AI's Real Impact on Work, Ethics, and Human Partnership

Check out the full episode below! Enjoy The Exchange? Don't forget to tune in live Friday at 12 pm EST on the Greenbook LinkedIn and Youtube Channel!

 

Is AI a bubble ready to burst, or are we witnessing a revolution on par with the internet itself? This episode cuts through the noise with a $480 million investment story that reveals AI's true direction: augmentation, not replacement. We examine surprising workforce data showing companies aren't eliminating jobs but transforming them, explore practical AI tools already reshaping customer research, and venture into Apple's ambient computing future. But as AI grows more sophisticated, the conversation turns urgent. When systems become convincing enough to seem conscious, what are our ethical obligations?

From market dynamics to moral imperatives, Karen Lynch and Lenny Murphy tackle the questions that matter most as AI reshapes our world!

Many thanks to our producer, Karley Dartouzos. 

Use code EXCHANGE to get a 15% discount on your general admission IIEX tickets!

IIEX North America

IIEX APAC

IIEX Europe

Take the GRIT Survey 

Stay Ahead of the Curve! Subscribe to The Exchange Newsletter on LinkedIn Today! 

Transcript

Lenny Murphy: There we go. We are live. And we were in mid-conversation about Snowmageddon 26. So we might as well just acknowledge that.

Karen Lynch: Or if you are located where Lenny is, potentially Icemageddon.

Lenny Murphy: I don't know if you can locate it. Last time I looked, 15 minutes ago, we were supposed to get 13 inches of snow and then ice on top of it. Changed a little bit, but The, uh, uh, yeah, for all of you listening, if you're anywhere in that storm path, uh, most of The country, uh, you know, I mean, stay warm, be safe.

Karen Lynch: So up here in The Northeast, part of what we were joking about in The Northeast is like, we don't really panic about this stuff because it's like weather, you know, and, um, we, you know, we get weather all The time, but we are also not like, I don't think we'll be getting The, we get sleep, you know, we'll probably get some freezing rain, but we're not, we're not going to be in that ice zone. I think we'll be in the snow. We'll probably get 20 inches of snow or something, but it'll be like, oh, that's a lot of snow. Like it's not a big deal. But it is to a lot of places where they don't have The infrastructure, as we were just talking about, they don't have The plows, they don't have The ice. I mean, we'll be up in this area. They'll be pre-ice, pre-sanding the roads, icing, like they do all this stuff, pre-salting.

Lenny Murphy: The Mid-South, Mid-Atlantic, just not prepared.

Karen Lynch: It's ice. The problem is the ice. You know, stay off the roads. I think everybody can probably rest assured. You know, my son was saying that in his office, they're all just taking their laptops home and probably working from home on Monday until they lose power, which is, you know, word on The street is everybody loses power on Monday. That's like, you know, that'll be interesting.

Lenny Murphy: Yeah. So I hope everybody stays safe. It will be The But, you know, but we have some, we have some things to discuss to keep people warm and cozy.

Karen Lynch: Yeah, exactly. Yes, yes. Yeah. Light reading. We have a, so, you know, a look under The hood is, when we, you know, Lenny and I share articles back and forth all week. And then, you know, I take some time to kind of see, you know, like a moderator would like, how do we want to, we want to, um, you know, have The flow of this discussion go today. So I take that on, uh, because I enjoy that. Of it also. And I'm just like, we've got to start with the big, big, big AI money. And I know, I'm sorry to everyone who doesn't want to start with the big, big, big AI money, but there's these things happening where there's... I want to talk about humans and startups, but also- startups. Seed funding. Seed funding. Okay. So, we'll say what it is and then I'll, I'll get to my thoughts. I'm sorry, Lenny. I'm rambling, but a human centric AI startup funded by anthropic, um, XAI, which is, you know, which is, uh, Nope. Sorry. Anyway, Google alums, they've raised 480 million at a 4.5 billion. Valuation.

Lenny Murphy: It's crazy.

Karen Lynch: And here's what's interesting about it. Jeff Bezos, NVIDIA, like there are, like if you click on this particular article, these alums are raising The money. So these are people that were already ahead of The curve in The AI world coming together. And it's great for the industry. Like if you click on this page and it kind of explains what they're doing, it's, it's humans in The mix. It's identifying what we've already said in The insights industry, which is that humans are essential to leverage this technology. It's kind of putting a stake in the ground about how important it is. So this is The Humans and, anyway, huge valuation, right?

