[00:00:00] Beth Schultz: Hello listeners, and thanks for tuning in to this edition of our metrosight podcast. I'm Beth Schultz, I'm the Vice President of Research and Principal Analyst at metrogy. And with me here today is Matt Craig. Matt is our Senior Director of Product Development and he also spearheads our quarterly AI Sentiment Index research. And that's why he's here with me today so we can talk about the latest findings. Welcome, Matt.
[00:00:24] Matt Craig: Hey, Pat. Thanks.
[00:00:26] Beth Schultz: Okay, so before we get to 2Q findings, let's kind of give our listeners a quick review of metrogy's methodology for the AI Sentiment Index.
You know, this is due to us, really. It's only the second quarter in, so let's remind everybody. Who are we surveying? What are we asking them? What's the goal?
[00:00:44] Matt Craig: Absolutely. And as you mentioned, this is only our second quarter of doing this. But so what we're doing is every quarter we're going out and surveying a third thousand US respondents demographically representative of the US population. So we're trying to get a snapshot of what, you know, regular people are thinking about AI, you know, what, what are their feelings towards it. We are asking them both about their current feelings and where they expect it to be in one to two years in future sentiment around the personal impact of AI that's had in their lives, what the societal impact has been in their opinion and the economic impact. So these are, you know, ranging from extremely positive to extremely negative in, in each of these. And we're taking those, those responses and using a pretty standard diffusion index methodology for that. We've tweaked it a little bit to make sure we're really capturing those who are at the extremes a little bit more, you know, that are very positive or very negative, giving them a little extra, extra weight that, just a general positive or negative. So doing that, putting it on regular 100 point index that we have here. So 100 is your baseline and we'll be tracking this every quarter going forward and seeing how, how sentiment changes from, from, you know, this 100 baseline in Q1, positive or negative.
[00:02:10] Beth Schultz: Okay. And we have, so we have the overall AI Sentiment Index, the current, correct me if I'm wrong here, current AI Sentiment Index and Future Expectations Index. Right, correct.
[00:02:21] Matt Craig: So those are, all three of them are there. So we have the current Sentiment and Future Expectations Index. Those are what I kind of deem as sub indexes. So those are looking specifically at what they're thinking now and what they're thinking for the one to two year timeframe. And then we kind of roll that up into an overall index by, you know, averaging them out. So then we'll have the overall.
[00:02:46] Beth Schultz: Okay, so as you said Q1, we set the baseline at 100. So let's talk about where we are with Q2 in those index scores.
[00:02:54] Matt Craig: Sure. If we're looking strictly at the overall, we saw a 1.4% bump, so we're up to 100.4. It's really easy at this point to calculate those quarter over quarter changes moving from 100. And while that's a decent shift forward in terms of positivity around AI, we're actually seeing the biggest bump, bump of that in the current sentiment index.
So that actually went up 2.2% to 102.2, whereas the future expectations saw a very, very slight shift up to 100.6. So I think the, you know, kind of incredibly rapid advancement that we've had, not just in the AI models themselves, but how they're being packaged and turned into usable applications and the fact that more and more people are using AI even now this quarter compared to last quarter, we're seeing that current sentiment kind of jump up a little bit more. You know, so we've, we've seen a 5% increase in those who are saying they're regularly, regularly using AI and an 11 point or 11% bump in those who are saying they're at least occasionally using it. So I think people are now, more people are seeing, hey, we have, we're seeing positive impacts now. We're seeing how helpful it can be. The future, it's a slightly different story. While not as, you know, while we're not seeing a drop in sentiment towards it, we're still seeing quite a bit uncertainty.
I think there's just so many possibilities of what could be an impact of AI, you know, both their personal, you know, benefits as well as societal and economic. You know, with that uncertainty, we're seeing a lot of things around job displacement concerns, misinformation concerns, data transparency, privacy issues. I think a lot of those are on the minds of a lot of people and there's no real clear cut answer to is, you know, is that a real concern or not? So people are taking a little wary approach to what it's going to be in a couple of years.
[00:04:56] Beth Schultz: You know, we're only 2/4 into this index. Right. So we have a fairly limited data set.
And you know, as you talked, you talked about, we've seen, you know, the, you know, what it was it a 1.4 point change in the baseline. Do we really think about can you think about this as a trend? You know, these indicative of a trend at this point, or is it too soon to start talking in those sorts of terms?
