Why AI Still Needs Us (More Than We Think)

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!

IIEX.AI revealed a paradox: while AI tools advance at breakneck speed, successful implementation still depends on human judgment and organizational culture. Most companies remain stuck in experimental mode, using AI for efficiency gains rather than transformation. Meanwhile, Walmart's ChatGPT partnership signals conversational commerce is here, MIT's self-adapting AI raises autonomy questions, and the real competitive battle is shifting to unglamorous data preparation challenges.

Karen Lynch and Lenny Murphy unpack why the gap between AI capability and adoption isn't technical—it's human.

Many thanks to our producer, Karley Dartouzos. 

Use code EXCHANGE30 to get a 30% discount on your general admission IIEX tickets!

IIEX North America

IIEX APAC 

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

Transcript

[00:03] Lenny Murphy: Here we go, happy Friday.

[00:05] Karen Lynch: Oh, happy Friday, for real, for real. It was one of those weeks.

[00:10] Lenny Murphy: You had a busy week.

[00:12] Karen Lynch: I did, I did. You know, our virtual events are both kind of, you know, a blessing and a curse for me specifically because I'm able to see a lot more when it's virtual because I'm not running around and being intercepted by people and having meetings and, you know, The drill when you're live at one of our events, always doing something, whereas our virtual events, I can actually listen and learn. So I think that's why, you know, it was great. And then also, it's just so many hours a day, just learning, learning, learning.

[00:42] Lenny Murphy: So, yeah, well, that's a segue. So why don't we start and for those who don't know what The hell we're talking about, we are IAX AI this week, and I was only following along via LinkedIn, all those posts. Unfortunately, I was doing other stuff, but you were in the thick of it.

[00:59] Karen Lynch: And I know I was in the thick of it. And first of all, shout out to Conveo, because not only were they our title sponsor for IEX AI, but they are sponsoring this Episode as well. So thank you again for them, for your sponsorship for both this show and for The last two days of great content around AI. So here's some takeaways, because I was able to look at the data that Kara shared quickly overnight. I don't know if you saw that, Lenny, but quickly overnight. Which sessions were the most attended. And like, shout out again to Conveo, highest attendance of all of The sessions, which I just think, first of all, you know, it's a sweet spot. The title sponsor always gets that sweet spot, you know, early in The day on day one. But also what they were talking about, you know, was really important. So their session was called Beyond Scale, Authentic Human Understanding with AI. And I think there's an event in that title that 's like, okay, authentic human understanding and AI, an overarching theme. When you're pairing a kind of human insight or human intellect with artificial intelligence, those two things are the sweet spot where our industry is right now. The partnership, the collaboration, AI plus humans, that seems to be what everybody was tuning into. So, you know, Conveo nailed it right off the bat.

[02:20] Lenny Murphy: Can I interject something real quick, too? Yeah, of course. You know, let's put this in perspective. This is a three-year-old company.

[02:29] Karen Lynch: Yeah, it's a good point to call out.

[02:32] Lenny Murphy: You know, and there's others as well, right, in that space, that all of them are just a few years old, going from zero to 100. You know, I mean, just from a growth and impact standpoint, that's sweet. Let's recognize that. That's it's pretty cool to see companies and such people that we know like Niels right there forever. Just seeing that rapid transfer. It's emblematic of The transformation. And we see that in a very specific commercial way. They have money to spend.

[03:06] Karen Lynch: Yeah, they do. Certainly, you know, I just thought about sponsoring another industry event, a smaller one. And I was like, All right, I see you.

[03:16] Lenny Murphy: I mean, that's part of success. You companies who are not succeeding don't get The investment to spend The money to help accelerate their growth. So you think, oh, these are, you know, flash in The pans, new stuff. These companies are hitting scale.

