1 00:00:00,640 --> 00:00:03,600 Speaker 1: If you've ever hit send on a document report or 2 00:00:03,640 --> 00:00:07,520 Speaker 1: slide deck and then immediately thought, hmmm, I hope that 3 00:00:07,680 --> 00:00:08,440 Speaker 1: lands okay. 4 00:00:08,960 --> 00:00:10,280 Speaker 2: This episode is. 5 00:00:10,280 --> 00:00:14,680 Speaker 1: For you because producing a high quantity of work isn't 6 00:00:14,680 --> 00:00:18,880 Speaker 1: the hard part anymore. It's producing high quality, which means 7 00:00:19,079 --> 00:00:22,920 Speaker 1: checking it, stress testing it, making sure it's clear, accurate, 8 00:00:23,040 --> 00:00:27,480 Speaker 1: and actually says what you meant before someone else points 9 00:00:27,520 --> 00:00:30,720 Speaker 1: out the gaps. Now, sure, you can enter a basic 10 00:00:30,800 --> 00:00:33,519 Speaker 1: prompt into AI to ask it to review your work. 11 00:00:33,720 --> 00:00:37,240 Speaker 1: But today Neo and I are going to level up 12 00:00:37,280 --> 00:00:39,880 Speaker 1: the way you can use AI to make your work 13 00:00:40,000 --> 00:00:44,159 Speaker 1: better and not to replace your judgment, but to sharpen it. 14 00:00:44,760 --> 00:00:47,000 Speaker 2: You're gonna learn how to get AI to. 15 00:00:46,960 --> 00:00:50,440 Speaker 1: Critique your work properly, not just polish it, ask for 16 00:00:50,440 --> 00:00:53,760 Speaker 1: feedback on clarity, logic, bias, and gaps, and we're gonna 17 00:00:53,800 --> 00:00:57,800 Speaker 1: give you one mega prompt to review emails, documents, slides, 18 00:00:57,880 --> 00:01:01,160 Speaker 1: or anything you like. Really, by the end of this episode, 19 00:01:01,200 --> 00:01:04,039 Speaker 1: you won't be guessing whether your work is good enough. 20 00:01:04,360 --> 00:01:07,240 Speaker 2: You're going to have a clear, repeatable. 21 00:01:06,640 --> 00:01:10,080 Speaker 1: Way to pressure test it before it leaves your desk, 22 00:01:10,680 --> 00:01:13,720 Speaker 1: and the confidence to decide what to change and what 23 00:01:13,880 --> 00:01:22,800 Speaker 1: to stand by. Welcome to how I AI with me 24 00:01:23,000 --> 00:01:27,280 Speaker 1: Doctor Amantha Imba and Neo Applin, head of Inventium AI. 25 00:01:27,880 --> 00:01:30,959 Speaker 1: Each episode we share one practical way to use AI 26 00:01:31,080 --> 00:01:35,240 Speaker 1: better at work and in life. No fluff, no dech jargon, 27 00:01:35,600 --> 00:01:40,080 Speaker 1: just things you can use straight away. So on previous episodes, Neo, 28 00:01:40,160 --> 00:01:42,120 Speaker 1: you've referred to how we can get AI. 29 00:01:42,080 --> 00:01:45,240 Speaker 2: To review our own work. 30 00:01:45,520 --> 00:01:47,600 Speaker 1: But I want to get into the nitty gritty of 31 00:01:48,040 --> 00:01:50,880 Speaker 1: how do we do that beyond just a simple prompt like, 32 00:01:51,200 --> 00:01:54,680 Speaker 1: you know, put your critical thinking hat on AI and 33 00:01:55,040 --> 00:01:56,920 Speaker 1: tell me what I could improve, Like how do we 34 00:01:56,960 --> 00:01:59,800 Speaker 1: get a little bit more sophisticated. 35 00:01:59,640 --> 00:02:01,440 Speaker 3: You need to tell it what you wanted to do 36 00:02:01,520 --> 00:02:04,360 Speaker 3: and also tell it exactly what you wanted to review 37 00:02:04,560 --> 00:02:08,720 Speaker 3: and check. So I've got I've called it Charles, So 38 00:02:08,720 --> 00:02:12,240 Speaker 3: I've created myself an agent to a GPT and Charles 39 00:02:12,280 --> 00:02:14,360 Speaker 3: reviews my stuff. Now I've put in a little bit 40 00:02:14,360 --> 00:02:17,920 Speaker 3: more specifics around it, but at its essence, this more 41 00:02:18,000 --> 00:02:20,960 Speaker 3: than two sentences is what you need to write. So 42 00:02:21,080 --> 00:02:23,600 Speaker 3: I've got review the last response. Of course, you could 43 00:02:23,639 --> 00:02:26,600 Speaker 3: say review the document, review this thing that I've uploaded, 44 00:02:26,639 --> 00:02:31,080 Speaker 3: whatever it is, perform a rigorous critique to find anything incorrect. 