[00:05:21] Matt Craig: I think we're a little early to say trend quite yet. Just because, and even going out, you know, another quarter, it still might be too early, more just because of how rapid the advancements have been. You know, the entire landscape can change within a month. You know, we could see a completely different landscape. If you OpenAI puts out a new model that blows everybody away, that's going to change how all of these tools and applications are going to be doing, or somebody comes out with a new way of utilizing the existing tools. You know, I would have been really curious to see if we had started this towards the end of last year, what the huge, you know, popularity, like rapid growth of openclaw would have been on this, because that was kind of a whole new paradigm of how people started using it. And really the personal agentic tool versus just the general agentic tool that is out there a lot. And I think those huge bumps is going to make it kind of wild for a little bit until we start seeing a slowdown in the progression of what these things can do.
[00:06:31] Beth Schultz: You mentioned openclaw. Maybe explain that just in case. Some of our listeners aren't familiar with openclaw.
[00:06:37] Matt Craig: So openclaw or Hermes is another popular one. They are focused more on actually using the AI models that you have now to go out and do things on your behalf. And they do it in a really clever way where the models actually will create memories and profiles for themselves of things that they need to remember in order to better future help you. So if you're working on something, whereas if you're going, you know, using cloud code right now to write, write some code, you know, there's going to be some things that you need to change, you know, every single time that it comes up, because, you know, it doesn't do it the way that you like. You know, are you, you know, capitalization or formatting, you know, even simple things like that of your preferences? Well, if using something like an Open Claw or Hermes, you could tell it, hey, don't want to do this, you know, in this way. I want it formatted in this way. It'll actually take that and store it in its memory and actually kind of build up a profile of what you like and what you want to do. And if it's something that you're doing, often it'll actually kind of cache that memory.
And so it'll start doing it without you even asking to do it. You know, if it's something that you come in every day and you run these certain, you know, routines every morning, well, in a few weeks it'll start saying, hey, do you just want us to do this for you without you asking us to do it? And it'll just be there when you,
[00:08:01] Beth Schultz: when you step in that ever increasing efficiency and productivity gains. Right. Okay. Yes, thanks. I think listeners who weren't familiar with OpenClaw will, will be familiar with it now. Okay, so Matt, as you talked about in the intro, we look across three dimensions.
Personal, social, economic.
So for, for Q2, what were some of the more interesting takeaways that you kind of uncovered as you did your analysis across these dimensions?
[00:08:31] Matt Craig: In, in personal, I, you know, we saw a little bit of a bump more towards positive sentiment in general. I think that goes pretty hand in hand with, as I mentioned, like more people actually using it. One of the things that we have found out is a lot of the people that say that, you know, AI is not helpful for any of these, you know, for them or, you know, it hasn't been positive. Most of them haven't actually used it, you know, saying that they don't have a need for it. But we're seeing as more people are using it more regularly, more people are getting a positive, you know, impact towards it. So I think those are going to go hand in hand. The societal and economic is a slightly different story because a lot of that is less on personal experience and more on what they're see, what they see or are told.
You know, I think media has a lot of impact over what the feelings people will be. If you're constantly bombarded with negative stories about something, you know, even if you are skeptical when that repetitive nature of it is still going to kind of subliminally leak in about certain things and putting more questions in your head. So we've seen both of those aspects veering a little more towards the negative. But one of the really interesting things that we saw on the economic side this, this quarter compared to last quarter is we're starting to see some of the, the people who said that they had kind of a, a negative impact or a negative feeling of what the economy is going to be in the next couple years, the impacts of AI and so just somewhat negative. So they're, they're not the, the doomers, but they're not necessarily the people who are thinking it's going to be a good thing.
We actually saw that shrink quite a bit. But at the same time we saw the, those that say extremely negative go up and also those who said that are positive go up. So you're seeing this kind of split in these people of how they're, they're shifting their feelings. So you're getting, some people getting more negative, some people getting more positive.
[00:10:33] Beth Schultz: And it'd be so interesting to see how this plays out, you know, quarter over quarter. Right. Because as you said, it's a little bit too early to say. Yeah, say the trend. Okay. That said, you know, you know, the Q1 data, you know, the Q2 data. So were there any surprises that, that came up when you were doing the, the Q2 analysis, things you really didn't expect to see?