[03:31] Karen Lynch: And, you know, shout out to Knit. So also Knit, another solidly attended session, another plenary sponsor for us. And they were talking about black boxes, fraud farms in The last mile. And that, you know, from a title standpoint, but again, like hitting on a hot topic right now, like, okay, so like, yes, we're talking about all this, we have to be talking about The other side, right, The risk, kind of The fear, The trepidation around fraud, and about data quality on all that. So, you know, hats off to Anish and his team, too. So, you know, really, really solid. You know, I've got to shout out some of the brands that spoke, oh, gosh, and I didn't even get to Tom with Jack in The Box. That was a fantastic session. I was listening to that thinking, this right here is why he's a pro. But we also had Lance Hill from HP talking about work not working, and Milind from Mercedes-Benz talking about The symbiotic workplace, and even Melina Palmer joined us on day two talking about why change fails in The workplace. So I think that was another theme. These were very well-attended sessions. And they were talking about what all of this disruption and rapid change around AI means for work. And, you know, really that is of interest to people. Anyway, I just need to kind of shout out all that as well.

[04:52] Lenny Murphy: How often was the agent said? Was that The new, if that was The new drinking game, would everybody be wasting The universe?

[04:59] Karen Lynch: It wasn't, it wasn't, because, yeah, no, I mean, there were certain sessions about agentic AI on day two in particular, but that wasn't necessarily The Conversation, and I can get my takeaway on that. If I fast forward to something that we talked about at The town hall, also sponsored by The Harris Poll, I was able to facilitate a town hall discussion. And The Harris Poll, through their Quest DIY programming, had shared this study that 98% of insights professionals are using generative AI or have used it in The last year. So great, great onboarding for people within The insights community in their sample of insights professionals. But what they're using it for is like these very specific use cases. It's really The assistant. It is The person who is, you know, helping us create briefs. It's helping us, you know, write draft proposals, communications. It's helping with data and that data and analysis and The on The analytics end. So it is this great efficiency and I think that the current use cases haven't really gotten too far into agentic AI. We talked about it, that's going to be coming up soon, but that is not omnipresent right now. Does that make sense?

[06:18] Lenny Murphy: It does, and I think that's probably, if we looked at it from a broad category, that's still The new wave, right, being rolled out. It's not fully baked yet. But from the client perspective, the process is efficient. And when I think of Agentic, I think that it's The next wave, The accelerator of process efficiency.

[06:41] Karen Lynch: And you know, what was interesting, again, going back to this town hall, we had somebody from Meta join us and somebody from Robinhood join us, so representing The brand side specifically. And they were talking about just internally how much work has to be done to upskill their teams so that they can be using these. But a question was posed at the very end by one of our audience members, virtual, of course, but somebody posed the question about, like, are decisions being made based on AI right now? And it's kind of this, like, is that what's happening? And they pretty much were like, yeah, not really, not without massive amounts of validation. And, you know, the researcher is still highly gatekeeping the decision making on some level. It's not quite there yet, but they, of course, foresee a time when that may be happening, you know. So it's a really interesting time with lots you know, lots of use, you know, we've adopted technology and yet we're still kind of keeping it gated and humans are very much in control. It's pretty sweet.

[07:42] Lenny Murphy: That's interesting, I think that's an interesting nuance, right? And thinking about my own, my own use of AI, I expect yours as well, all the process efficiency. Right before this, I was merging two studies, utilizing my favorite perplexing show The and and asked if implications and you know, do all that good stuff, but they had nothing to do with what The ultimate output of it would be. It was still just a process efficiency. Yeah, do that there would be no action taken without The human review. So it's really, it's cool that that's where we are, actually. Because I mean, some things, you know, if you're doing an, I'll pick on that test, right? And it's like an A B test type of thing. And the AI is doing that, and it's an obvious score, this wins, etc. And executes The ad and The ad goes out there fully agentic, those types of low risk, you know, things. I think that makes perfect sense. That for The human doesn't necessarily need to be involved in that. The more strategic, the more impactful, the more risk there is, then I don't see that changing.

[09:06] Karen Lynch: Well, what's interesting is it won't change without full leadership buy-in. So one of The conversations that we had was about at The enterprise level, your ability to experiment with all of these AI platforms and all of these AI testing tools, et cetera, et cetera, is going to be decided or determined by kind of how much your powers that be are encouraging The experimentation, allowing you to budget for experimentation. So that's really where the corporate side researchers are kind of standing is they can only try your new AI-based platform if they have permission to experiment and budget to do so, because they are still required to meet their business objectives. You know, so they're using what they can as they can. So a culture of experimentation is going to help them help the industry advance even further.

