Episode Transcript
[00:00:05] Welcome to Prompting Curiosity, a podcast for the AI curious. No coding background required. I'm your host, Dr. Shantae Cofield, also known as the Maestro, and I created this show to explore what these AI tools actually are. Really, though, are the files in the computer, how to use them, and what they might mean for how we think, work, create, and move through life. Whether you're skeptical, intrigued, or already experimenting, you're in the right place. All that I ask is that you stay curious. All right, let's get into it.
[00:00:36] Hello, hello, hello, my curious people, and welcome to episode 68. 68 of, uh, prompting Curiosity. I am your grateful host, the Maestro, and today we are talking about a term that I made up and that I'm very proud of. Heft, gnosis.
[00:00:51] So in writing this episode, I was looking for a word to describe how, you know, mesmerized. And they were, like, hypnotized. So some folks are, by the sheer volume of content that AI will produce for an answer.
[00:01:04] As I'm thinking about this word, I didn't want to include the term slop. This is very, very intentional. I hate that word, right? I hate the word slop. One, because I think it's become a very lazy, catchall term that's just used by folks who hate I. They're like, who. Who hate I, who hate AI they're just like, it's slop, it's slop, it's slop. And I'm like, actually, it's not.
[00:01:24] And two, slop speaks more to content that isn't actually any good. And sometimes when the volume, you know, sometimes as it relates to the volume of content, uh, you know, the output content, it can be good, but it's just too much. Right? What I wanted to highlight and emphasize with this word again, is not so much the actual content and what's in it, how good or bad it is, but rather what it does to the consumer and how, you know, these 100 page documents that it's in, 100 page reports that it publishes, irrespective of quality, literally hypnotize the person who's using AI and keeps them coming back for more. Right?
[00:02:04] And thus the word hypnosis was born. Right? Heft and hypnosis, and we put them together.
[00:02:12] So, per me, the definition of hypnosis is AI attraction and excitement that is caused by equating output, volume and value.
[00:02:22] My guess is that unless you suffer from this very real AI affliction, you've actually experienced the what. What's called verbosity bias of your LLM of choice and found it to be annoying. Right? Uh, you are absolutely not alone if you've ever felt like ChatGPT is too chatty. That's one of the reasons I left. I was like, this thing can be talking too much.
[00:02:43] Uh, one of the things that inspired me to write this episode, it was largely that inspired this episode was that last week, Anthropic announced that, you know, is having a limited rollout of a new agent called Claude Duet. And actually, I guess it was two episodes ago, I believe, when I was talking about, uh, open AI dot open AI dots. And at the end, I was like, you know, we'll see if Anthropic roles went out. We know that they're going to roll out their own agent like this. And they did. And it's called Claude Duet. And per the little release, it says a new app where Claude works as a proactive agent. If you want to read the release statement, it is in the the show notes. You can just give it a little click. You're welcome. Um, it looks to me from the outside to be a competitor to Metamuse and OpenAI Dots. It was just a matter of time. Um, but I actually have not dove in yet because, like I said, it's a limited rollout and so I don't have. I don't have access to it. Um, but more than that, and directly tying into this episode, I don't see the utility. Just like I haven't used Cloud Cowork yet, I just haven't had a specific need for it. And, um, I didn't use cloud code for quite some time now. I fucking love cloud code, but I didn't use it for a long time because I just didn't have the need for it. All right? And I think that. I think that, that, you know, what I'm talking about here directly ties into the main point of this episode, right? I don't see the utility. And just because you can doesn't mean you need it. Right?
[00:04:03] At one point in the release video, the fictional user tells Claude Duet to, you know, the person writes, add a line at the end that says how excited I am to bring her on.
[00:04:14] And the line that gets added by Claude Duet reads, when we're so excited for you to join the team, and it's like, it would have been faster for you to just write the line yourself. You said more than what the thing had to do. Like, why are we doing all this?
