t find the answer, then it will go look at, Tessitura, documentation more the more generally available. Oh, and just speaking of, Tessitura, and Spektrix, and TicketSolve, and Lineup, and all of these.

Great CRM ticketing companies that you’re probably some of you are working with. They are also starting to roll out a lot of those AI tools. So, someone mentioned earlier, they’ve got, you can chat with your data in Tessitura Analytics.

Spektrix and TicketSolve are bringing in very similar, so you’ll just be able to chat to your data, rather than Running complicated reports or queries, or needing a database query, you just need an answer to something, you can run a natural language query. And then Lineup, if anyone’s using Lineup, they’re doing some very, really interesting things around being able to build out Build out shows and seasons. So that, if you use Lineup, or if you go and have a look on LinkedIn about some of the very exciting things there.

What you’re doing.

Hans de Kretser: Thank you, Brooke. Yes, Fizz, anything to add to that? There was a lot of stuff there, so and I know we’re doing some of that here at HDK.

It can feel overwhelming because there are so many possibilities. To me, it feels like The opportunities, the, what we can do is absolutely endless, at the moment. I’m not finding those boundaries yet, but let’s try and bring it into practicalities that, we might be doing on a day-to-day basis.

Fizz: I suppose I would agree that, that can seem a bit overwhelming, and I’m definitely not using it in such a sophisticated way, at the agency. I think where I find it most useful, and where also I see the real benefits are at the very start of a project. helping me to plan and organise, and if I’ve got quite a complicated timeline, for example, feeding it in, feeding AI my usual process, and then getting it to plan the actions that I need to make.

That I need to do, has been really useful for me in breaking it down so I feel less overwhelmed. And actually, something that Hans showed me the other day, that I tried for the first time recently, was using AI and automatically putting those actions into Asana, which is our project management tool that we use here at HDK. So that saved me quite a lot of time, because usually I would go in and write all of those out individually.

So I definitely find that. really useful, and then at the end of a project, I think Brooke. Talked about this really well, but I definitely Also use it as a second voice.

I think that once I’ve finished a piece of work. I find it particularly useful to feed it in, and get it to assess what I’ve done. Have I missed anything based off the initial brief, for example, or the original goals that we had for a project?

So I definitely find the start and end. The points where I bring AI in the most, and then that middle chunk is more so my own my own thinking and my own thoughts.

Hans de Kretser: Brilliant, thank you. Thank you, Fizz. One useful distinction is the difference between traditional AI and generative AI.

And if anyone’s wondering, I guess traditional AI is analysing information to classify and predict and recommend things. So it’s using its vast data to predict those next steps. So it’s the, so you might use it, as with content, which content is going to engage, your audiences more.

Generative AI, as far as I understand, creates new text, new images, audio, and other materials in response. To a prompt. These are examples for great are great for large organisations, but what about small teams, non-venues, that don’t have the resources to develop all this knowledge without having an impact on their already very busy workloads?

I would say Just from person from my personal point of view, we’re a small team at HDK. It just needs someone to champion it a little bit and to find some time to invest in it. But it’s it’s something that you take at your own pace and you just try things.

You start off with small, simple things, and then and then you just move on to trying different things and just being inquisitive. And wanting to explore the options. I’ve gone from just using it to come generate ideas to using it in so many different ways in my personal life as well as work as well.

And as Fizz says, there’s always something new to learn. It’s just, I would say, is not be scared of it, but just have a go, just keep on. And it’s like all the other technologies that have developed over the years while I’ve been working. working in arts marketing.

It’s that Just giving it a go, trying it, exploring it, pushing it further, getting in there soon, because over time. Things are going to change again. So you just want to just, just keep plugging away, I think.

So I don’t think it’s about whether you’re a small team or a large team. I think it’s quite democratic in that sense. There’s something there for everyone to take advantage of.

Brooke Gallagher: And I would just add Hans, having worked with large teams and very, very small teams, it is a challenge for everyone to find the time to find your way with AI, because AI is not like Excel, you don’t just learn, this is how you type a VLOOKUP. The.

Joy and the promise of AI is that you teach the AI about you and how you work, and where your strengths are, and where you might need support, and it can help you with that, but that does take time, and It really does help having a dedicated, time slot, or if you have those if you can curate one or two cultivate, sorry, one or two internal champions who could run some sessions. There’ll definitely be people in your organisation that are experimenting and exploring this stuff in their own Lives.

And so, maybe bringing some of that in, just it is really difficult to dedicate time to this, because arts professionals are We know are already over capacity. Most people are doing the jobs of one and a half, two people, so I can completely empathize with that. But.

The gains that you may see, if you can invest a bit of time in, to take some of that friction, some of that really tedious stuff, some of that repetitive stuff, some of those errors. It really could be transformational, and that is what I see when I,, when I go in and work with organisations.