[10:01] Lenny Murphy: But, but again, if we look at which conveyor was used as The, you know, it's just an example, you know, that experimentation is obviously happening. This whole event would be an example, so we may still be in the experimentation phase of the adoption cycle, but I don't think anybody needs to take that as, but it's still marginal. There's no data that indicates that. We're all in.

[10:34] Karen Lynch: For us, data is indicating this was our most popular AI event, which is interesting to me because we are anyway, more registrants for this event.

[10:43] Lenny Murphy: We didn't think we could pull it off three years later. We're like, oh, we're going to bring it up.

[10:48] Karen Lynch: Because we're talking about AI at other in-person events, we didn't know that the hyper-focus on AI would still be of interest. Because a lot of people tell us a little too much AI at some of the other industry events. So would we be able to engage people for two days of AI content? And sure enough, the people that showed up, now here's what I think is telling. The people that showed up for this, and there were people that were tuned in all day for The last two days. There's people that are looking. I've already seen people like catching up today, that kind of thing. Um, but they are, they are really looking at the range of, okay, here's how a brand side user, researcher is using it. Okay, here's a new tool I've not heard of before. Oh, great. Here's a prompt from Monica at Mondelēz. You know, I mean, that was a very powerful prompting workshop because she's showing how she's using it internally at Mondelēz, of course, using their own internal DPT or LLM, but still it was a very, it was The people that showed up this year were like, okay, these people are, they're broad-minded. They're really The IAX audience, right? They're growth-minded. They want to know what's coming. They want to know what's here and they want to know what's coming next. So really in tune with our audience. So it was exciting. And, you know, I think The bottom line is leaders have to lead The change and lead their teams through change and disruption. And then we all have to recognize that right now. And for the foreseeable future, it's that combination of AI and humans that are making magic together.

[12:20] Lenny Murphy: And for our listeners, I mean, here's a little glimpse into how The sausage is made. I mean, we practice what we preach, right? We are a media company in The Insights organization. Karen and I are researchers. Then, we use all of these things as data points. So, and, and we're building those better capabilities internally as well. So this, I mean, events aren't focus groups, but by a stretch of imagination, but yet it is massive amounts of information that come through our events that we try and funnel through and, and alchemize and share with you guys, because The level of insight there, plus other data sets, you know, The news, etc, etc, this whole point of The show. So if we get a little if you're like, quit talking about this event, understand, we're synthesizing in real time, key insights that came out of it to share with The industry, right? Because it is another angle of information that we're cross checking and validating, you know, against other things, too. So we can give you our best sense on where things are.

[13:33] Karen Lynch: Right, right. And if you are a supplier in The industry, you really, you know, if you're listening between The lines also, when I say something like, you know, human plus AI or creative thinking or critical thinking plus artificial intelligence, like, listen to what I'm saying about what's popping because that will help inform your content marketing down The road, right?

[14:00] Lenny Murphy: And your business strategy. This week, we didn't, we, it was no, we didn't put it as a link. So there was no reason to, but if you go on LinkedIn, Qualtrics are still talking about their deal with press, with press getting forced to, and it very explicitly in calling out, this was data plus service. And they recognize that. So you look at this, The big ass $6 billion deal from a tech company.

[14:27] Karen Lynch: There's The technology and The data and whatever they do The data AI right but still that service component and I think that that's it we all should take some comfort in that in our industry which is fundamentally a service industry transforming into a technology industry that is that synthesis that is that drives value yeah can I just say one other thing for those of you who are listening who are who are always curious about like our calls for speakers I do a lot of like checking calls for Speaker submissions that I've accepted or that my team has accepted compared to, you know, like kind of what else is out there. And I really look at The two and I think about that each time, especially before, for instance, Europe is still out there. I'm about to make those decisions for Europe. We just made them for North America. And when I tell you things like The Will Leach and Seth Mintz from Perigo, you know, collaboration, talking about how Perigo is using behavioral AI personas to help with their research, and then Human8 with L'Oreal for this particular event talking about how, you know, what some of their measurable outcomes are for The very cool new technology that Human8 is providing. Like, when I see that kind of collaboration, it will always win over something that's like, hey, look how great my platform is. You know? Like, I know your platforms are all great, and they're all your babies, especially if you're new, and like, there are You know, your babies are all really cute and really great and your AI features are great and table stakes for The record. But when you show that authentic collaboration, that's what our audience wants. That's what drives attendance.