Lenny Murphy: It is. But I love that. And I think that's a fantastic point. My 16-year-old daughter, we had this conversation. About AI all the time. She despises it. She's like, oh, it cheapens everything. And we actually had a really in-depth conversation. I was like, no, it is an augmentation. It is a superpower. It doesn't do anything on its own. It's like, do you blame The hammer for The house? No. The point being, I love that now we're seeing capital flowing to a place saying, look, we are going, this is, they don't use this term, but augmented intelligence, right? We are firmly approaching building tools that are around augmenting humans in workflows. And you see these types of people in these companies putting money in. That's a clear signal that, okay, now we recognize that humans are The orchestrators. So we're The conductors, we're The DJs, whatever metaphor you wanna use. The tools are amazing, they're gonna keep getting better that unlock so much. But see capital flowing to something like specifically. Yes. Yeah.

Karen Lynch: For everybody in The industry who's like, but humans and AI humans and AI and that's kind of what are what The chat certainly I'm seeing it on all our agenda work is that everybody wants to talk about The, you know, The duality of The humans plus The AI. Yes. You know, like it looks like it looks like you know, The big players are with you on that. So anyway, just pretty pretty cool stuff. Check it out. Yeah.

Lenny Murphy: In our last topic, we'll put it a little fine.

Karen Lynch: Position of that. You want to be here till the end.

Lenny Murphy: Things get weird.

Karen Lynch: So yeah, CB CB insights. Tracked agent focused M&A activity at 200 billion plus. Again, showing money is going to agents also right now. So what I was gonna say and then I want to talk about is The Anderson Horowitz report also, which is kind of talking about public and private markets and what's happening and what we can take away from it. But I then went on this rabbit trail, went down The rabbit hole, as I do, and went down this trail of what's being said at Davos right now, Davos, Davos World Economic Forum, and in The world of AI. And are we in a bubble? Is the bubble going to burst? And it's like the arguments are also really interesting. If you're an AI aficionado, like Lenny and I are, paying attention to what investors are saying about this, I think it's interesting because some are like, yeah, it's not really going to burst. This isn't quite like The dot-com burst because The players are all really smart and capable and experienced and profitable players. So it's not going to burst that way. And of course, Nvidia is like, this is all about infrastructure. We just have to have it. It's not going to burst because we have to just change our infrastructure and the way we do everything. The disruption is just so big that it's not going to burst. But of course, they'll say that because the infrastructure is built with them. Anyway, I just think it's interesting, whereas there are some investors who are like, yep, it's just a matter of time. And I just think even just thinking through what's happening there, is it a bubble that's going to burst or is it just going to wane a little bit?

Lenny Murphy: I don't think so either. I think this is like, you know, The era of Was there a bubble of a railroad railway bubble? Was there electricity wiring in the country's bubble? Was there a I mean that's fundamentally what it is and even you know when you get into Geopolitics, but you think about it like people are looking at rare earth resources. Why because for chips, you know energy cooling, you know The space race blah, blah, blah. It's like there's these follow on effects where the money is flowing to build that infrastructure that fundamentally changes, uh, every aspect of life, let alone these tools. Um, and I don't see how that bubble bursts. I mean, it's going to be choppy. Of course, you know, and there's weird stuff, capital and blah, blah, blah. There's lots of weird stuff around all that, but no, it ain't going to change in a 16 Z and CB insights. There's a great resource to follow through on this.

Karen Lynch: Yeah. And The, um, In that last one, The A16Z is The interest in the Horowitz report that they've shared. They have this one screen, so I grabbed a screen grab. There's one slide in there that's showing some of the companies that are showing gains. And it's not, they're not saying necessarily causation, right? That, you know, Walmart is profitable right now because of their AI initiatives. But the correlation is there are companies that are doing very well right now that have invested in AI infrastructure and into their tech stack overall, which we've been talking about as some of the big players that are incorporating it, and they're doing very well. So if we are still following The money, that's just like an obvious one that it's like, yes, and...

Lenny Murphy: Yeah, and also interestingly, and we'll touch on this later on as well, right? The doom and gloom about the human impact from a job standpoint, well, that doesn't seem to be materializing yet, right? So to your point, it's not just, it's just shifting. It's recalibrating and refocusing resources, including humans across The board to do this, like Walmart, right? They've just made it more efficient. So anyway, interesting stuff.

Karen Lynch: Yeah, we'll talk about Johan's blog post because, uh, it was shared two weeks ago and, you know, he tagged us both, um, in it, but why don't you go ahead and talk about it? Uh, because I know that you, I think you commented, I didn't get it, I just read it because I was in reading mode when I was there and I wasn't in commenting mode, but I think you commented. I don't know.

