Skip to content
PodSubs
ShowsPeopleGuestsJobsNewsEvents
More
BlogPodCred
Menu
ShowsPeopleGuestsJobsNewsEventsBlogPodCredNetworksSign in
Join nowSign in
  • Shows
  • People
  • Jobs
  • News
  • More
    • Guests
    • Events
    • Blog
    • PodCred
    • Networks
    Join nowSign in

How I Work

How I Work artwork

Your prompts are stuck in 2025. Here are 5 fixes

Oct 18, 202612 minAudio
ShareXLinkedInFacebookWhatsAppEmail

Some prompts have been with you for a year or more. They open with a grand persona for the AI, add a few rules in capital letters, and finish with a request to think it all through. They worked well once, and lately the results feel slightly off. The AI models have changed, and a few of the habits we picked up in 2024 and 2025 now work against us. In this How I AI episode, Neo and I walk through five ways good prompting has changed as of October 2026. You will come away with a simple way to tighten up the prompts you rely on most. What you’ll learn: What a persona really changes in an AI response When a goal and a finished example beat step-by-step instructions Why asking AI to think through its answer can backfire How to add emphasis without boxing the AI in What a good checklist looks like before AI hands work back Practical AI tools for productivity and focus Real-world AI workflows used by high performers How to use AI at work without burning out Smart shortcuts for managing time and mental load Connect with Neo Aplin on LinkedIn (https://www.linkedin.com/in/neoaplin/) and via inventium.ai (https://inventium.ai), where he leads Inventium's AI training and upskilling work with organisations and teams. Find out more about Inventium AI’s training programs at inventium.ai How I AI with Dr Amantha Imber and Neo Aplin explores how high performers are using AI at work to boost productivity, make better decisions and reduce overwhelm. Want more practical ways to put AI to work? You can now follow How I AI separately on Spotify or Apple Podcasts so you never miss an episode. My latest book The Energy Game is out now. You can order a copy here: https://amzn.to/48ID29M Connect with me on the socials: Linkedin (https://www.linkedin.com/in/amanthaimber) Instagram (https://www.instagram.com/amanthai) If you are looking for more tips to improve the way you work and live, I write a weekly newsletter where I share practical and simple to apply tips to improve your life. You can sign up for that at https://amantha.substack.com/ Visit https://www.amantha.com/podcast for full show notes from all episodes. Get in touch at amantha@inventium.com.au Credits: Host: Amantha Imber Sound Engineer: Martin Imber

Audio and video play straight from the show's own hosting. PodSubs stores and serves no media.

Transcript

From the show. Tap a time to play from there.

Speaker 1Tell me if you can relate to this. You've got a prompt that you've been reusing for a year. It opens with, you are a world-class expert. It asks the AI to think through its answer, and it has a couple of rules in capital letters, so it knows you mean business. But lately, though, maybe those results have felt a little bit off. And the reason why is because the models have moved on and a lot of the habits that we picked up in 2024 and 2025 now

Speaker 1actually get in the way. Today's AI takes your words very literally and it already does the thinking that you used to have to ask it to do. In today's episode, Neo and I go through five ways that prompting has changed. You're going to learn why. what a persona really changes in a response, why a goal plus a finished example often beats spelling out every step. And we're also going to cover the importance of a clear checklist to get

Speaker 1AI to actually check its work before handing anything over. By the end of today's show, you are going to have prompts that are much shorter and clearer and also much more likely to get it right the first time. Welcome to How I AI with me, Dr. Amantha Imber and Neo Applin, head of Inventium AI. Each episode, we share one practical way to use AI better at work

Speaker 1and in life. No fluff, no tech jargon, just things you can use straight away. In our training programs at Inventium AI, we, of course, have always talked about prompting, but the way that we talk about prompting has changed a a lot since 2024 and has also changed since 2025. So, Neo, I would love to know what are the biggest changes that have happened in what good prompting looks like in

Speaker 1late 2026 as we are recording this in October versus, say, a year ago? Where should we start?