Hans de Kretser: I think that’s a really good point and I think that chimes with me in the sense that, you know. If myself as a business owner, one thing I’m taught is to, to carve out time to work on the business. Rather than in the business.

On the business rather than in the business. And in the business is doing your day-to-day stuff, but every now and then you’ve got to carve out that time to think, well, what are the things that I need to work on that will move the business forward? And I think that applies to any job in many ways, and we try and do that.

At HDK with a whole team in terms and we have a system called OKRs, which is about setting yourself quarterly objectives, and then working towards that. The idea of those is that we look for things that will save us that time, that will improve our day-to-day job, and make things easier for us. And I think What we’re finding is there is so much with AI that it can help with different elements of our job.

So, if you’re finding that you’ve not got time to spend learning AI. The my response is, well, if you can find the time, you’ll probably ways that you will be able to save your time. I’ve not articulated that very well, but I hope you’ve got what I’m trying to say there.

There was another point, but I’m sure it’ll come back to me. I. I actually have a question I would love to ask, which is, Brooke, if you think that one AI system or platform is better for smaller teams.

Than another one, or maybe more ethical as well.

Brooke Gallagher: Oh, that’s oh, that’s very interesting. I’ll be blunt, I don’t know if any of them are particularly what I would describe as ethical. you’ve got your.

You’ve got Anthropic taking somewhat of a stand against mass, monitoring and use of automated weapons in war, so that is somewhat of a stand. So many people moved across to Anthropic, for that. But I would say, so what I come across is, most arts organisations that I’ve worked with and cultural organisations are Microsoft houses, and so if you have Microsoft, you probably at least have Copilot Basic.

Am included, and I think Copilot has got a lot of, flack in the past for being not the not the best, but there is a lot that you can do with, even with Copilot Basic. You now have In Excel, in Word, in PowerPoint, you have the power of Copilot, the full power of Copilot in those apps, and I think probably for a small organisation, if you’re building PowerPoints, if you’re in Excel, if you’re using I’m not really one to advocate for AI for writing, but for definitely for Analysing, for creating dashboards in Excel, it is absolute magic, even Copilot Basic in Excel and in PowerPoint.

Go and use it there, it will genuinely make a huge difference. And so, I’m seeing that, and then if an organisation is prepared to invest a little bit and bump it up to Copilot Premium, it is an investment, but then you have the full context of your work Is right there in AI in Copilot. You have access to all of your files that you would have access to.

Only files you have access to in SharePoint or in OneDrive, but you have access to all of your work within there. There’s not this uploading, downloading, connecting this out. And the other, it’s all there and it’s all connected.

And the power of Copilot Premium, it’s what I’m working on mostly with organisations at the moment, even the smaller ones are finding that,, that upgrade fee per month really pays off in terms of efficiency. And getting stuff done. But I also come across brilliant organisations who are, so if you’re using something like Claude or ChatGPT, then you can take advantage of these other features where it will go and it can do a lot more for you, with the Microsoft, Copilot.

You would have to pay again for those features, so at the moment, I don’t really work with a lot of organisations who are doing that, but, if you use Claude or GPT in your personal life, you see the power of it going and Doing actually doing work for you. And again, I would say, get it to do the work that doesn’t bring you joy, doesn’t bring you energy, don’t get it to do the work that challenges you or excites you, but getting it to go and do actual work is the power of if you’ve if you’ve chosen to use GPT or Claude. Is that helpful?

Is that was that did I prevaricate and not really give you a clear my personal look, at the moment, I think ChatGPT, despite all of the issues, look, I definitely am no big fan. Of that gang, but the power of GPT at the moment is pretty impressive, but this changes every few months, and every 3 or 4 months, I’m switching my I have to have all of them, right? I have to understand how all of them work.

Every 3 or 4 months, I’m switching who my favourite changes every few months, and it was Claude for a long time. GPT is in the front for me at the moment, but I’m sure they’ll switch back. And then I’ve met people CEOs of digital agencies, no less, who are predicting that no matter what happens in the middle, Google will always win in the end.

So, we could all end up using Google in a year or two, who knows? But that’s my that’s my that’s my two cents, and that’s my,, my experience of working with arts and cultural organisations.

Hans de Kretser: Fizz, sorry, you go first.

Fizz: No, I was just going to say that was really interesting to hear your thoughts on that, because I know there’s been conversation in our office about which one is better, or which one we prefer using, etc. So really interesting.

Hans de Kretser: And there are people in the office who have team Claude. I’m with Brooke on this one. I’ve come back to be team ChatGPT, but I absolutely agree that it’s changing all the time and you need to keep your eye on different ones.

I’d love to carry on talking, but I’m very conscious of the time.

Thank you, Brooke and Fizz, for all your insights and useful information. I think that has been really illuminating, and I hope it has been useful for everyone joining us today.