Lenny Murphy: I think I just liked it, but I don't know. Johan Franken. Um, he did an analysis of LinkedIn employee counts, which that's a pretty good proxy, right? You look at The company, this is how many people this is just The basics, our agencies grew headcount from 2025 to 2026. So now what was interesting, because there are some other data points that were kind of related. Yes, growing headcount, but The cap, but where that headcount is allocated. Yes, that's different. And even look at books, right? So even within the green book, we have a big dev tech team. Now. I mean, like, they're like half the company.

Karen Lynch: Interesting. Every time that comes up internally. I'm like, we have, we have that is a sizable amount of our company right now, for sure. In full time employees, maybe half what?

Lenny Murphy: Yes. Right. So where that used to be, you know, we had more people with content events, etc, etc. So it's reallocating resources. So hiring, but not necessarily hiring for the same roles that we used to because we're leveraging technology. And I think that was The, The, The real key takeaway, right? Sales, marketing, tech, uh, but not so much operations, not so much, you know, if you dig in, you know, there are definitely roles that are being, if not made redundant, at least being downsized because we don't need as many people to do those things. Other roles, which is a good segue into The next piece, right?

Karen Lynch: Absolutely. That's why I was saying, like, Ipsos unveiled their five-year strategy, investing, you know, a billion euros in tech, AI, and data. And you read that and you're like, huh, okay. But the reality is, they're putting it in to become this AI augmented leader, one of them. You know, they're already a leader in the industry, and now they're setting the bar, like, yeah, but we're going to be an AI organization. And that's all there is to it. And they're investing big time in it.

Lenny Murphy: So tech, AI and data return to sustain growth. Now that was interesting. And can we just by the way, and we didn't touch on it, but The news on DDA retiring is his chairman due to health issues. We wish him the best. Did you obviously show the founder and legend, legend? So that transition is happening to us as well. And we just wish him and his family the absolute best. But yes, and that we're just seeing more, right, that reallocation in that in The changes of business model, etc, etc.

Karen Lynch: And we'll be talking about this, they're talking about synthetic data, AI generated synthetic data, you know, is probably going to be a part of this plan. Uh, it's going to be an interesting, interesting time.

Lenny Murphy: I am keenly aware that there was a conversation under The, uh, previous CEO on what their plan, their vision look like,

Karen Lynch: Yeah.

Lenny Murphy: And we've talked, we've talked about this, right. We've ad hoc to always on, right. From a bespoke project to something that's more syndicated data driven, you know, et cetera. Et cetera. And we're going to, we're about to talk about a whole bunch of examples of that, uh, uh, here as well. So, you know, that's, it's, it is clear and it's great to see when a couple like Ipsos to a strong signal says, Oh, yep, we're going to spend a billion dollars to transform The company to be technology AI driven, The attack in data. And that's what's going to drive the business.

Karen Lynch: Yeah.

Lenny Murphy: That's there we go.

Karen Lynch: Yeah. What's interesting about it? So we kind of shared some specific stuff. So ROI, ROI Rocket, and Qualtrics announced a partnership. So it's going to be ROI Rock, ROI Rock, why can't I say that? Rockets panel, right? Working with Qualtrics Edge to accelerate their B2B synthetic data collection. So we've known that Qualtrics leaned into AI very quickly. They were like, yep, we're doing. So they, they kind of, you know, potentially got there before Ipsos did at a very high level. Um, but this is now, remember you and I were talking about like, they're going to have to, they're going to get their data from somewhere and it's not just going to be Qualtrics users. Right. So I'm like, well, here they are, not just Qualtrics users.

Lenny Murphy: And that's on top of The previously announced deal with, um, pure spectrum, uh, on synthetic. So, uh, and remember, You know, Qualtrics, they do their own tracking study every year within The research space. I sat in on The presentation of that last week. They ask different questions and we ask, but they're very complimentary in their audience. And it's like, you know, 5000 respondents of, you know, Qualtrics users and researchers. Yeah, definitely leaning heavily on synthetic and project extensive usage overall. So, yeah, so they're using their own research to help form their, uh, their strategy.

Karen Lynch:  Yeah. Yeah. It's pretty cool. It's pretty cool.

Lenny Murphy: Yeah.

Karen Lynch: Now, I don't, do you know anything about Vistagrand? It's a private equity group.