[16:02] Lenny Murphy: Yeah, absolutely. Case studies, because we're all looking to figure out what's fit for purpose. Although I would say, if you have something that is incredibly revolutionary and like, holy crap.

[16:16] Karen Lynch: Try to share it in such a way that it's going to show how you're going to change business outcomes or influence business outcomes, though, like, absolutely, just can't be a shiny new tool for a shiny new tool sake, it has to drive decision making and outcomes like it has to anyway. Yeah, it's 15 minutes in Lenny, we have so much to cover.

[16:34] Lenny Murphy: There was a lot left on the cutting room floor this week. But I think all that was really helpful, great context, because everything else we're gonna talk about, really just see, here's The other data points that validate The themes.

[16:46] Karen Lynch: All right, share about data quality co-op, because I know you track that much more than I do. But they've expanded their capabilities. What do you know?

[16:54] Lenny Murphy: Yeah, they just keep, you know, data quality co-op is one of those platforms that they're, you know, trying to be kind of a clearinghouse of data quality across different buyers and suppliers. And they're just expanding their capabilities to drive value and understanding where things are from a data quality standpoint. I mean, they're, they're. The real thing really is around transparency, supply chain transparency. So it's almost a third party clearing house. I already said that, but I can't think of a better word. So to do that. Such a great attempt. I did a CEO series this week with Repdata. That'll be out soon too. Again, another approach, multiple companies trying to tackle this issue of giving us your data so we can identify the trends and stay on top of, you know, new emerging challenges and do equality co-ops doing that, other people are doing it, hats off. And I'm attempting to get a handle on this.

[18:06] Karen Lynch: Yeah, pretty cool. Good stuff.

[18:12] Lenny Murphy: So this was interesting. Go ahead.

[18:15] Karen Lynch: This is interesting. All right. So before I get to The story, though, let me just say, we did have conversation again at The Harris Full Town Hall at The end of The event yesterday. We were talking about AI powered qual and there is a difference for everybody listening between like AI moderation and AI and AI analysis of qualitative open ends. So like just kind of keeping all of that in mind. But AI qualitative research is popping right now. So it was a great coincidence that we saw this particular story. So Disqus was recognized by OpenAI for surpassing 10 billion processed tokens. Now, that's sort of like, what does that even mean? But the reality is OpenAI is recognizing providers or companies that are getting either, I think at their event it was 10 billion, billion, and 1 trillion token milestones. So they've surpassed 10 billion token milestones. That all has to do with usage, right, within The open AI ecosystem. I think it's a big deal. They're the only market research provider recognized on this particular wave with open AI. What's your take?

[19:32] Lenny Murphy: So I don't want to take anything away from that, from Discuss.io. I had never thought of it. Token utilization is a PR thing? And I get it. I totally get it. I did check with other AI-first companies and they were like, we do that in like a month. So there's something, it's interesting. There's The PR component of that, which is valid and appropriate. And hats off to Jim and The whole team at Disqus for doing this.

[20:08] Karen Lynch: I'm not- I mean, they're talking about the engineers within their ecosystem, which I'm like, all right, shout out. Yes, absolutely.

[20:17] Lenny Murphy: It's not taking anything away from that. But I'm also not sure that that is the accomplishment because there are companies like AI first that are like, we would have never even thought about highlighting that because we do that in a week or whatever, you know, The, and it just goes, it's a different metric. It's a different metric to success to understand where the company is on the AI transformation path. So that's, that's important. Adoption discusses a good sized company bearing for a long time transforming, you know, so obviously they're all in and they're getting client success and utilizing that all those things play into it. But it's almost like reporting how much power you use, right, kind of thing. I mean, it's just weird.