45 00:02:31,800 --> 00:02:39,600 Speaker 3: Weekly supported, unclear, incomplete, biased, impractical, verbose, wordy, or improbable. 46 00:02:39,680 --> 00:02:42,520 Speaker 3: So as you can tell, I'm going through all the bases, 47 00:02:42,560 --> 00:02:44,639 Speaker 3: I want you to review it and just go through 48 00:02:44,800 --> 00:02:47,679 Speaker 3: everything that could be wrong with it, then propose better 49 00:02:47,720 --> 00:02:52,680 Speaker 3: alternatives and provide a corrected slash revised version. So you 50 00:02:52,720 --> 00:02:56,880 Speaker 3: could just get away with that one line. So's that's 51 00:02:57,000 --> 00:02:58,640 Speaker 3: enough to be able to get it to review it. 52 00:02:58,639 --> 00:03:01,160 Speaker 3: Because what you're being here is about what you want 53 00:03:01,160 --> 00:03:03,280 Speaker 3: to review it. So you could say I want to 54 00:03:03,880 --> 00:03:07,000 Speaker 3: find instead off anything incorrect, weekly, supported, unclear, and complete. 55 00:03:07,200 --> 00:03:11,720 Speaker 3: You could say anything that is against our policies, anything 56 00:03:11,840 --> 00:03:16,359 Speaker 3: against you, how our operating process are, or anything against 57 00:03:16,480 --> 00:03:19,040 Speaker 3: how I think a CFO would want to review this paper, 58 00:03:19,240 --> 00:03:21,560 Speaker 3: those kind of things. So be specific about how you 59 00:03:21,639 --> 00:03:24,359 Speaker 3: want your agent, your chat thread to be able to 60 00:03:24,400 --> 00:03:27,440 Speaker 3: review your work. So that's an easy thing. So review 61 00:03:27,440 --> 00:03:29,440 Speaker 3: the last response and then put in all those things. 62 00:03:29,680 --> 00:03:32,679 Speaker 3: And we'll put that prompt in the show notes as well. 63 00:03:33,919 --> 00:03:35,880 Speaker 1: And in fact, we will actually do better than that, because 64 00:03:35,920 --> 00:03:39,160 Speaker 1: I think we're going to have the whole system instructions 65 00:03:39,200 --> 00:03:40,040 Speaker 1: for Charles. 66 00:03:40,160 --> 00:03:40,800 Speaker 2: Is that right? 67 00:03:41,120 --> 00:03:41,240 Speaker 1: Oh? 68 00:03:41,280 --> 00:03:44,120 Speaker 3: Okay? And if you're infratrate, okay, we're going to give 69 00:03:44,120 --> 00:03:47,120 Speaker 3: you Charles. So this is what Charles does. So I 70 00:03:47,200 --> 00:03:49,800 Speaker 3: call it Charles because I use it so often, I 71 00:03:49,800 --> 00:03:51,840 Speaker 3: may as well give it a name, So we'll give 72 00:03:51,880 --> 00:03:55,600 Speaker 3: you Charles. So Charles has that task, but Charles also 73 00:03:55,880 --> 00:03:58,480 Speaker 3: has a couple of steps after that that it follows. 74 00:03:58,720 --> 00:04:00,640 Speaker 3: So the first is it goes what do I think 75 00:04:00,720 --> 00:04:03,400 Speaker 3: Neo's really trying to do? And what do I think 76 00:04:03,400 --> 00:04:05,440 Speaker 3: He's really wanting me to check? And that's a really 77 00:04:05,480 --> 00:04:07,880 Speaker 3: important thing for you to see. Is AI actually checking 78 00:04:07,880 --> 00:04:09,600 Speaker 3: the things that I want it to check? And it 79 00:04:09,640 --> 00:04:13,680 Speaker 3: understands me. It goes to equality gates. So in other words, 80 00:04:13,720 --> 00:04:17,680 Speaker 3: what it says is have you passed these main categories 81 00:04:17,720 --> 00:04:24,520 Speaker 3: and the categories I've given it is factual accuracy, logical soundness, completeness, relevance, clarity, structure, safety, 