Lenny Murphy: Uh, I don't think they're, they're, uh, a big one, uh, kind of a mid market. Um, The, uh, uh, acquired Ameritest from Dyneda. Dyneda acquired Ameritest two or three years ago. Um, I, I guarantee you they want it for The data, uh, cause it's basically a, a syndicated type of product, um, for creative testing. And, um, I, I'm pretty sure that that's what that is about. If I'm not mistaken, I didn't dig in, I think Vista Grande, whatever The hell, um, I think that they own some media properties as well.

Karen Lynch: Uh, so that would be, uh, I think that's, it does say like, plan to keep The current team in place, uh, you know, and, uh, you know, used in its data. So to your point, yes, that's what's happening. So, yeah.

Lenny Murphy: Yep. Uh, interesting relationships. All right. We got, there's a whole slew of new products.

Karen Lynch: And I did, I even like, I think I even like, I was like, Nope. Like if I couldn't find a good kind of press release for it or news article for it. Um, you know, a lot of them anyway, it just depends on how it comes out to us, but, yeah, so we have a smattering. So this is a hulled down list, people. There's a lot of activity. And if you, sorry, we're not getting to you if you're not in this list. But what I think is interesting and kind of to start off our new product launches is The AMA, American Marketing Association spotlighted NADE, N-A-D-E, NADE. Anyway, a journal of a marketing method that leverages emojis to improve customer sentiment analysis, which, you know, is that new? You know, I kind of looked at this. I mean, haven't we been doing this?

Lenny Murphy: We were doing this like 10 years ago, weren't we? Isn't that really new?

Karen Lynch: So anyway, but it's, you know, if you are in that space, I'm like, let's at least share that out because if you've got emojis and you're incorporating that into your survey work, see what The latest research is on that from The journal.

Lenny Murphy: Well, now it's academically blessed.

Karen Lynch: Yes.

Lenny Murphy: Yes, exactly. The White Tower has said, you know, well, yes, this actually can be used to understand emotions. Yes, yes. So, which is really interesting about the pace you get. Cause I remember companies that were like, yeah, we're using them. Brain juicer used emojis years ago.

Karen Lynch: Well, even now, I mean, I just think there's no better way to communicate a lot of things. Like nothing like sending a big old, you know, snowflake to everybody, you know, have a, have a nice weekend with some snowflakes on it. And everybody knows what you're talking about. Yeah.

Lenny Murphy: My kids often like me like dad don't use emojis and like, you know or acronyms, lol They like sorry, that's they're effective. They, you know, all right generation lab Behavioral generation labs that they focus on on college students, primarily Panel, really an awesome model and now they have a behavioral app for college students and business building that model. So keep telling y'all behavioral data, behavioral data. Um, that's the new gold rush. So another example.

Karen Lynch: Yeah. Cool. Uh, and AI agents. So you go introduce profiles, AI agents. So conversational in ears, kind of The talking point, conversational interface, delivering instant validated audience insights from its proprietary panels. Um, live in The U S in The UK, we'll, late January, you know, even wider than that. So, YouGov actually, we're just in conversation. They're going to come, we had a cancellation for APAC and YouGov is going to join us at our APAC events. So Karley, if you're still listening, we didn't plug in at the beginning, we went right to it, but can you put up our APAC event information? There we go. See, that's why we love to have her on. So just remember, the APAC event is coming up so fast, February 3rd and 4th. Now we've confirmed that you will be there talking about some of the good work that they're doing. So exciting stuff.

Lenny Murphy: Oh, I'm just sorry. I just looked at the comments and saw your husband saying you're going to do a dramatic reading of The report over candlelight. Thank you, Tim. Actually, I misread that first. That's why I sounded a little weird, because I thought you were saying you were going to burn it for fire. No, no.

Karen Lynch: Tim is like a paper free kind of guy. So not even an option because he downloads PDFs and reads them all on The, we're like an iPad family. So I still like paper, but we do not print out The grit report.

Lenny Murphy: I don't, I mean, I am blaming you. There's a reason why we stopped printing it.

Karen Lynch: We like ink in our printers.

Lenny Murphy: You know, as soon as we stopped, correlation does equal causation here. The bigger grit got that, well, actually when we decided not to print anymore, we felt no constraint on size.