[00:10:51] Matt Craig: Not necessarily a surprise, but more eye opening, I think. So we did add a couple extra questions this quarter that we did not in the first quarter, you know, after we, you know, went through the first quarter data, there's, there was a couple questions that came up that were like, we, we really want to know a little bit more about this.
And the biggest kind of eye opening is how the general public is interacting with AI for the most part. You know, a lot of the conversations that we've had in the past, you know, have been with a lot of the producers of the technology, you know, those who are creating it or those who are implementing it within businesses. But when we go out and actually talk to the people who are using it, you know, there's kind of a disconnect in, you know, where they're focusing. So more than half of the use of any type of AI is more just general purpose tools. It's using ChatGPT, it's using Gemini, it's using Claude. You know, they're not necessarily using it for the kind of agentic features that are constantly being talked about, you know, on, on our side. And those who are kind of, you know, waist deep in what's happening in the industry, you know, they're, they're using it for general research, for search engine replacements, you know, some more of like those types of convenience. They haven't, they're not, most have not taken that extra step towards using it as essentially an assistant or using it to do things on their behalf. You know, they're, they're still, you know, a little, a little bit slower on that upkeep. Not that that's not a good use of it. It definitely is, but it can do so much more. So that, that is something I want to keep my eye on and see how that changes at all, especially with all of the, the talk happening about the real agentic aspects of AI and what it can do.
[00:12:42] Beth Schultz: Okay, one last question for you, Matt. And it may not be all, let's say it may not be a fair question, but I'm going to ask it anyways. Oh, gosh, no, no, it's, it's nothing scary. It just, you know, as you sort of keep your pulse on what's going on in the world with AI and you know, having done two quarters in. When you think about Q3, do you have any expectations of what you're going to see, how the survey is going to pan out for Q3 and maybe if you don't have any expectations necessarily, what are the kind of the key things you're watching for?
[00:13:15] Matt Craig: Yeah, so I would say on, from a personal aspect of it, I don't think we'll see a huge change for personal impacts. We may see a slight increase or a slight decrease, but a lot of times within the margin of error, you know, on that decrease side of it, just as we're getting the advancements and more people are using it. So I don't expect a huge, you know, shift in that and I definitely don't see it necessarily going negative on the personal side. On the society side. I'm very, I'm very curious to see what is going to happen. I would. Just three days ago we had the US Government deciding to ban Mythos from Anthropic. The Mythos model. That got a lot of, of news, you know, headlines out there, especially from major media publications, you know, and that can, the, the headlines of that can sound very scary. It's like it banned, it was banned for being too dangerous. Now if you read into to it, it's not necessarily as scary sounding very scary as it sounds, but that, you know, not, not everybody reads past the headlines. So I'm really curious to see what type of impact that's going to have on the society factor of things. Because if we're putting into people's minds that hey, this is a dangerous tool, it's not just, you know, making me nervous, but it's actively dangerous on the economic side. I think the, the biggest thing that may have an impact is what the, the US Stock market does. To be completely honest, we've, we've had about a year now of a lot of layoffs with the stated, you know, reasoning being for AI and it hasn't necessarily impacted it significantly. So I don't think any more of those will impact sentiment, you know, too much negative. But if all of a sudden the, the stock market goes down 15% in the next three months because, you know, because of AI. You know, we've seen the huge rise based on the back of AI. Now if we see a decline based on that as well, I think you're going to start seeing more people shift negative on that.
I don't know if we're, you know, I wish I knew that I could predict exactly what the valuations are going to be and what investors think. I wish I could predict that, but.
Right.
That would make life so much simpler. But I think that's going to be a leading indicator of what the sentiment is going to be just because that has such a huge impact on a lot of the people that are using AI.
[00:16:00] Beth Schultz: Okay, so that is all great stuff. We'll have to track how your, your, your thinking comes to fruition with our Q3 survey and then moving on from there. This is really exciting stuff. It's great to, it's great to talk to you about this and looking forward to your continued work on the index. So thank you so much, Matt and listeners. As a reminder, you can find our AI Sentiment Index reports on the Metro G website at www.metrogy.com and as always, we're happy to hear from you. So if you have any question Matt about the index, feel free to reach out to him. You can just shoot him an
[email protected] or you can email me at beth.bethetrogy.com or you can just use the contact button on the Metrogy website. And that is all for now. On behalf of Matt and the rest of the Metro G team, goodbye till next time and take care, everybody.
[00:17:02] Matt Craig: Thanks so much, Beth.