82 00:04:24,760 --> 00:04:27,360 Speaker 3: and practicality. So it goes through those and effectively gives 83 00:04:27,360 --> 00:04:29,200 Speaker 3: me a tick or a cross, so I can see 84 00:04:29,240 --> 00:04:31,200 Speaker 3: if he goes biased, Okay, I've got to go in 85 00:04:31,240 --> 00:04:33,840 Speaker 3: there and change those kind of things, or practical sound, 86 00:04:34,160 --> 00:04:37,320 Speaker 3: factual accuracy, those kind of things. Then it goes through 87 00:04:37,440 --> 00:04:41,040 Speaker 3: and tells me some of the main points that I 88 00:04:41,080 --> 00:04:43,839 Speaker 3: could have added to or I could have changed. So 89 00:04:43,839 --> 00:04:47,279 Speaker 3: it's effectively a diagnostic audit, so it says how important 90 00:04:47,320 --> 00:04:51,599 Speaker 3: this is, how confident it is about how important that is, 91 00:04:51,720 --> 00:04:54,480 Speaker 3: or the problems there. So if you've got a high 92 00:04:54,520 --> 00:04:57,680 Speaker 3: severity like this is really important, but a low confidence 93 00:04:57,760 --> 00:05:00,240 Speaker 3: is like, well, I'm not sure i've been to to 94 00:05:00,360 --> 00:05:04,040 Speaker 3: your work neo. But if I've interpreted right, high importance 95 00:05:04,200 --> 00:05:08,000 Speaker 3: those kind of things, it'll tell you what problem it's got, 96 00:05:08,000 --> 00:05:11,080 Speaker 3: whether it's like a user experience, a structure or clarity whatever, 97 00:05:11,360 --> 00:05:13,360 Speaker 3: and it tells you what the problem is where and 98 00:05:13,400 --> 00:05:16,280 Speaker 3: what you can do to fix it. Also gives you 99 00:05:16,320 --> 00:05:19,320 Speaker 3: a bunch of alternative approaches. So for these things that 100 00:05:19,400 --> 00:05:21,800 Speaker 3: you've got in this work, the document, the email you 101 00:05:21,800 --> 00:05:24,280 Speaker 3: wanted to review, whatever it is, it will then say, 102 00:05:24,360 --> 00:05:28,320 Speaker 3: here's some alternative approaches on how you can potentially fix this. 103 00:05:29,600 --> 00:05:32,760 Speaker 1: So can I understand just day to day because I 104 00:05:33,279 --> 00:05:36,240 Speaker 1: feel like I hear you use Charles a lot, So 105 00:05:36,560 --> 00:05:39,480 Speaker 1: what are the kinds of tasks that you're bringing in 106 00:05:40,000 --> 00:05:43,560 Speaker 1: Charles for? And just to confirm, Charles is a GPT 107 00:05:43,720 --> 00:05:46,880 Speaker 1: and chat GPT for you, Is that right? Yes, that's great, yes, 108 00:05:47,000 --> 00:05:50,000 Speaker 1: But equally it could be a project in Chlaude. It 109 00:05:50,040 --> 00:05:52,520 Speaker 1: could be an agent in co Pilot, it could be 110 00:05:52,520 --> 00:05:54,040 Speaker 1: a gem in Gemini. 111 00:05:54,040 --> 00:05:56,000 Speaker 3: Yes, absolutely, or it could just be a prompt in 112 00:05:56,040 --> 00:05:57,920 Speaker 3: your prompt library you dump in whenever you need. 113 00:05:58,320 --> 00:06:00,719 Speaker 1: Okay, so tell me what are the kinds of tasks 114 00:06:00,720 --> 00:06:01,800 Speaker 1: that you're using Charles for? 115 00:06:03,040 --> 00:06:07,080 Speaker 3: Anything with analysis, I'm getting it to check my analysis. 116 00:06:07,360 --> 00:06:10,800 Speaker 3: So if I'm doing a review or some kind of 117 00:06:10,800 --> 00:06:13,160 Speaker 3: a document or whatever, then I'm getting Charles to review 118 00:06:13,200 --> 00:06:15,960 Speaker 3: that and give me their or its opinion on what 119 00:06:16,080 --> 00:06:19,720 Speaker 3: I think what it thinks can be improved in this work. Now, 120 00:06:19,760 --> 00:06:22,599 Speaker 3: by the way, sometimes Charles comes up with suggestions like, yeah, 121 00:06:22,680 --> 00:06:24,920 Speaker 3: I get that, but no, you don't know the background 122 00:06:25,000 --> 00:06:26,600 Speaker 3: like I do, and so I don't have to change it. 123 00:06:27,640 --> 00:06:32,160 Speaker 3: But those kind of things I was running through Charles. 