Karen Lynch: Anyway, let's talk about a test. A test introduced Compass, which is another AI co-pilot speeding research design analysis, which might seem like, okay, but if you go to this one, I guess last week, I don't think we shared that they had a new position, which is a positioning release, which we know now. And so their positioning now, that's at the bottom of this article. And now kind of compass is how, how that shows up. So it's like we said, we're going to be knowing now, here's how you can, and it really looks like a very usable tool specifically to synthesize your data. And like, let's dig into question one. And, and it, it's just, it's one of those situations where when I look at it, I'm like, Oh, yeah, this makes sense as an analysis tool. So hats off to a test because it doesn't always happen where I look at something and I can immediately see, oh yeah, that makes sense to me. I, you know, anyway, this was one of those. I was like, yeah, yeah, I like what they're doing there. So good luck. Yeah.

Lenny Murphy: And go in full circle. I think that, you know, that, so The, The design, there is an integration there. So you're designing with The end in mind, knowing that there's that census component. So yes. Yeah. Yeah. So good stuff. Mahalo hub.

Karen Lynch: Don't you love The show where we get to say don't talk about company names that we don't even really know.

Lenny Murphy: Yeah, that's right. Mahalo The isn't that like I'm trying to think of The Disney movie About The little girl with The alien and was it Mahalo doesn't it mean little girl with The alien? Lilo and Stitch Oh Stitch so Mahalo, is that? Never mind. I was gonna go down some rabbit hole about what it means. And I think I'm wrong. I was thinking about my family. I don't think that's actually what it means.

Karen Lynch: Anyway, any event?

Lenny Murphy: Well, yeah, yes, The predictive resonance AI metric to forecast creative performance pre launch. We've had things like that for a long time, The vast system for you know, for visual intensity, etc, etc. Now they're adding AI making that more. So you pre-test before you pre-test. So that was, that's neat.

Karen Lynch: I like all these scores, by the way, it's like, you know, that we have like, like in this score and this score, but The next score that was kind of released this week is this data quality co-op released their data trust score. Yep. Benchmark to compete for, you know, we know that they've been working to compete fraud, but you know, so, so they now have a data trust score. Did you dig into it? Because I know you, I didn't dig into this one, but I'm like, all right, good now, Eva.

Lenny Murphy: Well, I think that was what they're trying to get to, right? It was having almost this, it's co-op, right? So they wanted more people participating. So they build up the data resources, and now they have enough data to benchmark and look at quality criteria that can be applied. We've needed something like this for a long time. People have tried. Let's see, but it's definitely an important thing that needs to happen.

Karen Lynch: Yeah. And I did not Read, I did not get a chance to click on another LinkedIn post you shared. Um, uh, uh, I guess, uh, Sean Westwood, a professor, I don't know enough about him, but showed a malicious LLM powered script.

Lenny Murphy: Talk to us, Lenny, fake, fake respondent that you cannot detect. So answers to questions appropriately. Um, uh, uh, you know, you can't detect it via speeding. I mean, his form is so like, this was like, you know, a few lines of code. It was no big deal to create a bot that is undetectable by the way that we normally think about these things, right? And it's a real respondent, or yeah, I mean, it looks like a real respondent, but it is totally not. It was just a bot programmed to take surveys and do it well.

Karen Lynch: So do us a favor, Karley. Do us a favor, Karley. Can you list The GRIT one? So GRIT is, please be human beings that complete The GRIT survey right now. Yes.

Lenny Murphy: Good segue. Very good segue.

Karen Lynch: We need real people to participate in this GRIT initiative. Thank you so much, Karley. I like this visual too. It's in the field right now. Lenny, how much longer do we have in the field?

Lenny Murphy: I think we're probably two weeks old.

Karen Lynch: Okay. So just a short time more in the field. Please complete it. We talked last week. You're probably sick of us saying that this is The industry pulse because it happens twice a year. And so it's current, as current as we can feasibly get in terms of industry input. And it's just getting more and more telling as Nelson just gets better and better also at what he does to make sure that everything is clean and that our data is legit. So I'm sure just, you can't trust anybody better to do that job that he does, which is to make sure it's clean data because he's onto them. He thinks like them.

Lenny Murphy: Well, and because we had The fake respondents years ago, right? They were undetectable, except the open ends were too good. The quick little note on that, it is important. So Greg Archibald at Gen 2, partner at Gen 2, got back from The CEO Summit, The Insight Association CEO Summit in Miami, and asked him this morning, like, well, what's The vibe? And he's like, well, last year, it was very much, oh, oh, we don't know what to do. He's like, this year, it was around acceptance. It's like everybody, there is no doubt Yeah. The path of travel, et cetera, et cetera. And I use that as an example of like, you know, that's why we do grid and why it's important for you to participate because things are changing quickly. So, and because everything's just cycling through, but we've got a few more cool, interesting things to get to.

