Latest / How I Work / AI is a terrible brainstormer (until you do this)
Transcript
- 0:00Speaker 1: So you ask AI for some ideas because you're doing
- 0:03Speaker 1: a bit of a brainstorm and it gives you back 10.
- 0:06Speaker 1: It tells you every one of them is brilliant and
- 0:11Speaker 1: you're flat out. So you just grab the first one
- 0:13Speaker 1: that looks half decent. I mean, we've all done it,
- 0:17Speaker 1: but the catch is that AI drifts towards the average.
- 0:22Speaker 1: It really pushes back and it's very easy to let
- 0:26Speaker 1: it do the thinking for you. So you end up
- 0:29Speaker 1: with ideas that look a lot like everyone else's. In
- 0:34Speaker 1: today's episode, Neo and I share how we use AI
- 0:37Speaker 1: as a brainstorming buddy that actually sharpens your own thinking.
- 0:42Speaker 1: You'll hear the four techniques that we use at Inventium,
- 0:45Speaker 1: and you're also going to learn how to get AI
- 0:48Speaker 1: to put them to you so you're the one coming
- 0:52Speaker 1: up with the ideas. And Neo also shares a bonus
- 0:55Speaker 1: prompt that makes AI ask something of you before it
- 0:59Speaker 1: hands you any kind of idea. By the end of
- 1:03Speaker 1: this episode, you will have a way to brainstorm with
- 1:06Speaker 1: AI that gives you ideas built for your business, your
- 1:10Speaker 1: customers and your goals. Welcome to How I AI with me, Dr.
- 1:21Speaker 1: Amantha Imber and Neo Applin, head of Inventium AI. Each episode,
- 1:26Speaker 1: we share one practical way to use AI better at
- 1:29Speaker 1: work and in life. No fluff, no tech jargon, just
- 1:33Speaker 1: things you can use straight away. So Neo, one of
- 1:36Speaker 1: the things that we know about good brainstorming practice is
- 1:41Speaker 1: that you should start solo because as soon as you
- 1:44Speaker 1: start involving other people or AI, You start to get
- 1:49Speaker 1: influenced and taken off in all these different directions. So
- 1:52Speaker 1: I'm curious, what are the problems that you're seeing people
- 1:56Speaker 1: encounter through ineffective AI use for creative problem solving?
- 2:02Speaker 2: Well, for one, AI thinks every one of your ideas
- 2:05Speaker 2: is awesome, doesn't it? Yeah, well, mine are, I think. Yeah, well, absolutely.
- 2:10Speaker 2: Maybe you are awesome on ideation. Certainly my AI thinks
- 2:14Speaker 2: that everything I do is a great idea. It doesn't
- 2:17Speaker 2: usually push back out of the way, out of the gate.
- 2:20Speaker 2: So therefore, that is a problem. The other problem you've
- 2:23Speaker 2: got is we as humans are busy slash lazy. And
- 2:26Speaker 2: so when AI gives us a couple of brainstorming things,
- 2:29Speaker 2: oh great, they're awesome, and then we'll just pick them.
- 2:32Speaker 2: There's a third problem as well, which is AI kind
- 2:36Speaker 2: of tends to go towards the average because it's got
- 2:40Speaker 2: all this knowledge and all those kind of things, but Also,
- 2:42Speaker 2: the kind of ideas that I would get with the
- 2:44Speaker 2: kind of problem using AI, you're going to get the
- 2:46Speaker 2: same kind of ideas out of AI as well. And
- 2:49Speaker 2: you are not me. Your business is not our business.
- 2:52Speaker 2: Everyone's quite different. And our needs, our customers, our goals
- 2:56Speaker 2: are really different. And so just getting AI to do
- 2:59Speaker 2: the brainstorming is really generic and you won't actually get
- 3:03Speaker 2: something that's really going to move the dial. It'll be
- 3:05Speaker 2: probably better than you just not doing much. But if
- 3:08Speaker 2: you really wanna get a great brainstorming session, there are
- 3:11Speaker 2: better ways to use AI.
- 3:13Speaker 1: Okay, so tell me, what are your top tips for
- 3:18Speaker 1: using AI as an effective brainstorming buddy?
- 3:22Speaker 2: The first is, and this is kind of principles for everything,
- 3:25Speaker 2: you need to tell AI what the whole point is
- 3:27Speaker 2: and what you're trying to achieve. So the context really
- 3:29Speaker 2: matters here. So if you've got no ideas, but you've
- 3:33Speaker 2: got a goal, then tell it the goal that you've got. Also,
- 3:37Speaker 2: I feel that most problems when we're doing a brainstorm,
- 3:41Speaker 2: it has an audience. There is a reason why we're
- 3:43Speaker 2: doing it. We're doing things for people. We don't work
- 3:46Speaker 2: for robots, maybe yet, but certainly we do things for people.