124 00:06:32,240 --> 00:06:35,440 Speaker 3: I'd done a social post. We've got online communities for 125 00:06:35,480 --> 00:06:38,640 Speaker 3: AI people learning AI, and I had already done a 126 00:06:38,640 --> 00:06:40,600 Speaker 3: social post, and I thought, oh, I forgot to put 127 00:06:40,640 --> 00:06:44,719 Speaker 3: that through Charles, and so I posting it. I put 128 00:06:44,760 --> 00:06:47,240 Speaker 3: the social post into Charles, and it gave me a 129 00:06:47,240 --> 00:06:49,080 Speaker 3: couple of suggestions on how I could quickly go in 130 00:06:49,120 --> 00:06:51,119 Speaker 3: there and do some tweaks to make my social post 131 00:06:51,240 --> 00:06:54,720 Speaker 3: clear out so it can be anything from social posts 132 00:06:54,800 --> 00:06:59,000 Speaker 3: to emails to documentsy working on analysis as you're doing 133 00:06:59,360 --> 00:07:02,720 Speaker 3: slide you can give I've given slide decks to in 134 00:07:02,760 --> 00:07:05,880 Speaker 3: the past, a whole bunch of different things, so pretty 135 00:07:05,920 --> 00:07:08,240 Speaker 3: much anything you're doing. If you would love to get 136 00:07:08,279 --> 00:07:11,200 Speaker 3: a second opinion, then get Charles to give you that 137 00:07:11,240 --> 00:07:15,280 Speaker 3: second opinion. So I'm not quite there. I still haven't 138 00:07:15,280 --> 00:07:18,560 Speaker 3: built it into everything I do, but anything important I 139 00:07:18,560 --> 00:07:20,400 Speaker 3: get Charles to have a look at. But I think 140 00:07:20,400 --> 00:07:24,400 Speaker 3: it's almost like a spell check, Like we get word 141 00:07:24,480 --> 00:07:26,600 Speaker 3: to do a spell check and a grammar check for 142 00:07:26,640 --> 00:07:29,120 Speaker 3: anything we then do and then upload or give to 143 00:07:30,560 --> 00:07:33,320 Speaker 3: a colleague. It's a good idea to get AI to 144 00:07:33,400 --> 00:07:36,600 Speaker 3: do a review to see whether you can improve those 145 00:07:36,640 --> 00:07:39,000 Speaker 3: things before you give it to those people. Why that 146 00:07:39,040 --> 00:07:41,080 Speaker 3: review doesn't take much because of course AI is just 147 00:07:41,080 --> 00:07:43,400 Speaker 3: doing the review. You can agree or disagree with those things, 148 00:07:43,480 --> 00:07:45,440 Speaker 3: and quite often it's like, oh, that's a really good pickup, 149 00:07:45,440 --> 00:07:47,680 Speaker 3: and I'm just going to do small tweaks. So it's 150 00:07:47,760 --> 00:07:50,640 Speaker 3: like the third thing we're doing spell check, grammar check, 151 00:07:50,840 --> 00:07:52,560 Speaker 3: get it to do a your Charle's check, and of 152 00:07:52,560 --> 00:07:53,800 Speaker 3: course we're giving you the we're going to give you 153 00:07:53,840 --> 00:07:55,800 Speaker 3: the COVID Charles, so you can run that yourself. 154 00:07:56,520 --> 00:07:59,280 Speaker 1: How well does it work if AI has been primarily 155 00:07:59,360 --> 00:08:04,200 Speaker 1: responsible for creating the output, Like, let's just say, you know, 156 00:08:04,360 --> 00:08:07,120 Speaker 1: as sadly many people are, they're just kind of taking 157 00:08:07,120 --> 00:08:08,000 Speaker 1: shortcuts with AI. 158 00:08:08,040 --> 00:08:10,400 Speaker 2: They're being a bit lazy. They're producing some AI slop. 159 00:08:11,280 --> 00:08:13,600 Speaker 1: You know, maybe they've put in three bullet points and 160 00:08:13,600 --> 00:08:15,800 Speaker 1: they're like, write me a three page report that I 161 00:08:15,840 --> 00:08:18,680 Speaker 1: can give to my bus on this