Speaker 2My first one that I'd look at is probably persona. It used to be that you would imbue this role into the AI. You are a world-class expert in X, Y, and Z. And then it would be able to give you a better result than just, hey, tell me the thing. And that certainly was the case back in 2024, a little bit of 2025, and it's certainly gone out of place. Where role or persona or perspective, which is what we probably use is really valuable, is It then comes to

Speaker 2perspective or the language that comes out of it, but it doesn't give you a better result. It just gives you a different result. It's kind of like if you said, explain it like a high school teacher versus explain it like a university professor. It's not a better result. It's just a different one that might help you in a different way. So perspective is good. But, yes, certainly don't say you're smarter than Einstein with a 2 million IQ. It's going to make up stuff hilariously and certainly won't actually improve anything. So, yes, give it a perspective rather

Speaker 2than imbue a role. Okay.

Speaker 1Tell me about micromanagement because previously, certainly in 2024 and 2025, we– we would need to micromanage AI's response and how it went about its job. But we don't need to do that anymore, do we?

Speaker 2Well, we do and we don't. I think I disagree a little bit with some of the advice that the companies have. And I'll tell you why. The tools have changed a bit so that they don't need to be told necessarily how to do things every single time. And they're relatively smart and they can go about and find their things. And so what you can do is just say, here's my goal and here's what good looks like. You sort out how you get there. And they've found that that is more effective because if you say, look, you must do it my way, that may be less effective

Speaker 2for the AI to do it and then effectively then hit against walls, rework, all of those kind of things. And I do love this. The goal and what good looks like, not just telling it, by the way, I love showing it what good looks like. So if you've got an example of that thing you've done before and say, it's got to be like this, then understand. So it's rather than pictures of a thousand words, a completed document is a million words, if you'd like. So you can get a much better outcome there. But sometimes... we do need to do things in a particular way for our jobs.

Speaker 2You might have a meeting report template and it needs to be done in a very particular way. And maybe some of the sources need to come from very specific areas or calculated in a particular way. So simply just going, hey, write me a report, here's what it needs to look like, doesn't necessarily give you the best results. So I think it's a continuum. Sometimes it's very much like, just give it the goal and what good looks like, Other times you do need to say, do this, then this, then that. But the ultimate thing is this, if you don't need

Speaker 2to be a micromanager, don't be a micromanager with the AI because you don't need to be anymore. Okay.

Speaker 1Now, something that we used to recommend in our programs when we were asking AI to do something that was potentially quite a complex task is we would ask it to Think through your answer. Can you talk about that and why we don't actually need to do that anymore?

Speaker 2The early models don't think. Thinking is a really odd word because you and I think in different ways than the AI does think. Let me tell you about this. The early models would predict words, and they still do now, and it would give you the result and say, here's what I think from my brain, or effectively just dump out the result there. but what they found was by saying, think through your answer, it would actually work out its thinking process in its brain. It would actually, a weird

Speaker 2vomit on your screen, so you could actually see how it was thinking. And that was before they created thinking models. And so thinking models still do that kind of thinking, but they just don't show you that boring text of, I'm trying to do this, and I'm not quite sure, is that right? And then this is the avenue I'm going through. That kind of thinking out loud still happens, you just don't see it on your screen. And so if you're saying think through your answer with something which is already doing a thinking behind the scenes, then you're actually getting to perform thinking, even though it's already done the job. And sometimes you get a worse response there.

Speaker 2So pretty much if you're on one of the tiny models, maybe like a quick response you might have on it or like a haiku, maybe you'll get a better response with think through your answer, but you're much better off just using a standard model that does the thinking these days. So Don't get to perform the thinking if it's already done the thinking.

Speaker 1Now, something that used to be helpful in 2024 and perhaps 2025 as well is yelling at your AI in all caps to make sure that it has heard you. Tell me, do we still need to yell at our AIs in late 2026?

Speaker 2If you want to, go right ahead. But what they've done is, I think a couple of people were doing things in different ways. So some people were just basically yelling the whole time, thinking that being Donald Trump's on all caps for your internal conversation would give you a better result. That certainly was not the case. The emphasis that yelling used to give you, always or never kind of thing, was really good in that AI could tell, wow, that's really important. Amantha doesn't usually... You'll use all caps. Maybe that's pretty important. So I'll make sure that I

Speaker 2do that. So showing emphasis absolutely is still required and still very good. What you might find these days is if you use the words never or always or must, it will take it as a absolute rule. And that is a problem because occasionally it might go, well, it's a better way to do it or a different way to do it. And so I can kind of get to the same goal without being told a particular process way of doing things. So I would say only use

Speaker 2absolutes when you really need to use those absolutes, always or never, and things like that. They've also said that even though that does happen, occasionally what they've also said is occasionally you actually get a better response, not by being absolute, but saying, don't do this and here's why. It does a much better job if you say why it's necessary important or why it shouldn't be that output. So you might say, needs to be 150 characters or less because this is only a title that people are

Speaker 2going to skim through and I'm going to have 30,000 titles on the screen that people need to scroll through to find the title. And it goes, oh, I understand it's quick readability. You'll find you get much better adherence there than an absolute rule. Okay.