[00:04:27] All right, so on the surface, as I'm saying this, you may be like, oh, that seems A little counter to this idea of heft and heftosis, but I view it very much in the same, largely in that folks are focusing on and being mesmerized by. Look at how much this thing can do, as opposed to asking themselves, do I actually need it to do this?
[00:04:48] Is it actually helpful?
[00:04:51] And this is a really big point that I, that I want to emphasize for this episode, right? To me, responsible AI, uh, usage involves you asking yourself, do I actually need all of this? Is it actually helpful?
[00:05:06] Sometimes maybe you do. Sometimes maybe you do. But quite, quite often I think you don't. And I want to use this episode. Episode to really caution against equating output volume and output value. Right? They're not the same.
[00:05:22] Now, clearly I am very pro AI, and I think that there are so many use cases. Like just last episode, I was encouraging y' all to be more ambitious with AI, and that directive still stands.
[00:05:33] But to me, a. A lot of these rollouts and a lot of the new features and the developments from these AI companies are really just marketing attempts to see what sticks.
[00:05:44] I've said it 11 billion times, and I will say it 11 billion times more, the marketing for AI is backwards. It continues to be, here's a solution. Go find a problem.
[00:05:56] Rolling out all these new agents and, you know, producing, uh, 100 page reports, and you have the ability to create an entire dashboard in order to present, you know, a response. You ask a question, it's like, I can make you a dashboard. And you're like, my. I don't need all that. Right? Uh, to me, that reads as here are 11 billion solutions. One of them has to solve a problem you didn't know you had. And now you'll find a new reason to keep using this tool. It's giving desperate more than it's giving helpful, especially with how capable these companies want you to believe their tools are. And if it's like, you're like, if it's so capable and it knows so much, should know that this is too much. I don't need all that. Like, just give me a fine answer. The. The other thing that inspired this episode, right, was a, uh, question from a friend of mine who wants to learn more about AI. And of course, I told her, just go use the search tool on Prompting Curiosity website. Ask your questions. But I asked her for specific questions or areas of interest. And the questions, the questions, excuse me, that she shared with me, they really got me thinking about how the answers might differ depending on who gives that answer, namely someone with hypnosis. Versus someone without.
[00:07:03] It seems very plausible to me that anyone enamored with the uh, Morris better approach would in general have more positive things to say about AI.
[00:07:13] To me, the reality is AI can be truly, truly incredible. And also more is not better.
[00:07:23] Better is better in all things.
[00:07:27] I continue to stand behind the approach of looking into and leaning on AI when you have an actual problem that you want solved. Hence why I said earlier I didn't. I still haven't used cloud cowork and I didn't use cloud code for a long time. And then I was like actually need to do something. Like maybe this is the solution I'm looking for.
[00:07:46] Oftentimes this, this motivating or inciting action or inciting event where you're like, okay, maybe I have a use for this is something you don't like doing.
[00:07:56] From there it's just about using your own judgment and determining determining when it should be AI assistance versus AI automation slash full outsourcing and which one is more appropriate for the for the circumstance.
[00:08:10] So to close the loop of learning here, per me, the definition of hypnosis is AI attraction and excitement that is caused by equating output, volume and value.
[00:08:24] Do not be charmed, fooled, hypnotized by the sheer volume that AI produces as an output.
[00:08:31] If you are annoyed by the verbosity bias, um, by the way, don't forget that you can give your AI instructions to simplify and produce shorter responses. You are in control again for all things be it AI outputs, data centers, model size, whatever. Bigger is not better, more is not better. Better is better.
[00:08:57] Okay, one last thing. This will be a shorter episode which I like. Uh, one last thing before our usual wrap up section of how I use AI.
[00:09:05] In case you haven't noticed, AI hacks and breaches and vulnerability exposures are becoming a very real thing and they are only going to increase.
[00:09:16] I am not saying this to scare you or be all doom and gloom. I just want to keep you informed and make you aware if you haven't already received emails from your banks about agent access slash agent activity. Those emails are coming. But like I said in a previous episode, no, I don't think that AI is going to kill us all. But I do think it's going to some shit up. When I don't know what exactly is it gonna up? I don't know.