- 3:51Speaker 2: So how do I do this thing for this group,
- 3:54Speaker 2: for this outcome, for these people? and have a good
- 3:57Speaker 2: chat with AI so it knows why you're doing that. Also,
- 4:01Speaker 2: if you've tried things before, you've got a kind of
- 4:04Speaker 2: like a limitation, all those kind of problems you might
- 4:07Speaker 2: be dealing with in your brain, get them into AI.
- 4:11Speaker 2: And we've mentioned this plenty of times before, dictate to
- 4:13Speaker 2: your AI. It's easier to get this information out. So
- 4:16Speaker 2: tell it your goals, tell it your audiences, but then
- 4:19Speaker 2: the most important thing is don't get it to brainstorm.
- 4:24Speaker 2: The reason for that is it's going to give you
- 4:26Speaker 2: 10 ideas. Now, maybe you'll realize that eight are terrible
- 4:29Speaker 2: straight out of the gate and you might go, yep,
- 4:31Speaker 2: two and four are pretty good and you can have
- 4:32Speaker 2: a conversation on that. That's not bad. That's fine. But
- 4:39Speaker 2: I prefer going a little bit deeper on that. And
- 4:41Speaker 2: what it is, is I'm using AI often instead of
- 4:44Speaker 2: just coming up with the ideas for me. It's actually
- 4:47Speaker 2: a brainstorming buddy to help me come up with the
- 4:50Speaker 2: ideas and then we have a conversation about it. And
- 4:53Speaker 2: the ways that I get it to do, I've got
- 4:55Speaker 2: four techniques that I use in my brainstorm buddy that
- 4:58Speaker 2: we use at Inventium. And we're happy to tell you
- 5:02Speaker 2: what they are now if you're interested.
- 5:04Speaker 1: Let's do it. Yeah.
- 5:05Speaker 2: All right. So a constraint. So effectively get AI to
- 5:12Speaker 2: put a few constraints in here to make you think
- 5:15Speaker 2: a bit differently. So instead of having $ 20 million to
- 5:19Speaker 2: make a movie, what if you had $ 10, 000 to make
- 5:22Speaker 2: a movie. Those kind of constraints come in. And we've
- 5:26Speaker 2: seen that often constraints increase creativity because you wouldn't find
- 5:31Speaker 2: another way around it.
- 5:32Speaker 1: And there's so much science to back this up.
- 5:35Speaker 2: So I really love this one. But don't get the
- 5:39Speaker 2: constraint put on the AI because get it to suggest
- 5:42Speaker 2: a few constraints for you based upon the problem you've got.
- 5:45Speaker 2: So it might be budgetary, it might be time, it
- 5:47Speaker 2: might be the types of people that you've got to
- 5:49Speaker 2: be able to work with you on the team. It
- 5:51Speaker 2: might be the resources you've got. It might be, it
- 5:55Speaker 2: could be you can't email and you've got to use
- 5:57Speaker 2: carrier pigeon. I don't know. Get it to suggest some
- 6:01Speaker 2: interesting constraints that might get you ideating. And then you
- 6:05Speaker 2: might go with the constraint, then you use AI, not
- 6:09Speaker 2: as to come up the ideas with you, but more
- 6:11Speaker 2: of an idea bouncer buddy. So what if I did this?
- 6:14Speaker 2: How do you reckon that would go? Get it to
- 6:16Speaker 2: go do some research for you. Go search the internet.
- 6:18Speaker 2: Go and find out someone's done something on that before,
- 6:21Speaker 2: but don't get it to do the thinking for you.
- 6:23Speaker 2: Get it to do the research and validation for you
- 6:26Speaker 2: while you're kind of leading the thinking here.
- 6:28Speaker 1: Okay. What else do you do?
- 6:31Speaker 2: Other ones.
- 6:31Speaker 1: Okay.
- 6:32Speaker 2: Another one is an odd analogy. So if you can
- 6:37Speaker 2: kind of see the kind of problem you've got, uh,
- 6:40Speaker 2: get AI to pose something which might be a little bit, uh,
- 6:44Speaker 2: wacky for you. So, um, instead of how would I, um,
- 6:49Speaker 2: how would I get this new strategy out? And I'm
- 6:51Speaker 2: going to really engage my team and, and make them
- 6:53Speaker 2: feel awesome about this new strategy. Um, maybe AI might
- 6:57Speaker 2: suggest something like, well, what would Ikea do if it
- 7:00Speaker 2: was going to give this strategy as instructions? And then
- 7:03Speaker 2: how do you communicate that as something you could explain
- 7:06Speaker 2: it to people? Um, How would you, instead of having
- 7:09Speaker 2: this as a sales problem you've got, it was a
- 7:11Speaker 2: video game. What would you actually look at to be
- 7:13Speaker 2: able to kind of gamify this or what would the
- 7:16Speaker 2: characters in the video game be? Now, this is not
- 7:19Speaker 2: for you to come up with. AI will come up
- 7:21Speaker 2: with these odd analogies and then it's just other ways
- 7:24Speaker 2: for you to start to think about your problem in
- 7:25Speaker 2: different ways.