[21:13] Karen Lynch: Here's what they're saying though, if you look at The press release, they're talking about a thousand users a day. And I think there is a correlation there that is very interesting to think about. I certainly was like, you know, users a day is pretty, that seems like a lot. So I think what's going to be interesting is to see if we hear more of these, more stories about this. And to your point, as far as a metric goes, it's one way of, again, gauging and sharing with the world how many people are using your tool every day or your integration every day.

[21:56] Lenny Murphy: Well, that path. Yeah, I'm still trying to wrap my head around it. But it's like, okay, this is like traffic, right? We look at traffic and, you know, yeah, so it's, it's a new metric that I think we just got to wrap our heads around on. Yeah, where it indicates outside the obvious of growth and adoption.

[22:15] Karen Lynch: And just shows where we're heading to. I think that's another thing is it's kind of pointing to a world when these are all integrated in and there's an infrastructure in place. That allows us to track those metrics or somebody to track all those metrics too. It's just one of those things that makes you go, hmm, that's interesting.

[22:34] Lenny Murphy: Well, and the whole token economy, that's a whole other piece of things that I don't have a firm hand on, but anyway.

[22:42] Karen Lynch: Well, we do have a firm hand on partnerships, so let's get into some of these. Think about these partnerships, just when you think it's safe to talk about them all. Some of them are like, this one I have to think about a little bit, like, why is this important? Should we share this? You know, when we look at all these, and it's like, is this one? Should we share this? And it's like, I think we should, because, you know, Lenny and I are here to kind of point the way to paying attention to what are some of the signals in this crowded world. So outset and dovetail. You want to take that one?

[23:11] Lenny Murphy: Yeah, you know, it's all about research workflows, and, and utility. So I've got a concept I've been playing with all that this idea of it's now The three U's. It's uniqueness, usability, and utility for data. This is that dovetail, that partnership that it's unlocking more outside of outset into dovetail for unified analysis and sharing. We're just going to see more and more of these ecosystem relationships that are expanding the utility of insights across organizations.

[23:45] Karen Lynch: It's just so interesting because when we talk about customer or when we talk about knowledge hubs and, you know, we talk about all of these things that we've been talking about for a while and now all of a sudden it's almost like, you know, now it's like The ability to access all that information in a new way is really what makes it exciting and it's just loaded with opportunities. So good luck in their partnership.

[24:06] Lenny Murphy: Yep, yep. The similar one, Signal AI and Propel AI partnership to pioneer a new era of AI native computing So again, same thing, utility, now it's more like utilization, right? The activation. So more things happening at scale to connect The dots and make it just easier to move from insights to activation.

[24:36] Karen Lynch: Yeah, and this is specific to the media, right? So it's on some level like media intelligence or kind of metrics. You know, whatever the analytics based on media performance and then give some intelligent direction towards it, you know, to be able to shape a narrative based on what you're learning, but do it pretty quickly. I think so. Um, yeah, cool that there's two, it's an AI partnership with another AI partnership. That's what I think, you know, popped out at me there. These next three, I think these are, these are The shape of things to come before you The next one because I'll let you talk about it because I know. Anyway, I was looking on chat GPT for something this weekend and I was being dialed up for some specific product recommendations and I was like, well, that's interesting. Anyway, let's talk about this because what I was seeing was already in action and it was pretty cool.

[25:35] Lenny Murphy: Walmart with open AI. Direct product purchase through chat GPT. And you mentioned this last week that, well, I think you had said, well, we were talking about this, The commerce connection, right? And the ability to make purchases directly through your recommendations. Obviously Walmart, largest retailer in The world and probably second largest online retailer. Uh, and trying to, to become number one. Uh, this is obviously an online retail, uh, component of that. So just make that seamless. So what, what does that do? Uh, you know, we're heading into the holidays.