Speaker 1We've got one final tip to give people about how to do prompting better at time of recording in October, 2026. And that relates to asking it to check things. Can you explain a little bit about this?

Speaker 2It's part of that. It's the dovetail of giving it a goal and what good looks like is saying, here's a checklist that I'd like you to check off. So before you give me that work, then please make sure you've done all these things because AI now can work longer than it used to before. Certainly if you've got say, co-work or work, and now they're obviously putting these two kind of features together, but it can work longer on things. And what it can do is think, is this ready for me to deliver to my friendly human? And you can help it to do that by saying, look, here's

Speaker 2a checklist. So does it have this? Does it have that? Is this section there? Is this calculation done? And just say, look, this is the checklist I'd like you to have before you deliver it to me. and it can actually review and check on that checklist. You can even say, look, tell me the checklist. And so if it hasn't hit the checklist, sometimes it'll work harder until it actually gets that checklist absolutely nailed. So give it that checklist, though. Make sure it's clear rather than opinion-based. So is it good? That's not a good checklist item. Is there a summary?

Speaker 2Is there an executive summary that's less than three paragraphs? That is something that it can actually deal with. And so you'll find that a checklist can give you a better result without the AI going backwards and forwards and you arguing with it so much. It can do its own check effectively.

Speaker 1Neo, thank you so much for sharing five ways that we can be changing how we prompt to get better responses with the models that we are now working with in late 2026. Thank you so much for listening to today's episode of How I AI. If this show got you thinking about how you could use AI better, that is exactly what we help people do at Inventium AI. We run AI training programs for organizations and also for individuals.

Speaker 1So if you want to upskill your entire organization or maybe just yourself, you can find out more at inventium.ai. Hey, and once more favor before you go, if you've got a moment, like just five seconds, I'd love it if you could leave us a review and hit follow wherever you're listening. It really helps the show reach more people and it helps us create even better episodes for you. Thank you so much for listening and we'll see you next time on How I AI. How I AI was hosted by me, Amantha Imber and Neo Applin. A big

Speaker 1thank you to Martin Imber who does our sound editing and Jem Rubio for production support. And thank you to John Kilby who composed the theme music.

Open the transcript file

Hosts and Guests

Nobody is credited on this episode yet.

Credit the people on this episode

Be the first to rate this episode

More episodes

All episodes
  1. The 2-week AI experiment that helps you remember every conversation that matters, with Chris Hutchins (Part 1)

    Nov 18, 202626 minAudio

    Think about how many useful conversations you have in a week. Now, how much of that advice do you actually remember? Chris Hutchins, host of All the Hacks podcast, started wearing an AI device that transcribes his...

    Credit the people on this episode

  2. AI skills are unreliable (until you fix these 3 things)

    Nov 15, 202614 minAudio

    **Join the AI Mastery program here: https://www.inventium.ai/ai-mastery ** You built an AI skill, tried it once, and it was brilliant. Then you ran it again and got something completely different, or handed it to a...

    Credit the people on this episode

  3. Prioritisation and saying no: the best advice from the How I Work archive

    Nov 11, 202625 minAudio

    Saying no to a bad idea takes no effort. Saying no to a good one is where prioritising gets hard. Too much to do, everything feeling urgent, and no clear idea of what to let go of: this is the problem I hear about more...

    Credit the people on this episode

  4. What is an AI skill (and why you should care)

    Nov 8, 202611 minAudio

    **Join our AI Mastery program: https://www.inventium.ai/ai-mastery ** You’ve typed the same instructions into AI for the third week running, tweaked the wording slightly, and got a slightly different result again. It’s...

    Credit the people on this episode

  5. How to make your frequent flyer points go further in 2027

    Nov 4, 202636 minAudio

    You've spent years collecting frequent flyer points. Then you finally go to book a flight, and somehow the seats you want are never available. Or the points required seem ridiculous. Or you give up and pay cash anyway....

    Credit the people on this episode