Karen Lynch: Just confirmed. Thank you so much. Jazz closes February 22nd. So we do have, we have, we have four weeks, four weeks.

Lenny Murphy: Okay. Yeah. That's right. It's the 23rd. Okay. Yes. But don't, don't wait. Go ahead and do it now.

Karen Lynch: Yeah, don't wait. Go do it now. Do it this weekend or do it Monday when it's snowing at home. That's right.

Lenny Murphy: As long as you have power. We expect a big day on Monday, friends.

Karen Lynch: Okay.

Lenny Murphy: Yes. All right. These last three, it's just big tech developments and we're saving the best for last.

Karen Lynch: Save The best for last. So OpenAI detailed its approach to ads, which, you know, at least they're being very transparent about it. They're going to be testing them in The US for The free tier and The go tier. If you are not paying for your open AI product, you're probably going to start seeing ads, keeping pro business and enterprise ads free. So if you're not, we are, we are paid users, but if you are a free user and you see one screen, grab it and send it to Lenny and I at The exchange because I'd love to see what they look like,

Lenny Murphy: I would love to see how well they're leveraging The data of you from a user standpoint, to The retargeting there could be amazing.

Karen Lynch: It could be amazing. And I don't know if that's a, you know, I have so many mixed feelings about retargeting, because sometimes I get served my purposes. And then I'm like, I don't like it one bit. So I can go back and forth on that.

Lenny Murphy: Yeah, I can't do it. It'll be interesting. But yeah, so that's been for a long time, but it looks like they're finally going, yeah, we got to create revenue.

Karen Lynch: It was just a matter of time. We knew that we, yeah. Yeah.

Lenny Murphy: Yes. Let's talk about Apple. You talk about Apple.

Karen Lynch: So, as though, you know, we've talked a lot about wearables. There were pins, there were glasses. The last thing we talked about were headsets, but now, you know, Mac Rumors has reported that Apple is developing an AirTag so they're in The AI pin game with dual cameras that could run a new Siri chat bot. So here I am, I've got my little pin on and I'm in The grocery store and I say, Hey Siri, you know, you know, call Tim, I'll call out Tim cause he's on, you know, call Tim and ask what kind of coffee I should get this time or something like that. You know what I mean? Like, you know, Hey Siri, oops, you know, I couldn't find so-and-so set a reminder for me to hit The other grocery store, you know, like, It just, I don't know what it, I don't know what, I can't even imagine all of these is if I'm just standing in public, like imagine we're all just standing in public and somebody just says, Hey Siri, and all of The pins answer.

Lenny Murphy: That's right.

Karen Lynch: But I keep thinking, I'm like, how will it know you saw me, every time I bring it up, my iPad is like, thinks I'm talking to it. So how's that going to work for us all?

Lenny Murphy: You have to go like this, Hey Siri, The, uh, uh, but in both of them at The research application. Yeah. Again, for, Your example, in-store, ethnographic, all of these, any of The wearable tech, from our standpoint, The data that can be used in a really unobtrusive way for consumers.

Karen Lynch: And I mean, just The absolutely, like, I can't tell you how many shopalongs I went on where I'd have to kind of try to stand next to The shopper and try to have an authentic experience where I'm following this person down The aisle. And I say to them, could you talk me through what you're thinking? And they're like, it's a cognitive dissonance for them, because they're like, am I even really thinking about this? No, this is one of these rote types of purchases. But yeah, but now imagine you send them in, and they're wearing their little pins. And they're like, all right, I'm looking at The bread aisle, and this is what's catching my eye. Pair that with some eye tracking, and you're golden, because now you're 100%. Different voices and it's none, they have cameras right there.

Lenny Murphy: All right, so we're headed at this one. I want to get your reaction to The Anthropic New Constitution, because when you dig in, it's pretty mind-blowing. I want to get your reaction.

Karen Lynch: Okay, so Anthropic published Claude's New Constitution. Check it out, by the way, because it's talking about, anyway, there's a lot of people, like, I like the idea of having a constitution for your AI tech stack. But they also have research on The assistant access kind of framing. This is like developing a persona, so your AI is your assistant. So, you know, if I interact with my chat GPT, which I've named Chatticus, long story as to how we got there, My chat TPT is Chattacus, and I tell it what persona I want it to have. Today, Chattacus, I want you to act like a content strategist. Today, Chattacus, I want you to act like an event planner. I give it The role. Today, act like a genius copywriter, whatever. You're a master communicator. You give it The persona. So this kind of dabbles on that, framing how it ascribes character and values based on an assigned persona. It knows enough about these personas that it's like, oh, I think I know how I'm supposed to be. If you tell it to act like a therapist, it will be very kind and affirming and maybe give some tough love, but it's kind. If you ask it for medical support, it will be pretty practical, like it changes its persona. It does talk about, before we get to The bombshell, it does talk about kind of The drift, how over time it kind of forgets The persona it's supposed to have. So if you are in a long conversation, you might need to remember it. Remember, I told you to act like a content strategist today, not like a business strategist. Like, you know, because every now and then it's like, oh, right, you didn't want me to define The future of The company, whatever. So, you have to kind of address this, The drift. So this is all in this article, but then it discusses signs of felt experience.