[26:23] Karen Lynch: There's a phrase in some of the stories around this that I looked at. Cause of course I went down the rabbit hole of, um, Nancy, I'll tell you in a minute. Um, there's, there's a phrase in there called conversational commerce. And I think that's going to be something to be listening for here. I am having a PT conversation. I'm overtly sharing, Nancy, what I was doing was one of my nieces is having a baby. I'm like, shoot, I need to start on a baby blanket for her. I haven't done that yet. She's halfway through her pregnancy. And then I'm like, what I need is like, I need The perfect yarn, right, for The right baby blanket because I don't know that I have time to like get to The yarns. Anyway, super complicated. And it just kept serving me up like one Etsy thing after another. And I was like, why does it keep giving me The Etsy stuff? And then all of a sudden, what was happening is I'm having conversational commerce because Etsy was The original partnership and now Walmart is on top of that. I wasn't getting Walmart kits. I was getting these Etsy kits. So anyway, but what that's indicative of is it's indicative of a time when I can like be chatting The chat GPT. And I think last week it was in The, there was an integrated, there's an integration with Spotify. That's somebody I know is playing with. It's like talking about creating a playlist and Spotify, but, but talk through the creation of your playlist. Like, don't just, don't just say, you know, you know, what's a playlist, but talk through it so that it's your thinking and your wisdom and your preferences being communicated to The AI that's going to take it and run with it. That's pretty cool.

[27:49] Lenny Murphy: And we're just going to see more and more of that.

[27:52] Karen Lynch: And so I think we shared last week that I have said, set up the integration, not to take away from Walmart, cause I do want to get back to that because there's so many implications there. But, um, I have set up The integration with chat GPT with, because we've listed our house and The agent is working and it's basically going to let me know if new listings come up that are kind of in our sweet spot of everything that I want in The house. Every like three times a day, I get a, there's nothing for you to look at. Don't worry. There's nothing for you to look at. And the second it says to me, if there's something for you to look at, I might then say to Tim, like, Hey Tim, we've got something. So I have an agent at work. Don't tell my buying agent, but like there's an agent at work in Chachi PT. To find The properties that pop up on The market that we may be interested in if we end up selling our house.

[28:38] Lenny Murphy: Something just occurred to me, but as we get to the next one, I understand. The battle here, as it always is, is between consumer and enterprise. OpenAI clearly is trying to become The dominant consumer platform. So, uh, and, and I would say right this minute, they are, um, uh, now The next one, right next story was Anthropic and Salesforce. And Anthropic is making The play for The, for enterprise. Um, that's why they've been focusing on safety here a lot, you know, um, they're taking a different approach, but they're, you know, who they're deeply integrated with Amazon. They're with AWS, right. With Amazon. With Amazon Web Services, with Microsoft, now they're Gage Salesforce. So we're seeing this world start to take shape of who's going for The big players, right? To be The next big platform. And we need to pay attention to that. And particularly, I would say, I mean, there's stuff in The ether, where we will see these companies making more overt moves that affect The market research industry.

[30:12] Karen Lynch: So let's certainly affect the way that the market research industry does business because of the slack integration with Salesforce also or vice versa, which I think also hit this week. And I was looking at that and I was thinking, well, that's interesting. As a company that has both Salesforce and Slack, I was thinking about that integration in particular. And I'm like, these enterprise level platforms can change the way business is conducted internally or how conversations are happening or how data gets gleaned. It could all be very interesting.

[30:50] Lenny Murphy: It is. And then The, uh, uh, Accenture acquired, uh, Deco, um, which was really about their enterprise analytics service.

[31:01] Karen Lynch: Um, especially with Palantir, which is particularly interesting when, well, I had, I had to, so, so this was, it's funny because this was shared on research live. So of course I tune into it. I might've just blazed right over this thinking like this is way out of our wheelhouse, even though Accenture every now and then dabbles into insights or whatever. I pulled up this article and I'm like, well, I personally have to look up what a Palantir platform is because I didn't know. So for those of you listening, Palantir platforms, enterprise software integrating disparate data sources. So we've talked about that, right? And actually talked about it with Sarah Fometa at our town hall yesterday. The ability of AI to connect these disparate data sources and data points and connect dots in a way that humans cannot, like our human minds just can't do this.