- 7:26Speaker 1: I love this because at Inventia moment, we have run hundreds,
- 7:30Speaker 1: if not thousands of ideation workshops and pre-AI, that was
- 7:34Speaker 1: a really common technique that we used, where we got
- 7:37Speaker 1: people to put themselves in someone else's shoes who was
- 7:40Speaker 1: really different to them.
- 7:43Speaker 2: Absolutely. And this is then a completely different market or
- 7:46Speaker 2: a different place that you've worked in as well. But
- 7:49Speaker 2: it's you coming up with these other ideas or the
- 7:52Speaker 2: other things that you don't think of. By definition, AI
- 7:55Speaker 2: is good at posing those in. Okay.
- 7:57Speaker 1: What's the next one?
- 7:58Speaker 2: Number three is inversion. So, inversion is effectively what if
- 8:03Speaker 2: Instead of it succeeding, it was a spectacular failure. And
- 8:07Speaker 2: what would make it a failure? And so you might
- 8:09Speaker 2: have some ideas. Okay, if it's a really big failure,
- 8:12Speaker 2: then I would do X, Y, and Z. And you
- 8:14Speaker 2: might find by actually planning the failure, you might have
- 8:18Speaker 2: some ideas on how to make that a success. It's
- 8:21Speaker 2: a different way for your brain to think. And then,
- 8:23Speaker 2: of course, you can have a chat with your brainstorm
- 8:25Speaker 2: buddy to be able to go, well, if this was
- 8:28Speaker 2: a problem, how big of a problem would this be?
- 8:32Speaker 2: It's also really interesting with the constraints as well. Constraints
- 8:36Speaker 2: plus failure is fun.
- 8:37Speaker 1: Yes. Also known as a premortem as well. Again, something
- 8:41Speaker 1: that we've done quite a bit at Inventium when we've
- 8:43Speaker 1: been looking at new ideas. We will run a premortem
- 8:47Speaker 1: facilitated by a human, but the joy is Now you
- 8:50Speaker 1: can just do this on your own and get your
- 8:53Speaker 1: AI to facilitate you through it.
- 8:55Speaker 2: And then you can throw ideas in with that. How
- 8:58Speaker 2: big of a problem do you think that would be?
- 9:01Speaker 2: Do research on what the impact could be with this market,
- 9:04Speaker 2: this segment or whatever, and it can get some science
- 9:06Speaker 2: behind that for you. So rather than you just sitting
- 9:09Speaker 2: in a room bouncing ideas off with yourself, which is
- 9:11Speaker 2: not fun, AI then becomes that buddy to help you
- 9:15Speaker 2: with the innovation part. Number four, are you ready for
- 9:18Speaker 2: the last one?
- 9:18Speaker 1: I am.
- 9:19Speaker 2: Hard trade-offs. Sometimes we think we should and could do everything,
- 9:24Speaker 2: but instead, we've always got things that can be trade-off.
- 9:28Speaker 2: Budget's a really easy one. Instead of having $ 5, 000 budget,
- 9:31Speaker 2: you've only got a $ 2, 000 budget. What is the importance?
- 9:35Speaker 2: Or if you can have more budget, but you have
- 9:39Speaker 2: fewer people. So if you're AI can pose to you
- 9:44Speaker 2: different trade-offs, you might find what's really key at the
- 9:49Speaker 2: thing you're trying to achieve or the problem you're trying
- 9:51Speaker 2: to solve or the thing that your customers, clients, or
- 9:54Speaker 2: your teammates need in that thing you're trying to innovate
- 9:57Speaker 2: or to create or to brainstorm. and so the hard
- 10:01Speaker 2: trade-off often gets you to think in different ways of
- 10:03Speaker 2: like that's important that's not and so we can really
- 10:07Speaker 2: hone in on what's important and then of course with
- 10:09Speaker 2: all of these then you've got a conversation with an
- 10:12Speaker 2: ai to say is that good is that bad is
- 10:14Speaker 2: that going to work all those kind of things um Yeah,
- 10:17Speaker 2: so there's the four different categories I'd look at, and
- 10:19Speaker 2: there's a bonus one I'll give you at the end
- 10:21Speaker 2: if you're interested.