[00:09:46] I do not know. But it feels very inevitable at this point in terms of acronyms, in case you're listening, you're like okay, what the do I do? It feels to me a Bit like having a bike lock. Like if someone wants to take your bike, they gonna take it, but it could make things a little bit harder. So that's what, that's kind of the approach that I have with this. So to that end, uh, action item is number one. Just be mindful of what you give AI access to. Probably not your bank.
[00:10:12] Uh, be mindful with permissions, be prudent with what it has. Write access, W R I T E access as opposed to just read or read only access. Those are different things.
[00:10:25] I, you uh, know, real talk, I'm not like super stressed about this or anything. I'm just more so aware because it feels like you know, to me being stressed out about this feels like when folks are stressed out about surveillance but they have every Apple product ever, they use a phone, they use their computer, they have a tablet, they live in a major city, they use social media. And I'm like, you are very much being surveilled all the time like what's happening here? Uh, so I'm not like super stressed, I just am aware of it.
[00:10:56] Um, and I'm continuing to do things to stay aware. Uh, my favorite person to follow is Liz the developer on Instagram. Once again that's Liz the developer and uh, she shared something last week or at the time, the time you listen to this, it'll be, it will be last week but she shared it. Today is the day that I'm recording this. Uh, and it was just like to Update Chrome because AI found like 62 vulnerabilities or something like that. And I was like, well it's an easy thing, we can just update. It takes two seconds. Uh, so just little things like that.
[00:11:28] Um, you know, just doing what I, what I can do. So transition there, segue there. Speaking of doing what I can do, that brings us to the final part of today's episode, which is how I used AI this week. In case you're new here, welcome. Each episode I share a quick example of how I used AI this week.
[00:11:47] This week, last week I used AI to do some major, major, major back end work on both my websites. One in particular, um, to patch a security vulnerability that was announced last week.
[00:11:59] So I'm not going to get into the nitty gritty because I don't think most of you actually care. But I run two WordPress websites and last week there was a fairly significant security issue that was announced and it affected WordPress. Um, I self host my sites which means that I need to fix the things myself. A company is not going to do it for me. Uh, and thus my French assistant Claude was put to work. And that guy, he walked me through everything and he held my hand every step of the way. And it was actually incredible. Like, I really. What's one of the. It's been one of the biggest things. I know I've shared this in previous episodes in the, uh, how I use AI segment, but having it help me just get more comfortable with WordPress and PHP and web development as a whole has been awesome. Just a really, really cool experience and feeling so much more comfortable. And what I had to do was actually a fairly significant undertaking.
[00:12:56] And I was like, let's go. Like, I did the easier fix first because I expressed it to Claude. I was like, this is. Feels a little scary to me. And Cloud was like, you can do this. Um, but good call to do this, the shorter one first. And just it took me like two days, uh, to do everything. But we're all good and it has been a really, really cool experience. Just getting more comfortable and familiar, uh, and proficient with, with web development.
[00:13:24] Uh, so, yeah, that is, uh, again, coding and the like is where AI really, really shines. And I was very, very big grateful for that fact last week. All right, that, my friends, is all for today. Hopefully you found this episode helpful. And if you did consider sharing it with someone you know who is curious about AI, don't forget I have a m. Companion newsletter and blog, the Curious Companion, that drops every Thursday. That is basically the podcast and episode. The podcast and episode. What am I saying? That is basically the podcast episode. By basically, I mean exactly the podcast episode in text format. So if you prefer to read or you just want a written record of the things, join the newsletter fam or check out the Bloggity blog, you can head to prompting Curiosity.com forward/newsletter or Forward slash Blog. Or keep it simple and just click the links in the show notes.
[00:14:21] As always, endlessly, endlessly, endlessly appreciative for everybody. Single one of you. Until we chat again next Thursday, stay curious.