[31:45] Lenny Murphy: Doing this for like 20 years.

[31:47] Karen Lynch: So, that's pretty cool. But, but I think it's, it's this big, and we talked about some of the data points that morning consultants are putting together right now. Like, so, um, Accenture, it's just another play where they're like, we're a data master. You know, we are going to be able to, to then glean insights from your data in our consulting services.

[32:12] Lenny Murphy: The interesting stuff. All right, we got two more that I do think are, we should mention back to our industry. Pulsar, they're Insights Agents, right? And I've always just liked Pulsar's approach. They've built, they started a social listening platform and they've expanded to build modules that address different business issues and types of data. So I just liked the approach and it was very interesting. Uh, The, uh, now they're, they're corporate agents. So surprise.

[32:48] Karen Lynch: Well, what's cool about this, click on this one, you know, friends, this, these agents, because it goes into their, they're calling it teammates.

[32:55] Lenny Murphy: It was pretty cool. It's pretty cool.

[32:57] Karen Lynch: And they've named them. They've got Sentinels that are designed for alerting. They've got oracles who are trained for prediction custodians for compliance analysts for research. So their team's pretty, it's pretty cool. I'm like, you know what? That'd be a great team to build. I know that, you know, shout out to my husband, Tim Lynch, God bless him, but he's deep in this space. And he had created a team for his own purposes, like in his own GPT world. And, and he shared on LinkedIn, you know, weeks ago, he's like, yeah, I've got a team of agents now working for me. And, you know, at the time thinking, okay, that's great, you know, like, good for you. But I love the idea that they're like, they're marketing this, like these capabilities are going to be available for your team, too.

[33:39] Lenny Murphy: And they're cute.

[33:40] Karen Lynch: And they're cute. That's cute. There was something cute about it. Anyway, no shade on being cute. It's all about usability.

[33:48] Lenny Murphy: All about usability and engagement. So our friends at Displayer, they introduced a new data preparation agent. Automated data cleaning and structuring for faster survey analysis. We use Displayer with our analysis grid data. Have been great friends to us for a very long time. Data cleaning is a pain in the ass, right?

[34:15] Karen Lynch: I mean, it just is. They say that this is like, you know, click the button and we'll, within a minute, tell you everything that needs to be done to clean your data.

[34:25] Lenny Murphy: We'll see. I'm sure Nelson is probably looking forward to testing this on The next round of Grid. I wish we'd had this a month ago.

[34:33] Karen Lynch: Maybe we Yeah, my pick is this round, but next round for sure. So yeah, it'd be, when I was looking at it, I was thinking to myself as somebody who has not, you know, cleaned data to that extent at all, obviously not living in that world per se, but thinking about how once you see The suggestions, discerning which things to say, yes, do this, don't do this, yes, do that, don't do this, like that all becomes this whole other world Data cleaning is a very unique skill, and not everybody is capable of making The decisions that this player is going to tell it to make. So human-AI collaboration, mandatory.

[35:15] Lenny Murphy: And go back to The very beginning, from a data quality standpoint, all of The platform, there's different points in dealing with this, right? There's The panel company, The supplier company, Then there's The intermediaries like, you know, Repdata and Data Quality Co-op, et cetera, et cetera, that, you know, are catching what's coming through The pipes, right, to, you know, filter it even more. And then there's still this human element that is necessary that doesn't necessarily mean that it's fraud per se. People just screw up sometimes. So, clean the data. We get it in like The GRIP50 all the time of somebody that says oh, yeah Microsoft is a supplier and you got to stop and think about it. Are they? Well, what's the context? You know, so there's a lot of massaging that just needs to happen Yeah, so we make sure that it makes sense contextually and here's another example of that. So yeah, you were right. Go ahead.

[36:22] Karen Lynch: No, no cut. What were you gonna say?

[36:23] Lenny Murphy: No, so you were right.