- 10:22Speaker 1: Oh, I always love a bonus.
- 10:24Speaker 2: What is the bonus one? The bonus is tell AI,
- 10:29Speaker 2: and we've got this in our brainstorm buddy.
- 10:32Speaker 1: Oh, you mean our brainstorm buddy skill that you use
- 10:37Speaker 1: with our clients?
- 10:37Speaker 2: That's the one, yes.
- 10:38Speaker 1: Yes, lovely.
- 10:39Speaker 2: So we cost money. In other words, like sometimes you
- 10:45Speaker 2: don't know and you think, what about this and this?
- 10:48Speaker 2: Or what do you think about X and Y? And
- 10:51Speaker 2: it's not unreasonable to ask a human being, look, I
- 10:54Speaker 2: can't pick between these two. But brainstorm buddy won't just go, yeah, fine,
- 10:59Speaker 2: I'll give you that. it costs everything you ask of it.
- 11:03Speaker 2: In other words, it'll say, all right, I can give
- 11:05Speaker 2: you five, but you need to tell me more about
- 11:07Speaker 2: your client, or you've got to tell me more about
- 11:08Speaker 2: what you're trying to achieve with this particular thing. Because
- 11:12Speaker 2: we really quite often get into this point where we're
- 11:15Speaker 2: having a conversation with it, and then we will get
- 11:17Speaker 2: sucked into the void of AI helping us to do
- 11:19Speaker 2: all the things. Don't get sucked in the void. And
- 11:22Speaker 2: the way to do that is get AI to cost
- 11:24Speaker 2: everything that it just delivers to you. So If you say, look,
- 11:27Speaker 2: I think there might be three options here. Can you
- 11:29Speaker 2: give me three options to get started with? It might
- 11:31Speaker 2: go fine, I'll give you the three, but you need
- 11:33Speaker 2: to tell me more about what good looks like for
- 11:35Speaker 2: this customer, client or problem. And so how do you.
- 11:39Speaker 1: prompt the AI to do that for you?
- 11:42Speaker 2: Yeah, it takes a little bit, but basically, you've got
- 11:46Speaker 2: to be clear. Your job is to help me, but
- 11:49Speaker 2: not to do the work for me. And so if
- 11:52Speaker 2: I do get stuck, absolutely help me get unstuck from this,
- 11:55Speaker 2: but don't do the thinking for me. And so instead,
- 11:58Speaker 2: try to get me to think along a line, and
- 12:01Speaker 2: then you can understand more about the line for you
- 12:03Speaker 2: to then ideate or give me some next steps on.
- 12:06Speaker 2: So it's more about I need to set at least
- 12:09Speaker 2: a step in the direction before it then opens up
- 12:11Speaker 2: the next couple of options in that direction. Awesome.
- 12:15Speaker 1: I love that because it's kind of like a guardrail
- 12:17Speaker 1: against human laziness that you've built into that skill.
- 12:20Speaker 2: Yeah. And you can put this within the work you
- 12:22Speaker 2: do because often with anything, AI will want to deliver
- 12:27Speaker 2: you the work and sometimes deliver stuff before you even
- 12:30Speaker 2: ask for it. And so here's a way for you
- 12:32Speaker 2: to say, no, no, no, no, you are not going
- 12:34Speaker 2: to do that. I need to be kind of leading
- 12:37Speaker 2: this I don't mind if you do some of the legwork,
- 12:39Speaker 2: but I kind of need to lead it. And forced
- 12:41Speaker 2: me to lead it, forced me to know what the
- 12:43Speaker 2: customer and client knows first. So it's a way to
- 12:45Speaker 2: do that.
- 12:46Speaker 1: Awesome. Neo, thank you for sharing these very practical tips.
- 12:50Speaker 1: And hopefully our listeners' creative juices are now flowing, or
- 12:54Speaker 1: at least they will be when they try out these
- 12:56Speaker 1: great strategies. If you enjoyed this episode, I reckon you
- 13:00Speaker 1: will really love an interview that I did with Exploding
- 13:05Speaker 1: Kittens CEO, where he talks about how he has a
- 13:09Speaker 1: walking with AI ritual to unlock great ideas. I absolutely
- 13:14Speaker 1: loved this use case for AI, and you can find
- 13:17Speaker 1: a link to that episode in.
- 13:19Speaker 2: The show notes.
- 13:20Speaker 1: How I AI was hosted by me, Amanda Imber and
- 13:23Speaker 1: Neo Applin. A big thank you to Martin Imber, who
- 13:26Speaker 1: does our sound editing, and Jem Rubio for production support.
- 13:30Speaker 1: And thank you to John Kilby, who composed the theme music.