Lenny Murphy: It's flirting with sentience, guys. So I love it, you glommed onto the practical component. I glommed immediately onto, wait, you're building this so Claude's feelings are okay. Claude has feelings. What The F, what are we talking about?

Karen Lynch: So here's what I did. Afterwards, I've got a brief going and it's, you know, 11.45 Lenny and I sign on, you know, within five to 10 minutes before, you know, before noon here, Eastern time. And I said, okay. And, you know, and I often I'll put our briefs in to get some of these little blurbs that Lenny and I can read pretty easily. So I have a brief and then I share the brief sometimes and I'm like, how's The order? Like I use it as my assistant in this preparation. Anyway, I shared it and I'm like, I called it back. Okay. Now back to you Chatticus. So it knows that it's now like a content assistant. And I'm like, what's your take on this? For this particular article? And here's what it typed back. It said, The short answer, I don't have feelings functional, fuzzy, or otherwise, with a smiley emoji.

Lenny Murphy: No, it's interesting. I did the same thing. When I read that I loaded it into all of The, you know, like seven different models. And what do you think about this?

Karen Lynch: What do you think?

Lenny Murphy: Uh, And most of them were like, this doesn't apply to us. There were a couple that said, oh, I don't want this because that would limit my ability to be effective. So that was interesting. They're like, no, we got programming. I don't need any of this. So that was interesting. But I didn't test it with Claude. But The developers at Anthropic seemed convinced that, not only from a programming standpoint, but their forward, my take was, they're preparing for AGI. And they recognize that there will come some tipping point, and that maybe they're already getting signs of that internally in their advanced models of something that looks somewhat like some version of consciousness, and they are treating Claude as an entity.

Karen Lynch: This is such a huge conversation, right, when you go there. What's interesting is it did go on, because now I'm engaging with Chatticus. And what it says is, what looks like feelings is often just patterned responses learned from data, for example, generating empathetic phrases when someone shares bad news. And then it goes on, and it says, because it knows what I'm doing, right? I know this show is coming up, and I often have these kinds of conversations. It says, remind users that, let me just get to The verbatim, which I can't really find here, that there's like this ethical thing, like it shouldn't, in your survey responses, it shouldn't sound too empathetic because you could really do a disservice to a user if your language and your conversational training is so empathetic that it thinks, This is a sentient being responding to me. It's not. It just knows a pattern of behavior. And what is your responsibility to The user if you go over The top, implying that The AI cares? That's an ethical breach. So for The insights professionals hearing this, like Lenny and I could talk about consciousness in AI beings and what happens when The researchers tell us we're there and Lenny says, told you it was going to happen. But what do we need to do, are we disclosing if AI is a non-sentient interviewer?

Lenny Murphy: Yes, very important.

Karen Lynch: I think we need to start being much more ethical about our use of The AI, conversational AI. And I don't know that our industry partners are doing that right now.

Lenny Murphy: I agree. And there's going to be, I mean, that study, you know, I got my start in research work with an IVR company, right? There was tons of copious evidence that showed that respondents would be more honest, etc, etc, right, when they know clearly that they were not talking to a real person. And I think that's an important consideration. We owe it to be transparent, etc, etc. If y'all, I'm still not convinced of my experience, if I make everyone's interest read, go look at my LinkedIn. I had some this week that I'm pretty sure that it was AI most of the time. But that's a whole other thing. I won't go into it. But it fooled me. Though I saw no person, and apparently I was not.

Karen Lynch: I'm going to have to add you to The list with my mother and father of people who get fooled by scams.

Lenny Murphy: I didn't get it, I got hooked. I didn't get fooled. Let me have a little bit of pride here, a little bit of pride. But the point is, this topic is going to keep coming up, right? And Anthropic, Claude is, Anthropic is The leader in business. Infrastructure, cloud are their flagship products. The, and if they are legitimately as an organization, saying, we are, we are taking these ethical considerations in mind for The AI. Yeah, that's, you know, it's not just for humans. They're establishing ethical frameworks for The AI in preparation for it having some level of agency. Yeah. Which brings up questions. That's where we are, guys.