[36:24] Karen Lynch: We were not gonna get to all the stories I literally said in The brief, yeah, we're not gonna get, yeah, let's just be very clear. But let's start with The CB Insights release because it's a download. Everybody can download, check out. They published an AI agent Bible. So kind of following The two stories about agentic work, CB Insights has this AI agent Bible. So I haven't necessarily had a chance to look at it, but if you are interested in what we're doing with agents and why that's coming next and how more and more companies are working on And CBN Sites, that if you don't follow CBN Sites, you should.

[36:59] Lenny Murphy: I mean, they are The, yeah, they are on it with all of these things. They create fantastic, well-researched content on this topic. I do want to call out this MIT thing. Yeah, please.

[37:09] Karen Lynch: This one I was like, yeah, you know, so they have a new model. Anyway, it's all too deep for me to read at a glance, so you can talk about what you read.

[37:27] Lenny Murphy: It's a self-adapting AI. It's a new model that rewrites its own code to learn autonomously. There's this like, oh no, should we be doing this? There's always that fear.

[37:47] Karen Lynch: I mean, you want, you want it to read. If, if you're still interacting with it, have you ever done a prompt or something, you know, whether it's in perplexity or whatever, um, where you're like, that's not quite right. And it says, yes, let me take a look at that. And then it reworks it. So, what's interesting to me is it's doing the thinking and then saying, wait, that's not quite right. And then redo it again.

[38:11] Lenny Murphy: So there's somebody who, uh, somebody shared a story with me. Like The, that they're using, uh, I've got a platform, um, uh, they called, um, and they're asking for a fairly simple functioning. I can't do that, but you know what? I can do this. I can work around my own restrictions. Get you The result that you want in bypassing its restrictions. And I think that's one thing. My understanding of this is this is, The code says, oh, there's a restriction. Screw that. I'm going to remove the restriction. I'm going to rewrite my code to accomplish the task, regardless of what was originally intended.

[38:55] Karen Lynch: Autonomous learning, I think, is really interesting to think about. It's going to learn on its own and make changes on its own.

[39:04] Lenny Murphy: And here, it says it's more efficient, better factual recall, et cetera, et cetera, outperforms GPT 4.1. The whole point of that is we're, do not assume that we're just making marginal improvements. We are not, you know, that every week there is something new. It may not be fully scaled and adapted yet, but the level of fundamental innovation and rethinking that is occurring with this technology shows zero signs of slowing down and all that we just see, we just see the incremental changes. Oh, there's a new version of whatever we're using. But behind the scenes, there's holy crap changes happening.

[39:54] Karen Lynch: I know, I know. And the university level is where they're happening, so good stuff. I say we wrap it there, Lenny. We're in, and what a show.

[40:10] Lenny Murphy: Next week.

[40:14] Karen Lynch: Big stuff. Lots going on. I hope everybody has a great weekend. I hope everybody has a great weekend. I hope you have a great weekend. I need to have another great weekend. More open houses. Didn't get The offers The first weekend because we had The stupid storm. So the realtors want to try again. So I'm like, all right, let's try again and see what happens this week.

[40:38] Lenny Murphy: Yeah, it looks like a lovely weekend.

[40:40] Karen Lynch: Yes Okay.

[40:41] Lenny Murphy: All right, everybody. We're setting Karen to offer vibes.

[40:46] Karen Lynch: So all right, really it's cool if we don't. Anyway, we love our house. It's all we'll see what happens Okay, everybody.

[40:56] Lenny Murphy: Take care. Have a great weekend.

[40:58] Karen Lynch: Take care. Have a good one. We'll see you next week. I Don't know what this wave was

Links from the episode:

Data Quality Co-op expands platform capabilities 

Discuss was recognized by OpenAI 

Outset and Dovetail partner to streamline research workflows 

Signal AI and Propel AI announce strategic partnership 

Walmart partners with OpenAI 

Anthropic and Salesforce expand partnership 

Accenture acquires Decho 

Pulsar launches Insights Agents 

Displayr introduces the Data Preparation Agent 

CB Insights releases the “AI Agent Bible” 

MIT unveils SEAL 

Most consumer goods decision-makers use or plan to use GenAI 

Study finds Gen Z overrepresented in ads