particular topic or 162 00:08:18,720 --> 00:08:22,040 Speaker 1: project update. Is it good at critiquing its own work 163 00:08:22,160 --> 00:08:26,480 Speaker 1: when the prompter's been lackluster and you know, maybe the 164 00:08:26,480 --> 00:08:28,320 Speaker 1: AI's output is actually not that good. 165 00:08:28,960 --> 00:08:31,880 Speaker 3: Yeah, I was a bet to joke and say it'll 166 00:08:31,960 --> 00:08:34,640 Speaker 3: recognize its own handywork and say that's awesome. No, it 167 00:08:34,679 --> 00:08:39,040 Speaker 3: doesn't do that. It'll even improve AI's output as well. 168 00:08:39,720 --> 00:08:42,160 Speaker 3: So it's a good thing to get this as a 169 00:08:42,160 --> 00:08:44,760 Speaker 3: second opinion AI. So if you've been working with A 170 00:08:44,920 --> 00:08:46,960 Speaker 3: on a document or whatever, even if you have just 171 00:08:47,000 --> 00:08:49,160 Speaker 3: been like phoning it in and saying five top points, 172 00:08:49,240 --> 00:08:51,520 Speaker 3: give me a document, it will still review that and 173 00:08:51,760 --> 00:08:55,440 Speaker 3: tell you how that could be improved. So don't take 174 00:08:55,520 --> 00:09:00,160 Speaker 3: AI's output first output as gospel ever, like I prefer, 175 00:09:00,280 --> 00:09:03,600 Speaker 3: using AI's output is to augment me, not to do 176 00:09:03,640 --> 00:09:06,280 Speaker 3: the work for me. But yes, you can get it 177 00:09:06,320 --> 00:09:08,200 Speaker 3: to review AI's output and we'll actually give you a 178 00:09:08,200 --> 00:09:10,640 Speaker 3: better response and'll actually tell you where the flaws are 179 00:09:10,880 --> 00:09:14,319 Speaker 3: where it could be improved. So yes, absolutely use it for. 180 00:09:14,320 --> 00:09:20,440 Speaker 1: That amazing So for anyone listening, we are very or 181 00:09:20,480 --> 00:09:25,040 Speaker 1: i should say NEO is very generously sharing Charles with you. All. 182 00:09:25,280 --> 00:09:27,439 Speaker 1: Just pop into the show notes and there will be 183 00:09:27,800 --> 00:09:31,440 Speaker 1: a link to get the full text in markdown to 184 00:09:31,600 --> 00:09:36,400 Speaker 1: paste into your GPT agent project of choice, or. 185 00:09:36,400 --> 00:09:38,800 Speaker 2: You can just use it as a mega prompt. 186 00:09:38,840 --> 00:09:43,079 Speaker 1: So enjoy Charles, maybe give it its own name, and 187 00:09:43,120 --> 00:09:47,440 Speaker 1: we will see you next Monday. If you found this useful, 188 00:09:47,480 --> 00:09:50,480 Speaker 1: please help us spread the AI love and share it 189 00:09:50,520 --> 00:09:53,240 Speaker 1: with someone who you think would benefit from knowing what 190 00:09:53,400 --> 00:09:56,520 Speaker 1: you now know. And if you're ready to really start 191 00:09:56,600 --> 00:10:01,640 Speaker 1: mastering AI, check out inventium dot ai. We help individuals, teams, 192 00:10:01,679 --> 00:10:06,959 Speaker 1: and organizations turn Jenai into a real work superpower, saving 193 00:10:07,200 --> 00:10:10,240 Speaker 1: ten plus hours a week and staying future already without 194 00:10:10,240 --> 00:10:11,880 Speaker 1: the jargon or overwhelmed. 195 00:10:12,160 --> 00:10:14,120 Speaker 2: Thanks so much for listening, and we'll see you next 196 00:10:14,160 --> 00:10:14,760 Speaker 2: time on. 197 00:10:14,679 --> 00:10:19,280 Speaker 1: How IAI how i AI was hosted by me Amantha 198 00:10:19,320 --> 00:10:20,760 Speaker 1: Imber and Neo Applan. 199 00:10:21,040 --> 00:10:23,000 Speaker 2: A big thank you to Martin Imber who does our 200 00:10:23,080 --> 00:10:24,120 Speaker 2: sound editing, and 201 00:10:24,240 --> 00:10:27,640 Speaker 1: Jem Rubio for production support, and thank you to John 202 00:10:27,720 --> 00:10:29,680 Speaker 1: Kilby who composed the theme music.