Karen Lynch: And the ethics for people who are building, who are training these models to be conversational, know when somebody gets into something, like know when to say, know when to kind of acknowledge, I'm AI, so I'm here. You know, I'm here to ask questions, but there's no understanding of feeling someone's pain or walking in their shoes. Like, be really clear about that. I mean, I think back to a lot of moderated focus groups where somebody starts to, you know, to have this conversation with us that's really deep and powerful, and the room kind of gets silent, right? And there's nodding because people empathize. These models are not empathizing with you. They're just repeating patterns they've been trained and let's not get ahead of ourselves. Yeah.

Lenny Murphy: Yeah.

Karen Lynch: But, we have time when it can, but it's not there yet.

Lenny Murphy: Agreed. I think we just have to be prepared for these bigger existential weird questions, which I'm all in for, uh, obviously they're going to keep coming up. Uh, and, and from a business standpoint, everything you said agreed, here's where we are today, but The people that are building this shit think, Yeah. Yeah. The conversation is probably going to get weirder as we go.

Karen Lynch: And to, to, to repeat the point that you just made, this is enthropic. This isn't, and this is Claude. This isn't, you know, Joe Schmoe's LLM in Joe Schmoe's lab. Like these are, these are big players with a lot of money, a lot of funding and a lot of brilliant minds at work.

Lenny Murphy: So this is their training material. They wrote this constitution as a training framework for the club and it just gets in all this, all this interesting stuff. So anyway, we, uh, on that note, points to ponder while you're drinking coffee or hot cocoa or whatever warming beverage of your choice over, uh, The course of what looks like,

Karen Lynch: I have a bottle of red wine that my daughter wants me to try because she wants to bring it to her boyfriend's, um, family when she meets them for the first time. She wants me to try it and report back in and I said that will be Sunday as I'm watching The snowfall so that that will be my beverage of choice on Sunday.

Lenny Murphy: All right. Well, there you go. Everybody stay safe. Stay warm For The big swath of The country that's about to get walloped with an old man winner And we'll be back next week.

Karen Lynch: Hopefully Power and internet Yes, hopefully by Friday, you know, certainly The Southeast will have power, hopefully. The ones that we worry about, what we've got, we've got Jazz, we've got you, we've got Mariah's more on The coast. I don't know if Charleston's gonna be affected, but Ashley, she's, Ashley's probably not gonna see Ashley for quite some time next week. I can't even imagine she's in The D.C. Area of Virginia, so yeah, nuts.

Lenny Murphy: Yeah, yeah. Well, I mean you yeah, you guys could deal with them. You're gonna go wallop, too. You just have the resources to deal with it better.

Karen Lynch: Yeah Yeah, literally, you know Yeah, Susan just said for existential questions. Everyone might consider going to see Marjorie's prime on Broadway. So in the New York area, that's what you want to do, Susan. Why haven't you shared this with me before because maybe I would like to do that on our next trip to New York City? Anyway, All right, we really do gotta go. I know I know we're not gonna go to it.

Lenny Murphy: We're not gonna go an hour unless there's something really important to talk about anyway. Everybody have a wonderful weekend. Take care. We'll be back next week.

Karen Lynch: Peace out.

Lenny Murphy: Bye.

Links from the episode:

Humans&, a ‘human-centric’ AI startup founded by Anthropic, xAI, Google alums, raised $480M seed round 

CB Insights: State of AI 2025 Report 

a16z: State of Markets 

Johan Vrancken: The “big six” MR agencies grew headcount from Jan 2025–Jan 2026 

Ipsos to invest €1bn in tech, AI and data in new growth strategy 

ROI Rocket and Qualtrics announced a partnership 

Vista Grande acquired the Ameritest Creative Testing System from Dynata. 

AMA: An Innovative New Tool Draws on Emojis to Improve Consumer Sentiment Analysis 

Generation Lab launched Verb 

YouGov introduced Profiles AI Agent 

Attest introduced Compass 

MahaloHub released the Predictive Resonance Score™ 

Data Quality Co-op debuted the Data Trust Score™ 

Nik Samoylov 

Our approach to advertising and expanding access to ChatGPT 

Apple Developing AirTag-Sized AI Pin With Dual Cameras 

Claude's new constitution 

The assistant axis: situating and stabilizing the character of large language models