Products

EMARKETER delivers leading-edge research to clients in a variety of forms, including full-length reports and data visualizations to equip you with actionable takeaways for better business decisions.
Reports
In-depth analysis, benchmarks and shorter spotlights on digital trends.
Learn More
Forecasts
Interactive projections with 10k+ metrics on market trends, & consumer behavior.
Learn More
Charts
Proprietary data and over 3,000 third-party sources about the most important topics.
Learn More
Industry KPIs
Industry benchmarks for the most important KPIs in digital marketing, advertising, retail and ecommerce.
Learn More
Briefings
Client-only email newsletters with analysis and takeaways from the daily news.
Learn More
Analyst Access Program
Exclusive time with the thought leaders who craft our research.
Learn More

About EMARKETER

Our goal at EMARKETER is to unlock digital opportunities for our clients with the world’s most trusted forecasts, analysis, and benchmarks. Spanning five core coverage areas and dozens of industries, our research on digital transformation is exhaustive.
Our Story
Learn more about our mission and how EMARKETER came to be.
Learn More
Methodology
Rigorous proprietary data vetting strips biases and produces superior insights.
Learn More
Our People
Take a look into our corporate culture and view our open roles.
Join the Team
Contact Us
Speak to a member of our team to learn more about EMARKETER.
Contact Us
Newsroom
See our latest press releases, news articles or download our press kit.
Learn More
Advertising & Sponsorship Opportunities
Reach an engaged audience of decision-makers.
Learn More
Events
Browse our upcoming and past events, recent podcasts, and other featured resources.
Learn More
Podcasts
Tune in to eMarketer's daily, weekly, and monthly podcasts.
Learn More

The Daily: GenAI adoption, AI PCs, and will ChatGPT hurt publishers?

On today's podcast episode, we discuss how marketers have been adopting generative AI (genAI), what ordinary Americans want from it, and what the genAI frenzy could face in 2024. "In Other News," we talk about how much ChatGPT will hurt publishers now that it has real-time information and what we can expect AI PCs to do differently. Tune in to the discussion with our analysts Jacob Bourne and Gadjo Sevilla.

Subscribe to the “Behind the Numbers” podcast on Apple Podcasts, Spotify, Pandora, Stitcher, Podbean or wherever you listen to podcasts. Follow us on Instagram

Made possible by

MailChimp

Intuit Mailchimp is an email and marketing automations platform for growing businesses. We empower millions of customers around the world to start and grow their businesses with world-class marketing technology, award-winning customer support, and inspiring content. Mailchimp puts data-backed recommendations at the heart of your marketing, so you can find and engage customers across channels— automatically and with the power of AI.

Learn More

Episode Transcript:

Marcus Johnson:

Ever heard of a clustomer? It's the result of marketers grouping customers With different behaviors into one big mess. But with Mailchimp, you can use real-time behavior data to personalize emails for every customer based on their browsing history and buying behavior, turning your clustomers into customers. Intuit Mailchimp, the number one email marketing and automations brand. Visit mailchimp.com/personalize for more information. It's based on competitor brands' publicly available data on worldwide numbers of customers in 2021, 2022. Availability of features and functionality vary by plan, which are subject to change.

Jacob Bourne:

I'm not completely convinced that there's going to be this crisis for the sector come 2024. What this all really depends on is high-performance, efficient, low-power-consuming AI chips, and that's where a lot of investment is going into.

Marcus Johnson:

Hey, gang. It's Tuesday, October 17th. Jacob, Gadjo, and listeners, welcome to the Behind the Numbers Daily, an eMarketer podcast made possible by Intuit Mailchimp. I'm Marcus. Today, I'm joined by two folks, both of them right for our connectivity and tech briefing. One of them is a senior analyst based on the right-hand side of the country. In New York City, it's Gadjo Sevilla.

Gadjo Sevilla:

Hi, Marcus. Hi, Jacob. Happy to be back.

Marcus Johnson:

Hey, chap. Good to see you. The other one is on the opposite side of the country on the left-hand side. Based in California, it's one of our analysts. It's Jacob Bourne.

Jacob Bourne:

Hey, Marcus. Hey, Gadjo. Glad to be here.

Marcus Johnson:

Good to see you. Today's fact. What do you guys think are the most common passwords? If you had to guess what you think one of the most common passwords is, what would you say?

Jacob Bourne:

Password.

Marcus Johnson:

Password's second? Yes.

Jacob Bourne:

Yeah.

Marcus Johnson:

According to-

Gadjo Sevilla:

12345?

Jacob Bourne:

Birthdays.

Marcus Johnson:

12345 is fourth.

Jacob Bourne:

Names of significant others, family members, things....

Marcus Johnson:

No, it's not that-

Jacob Bourne:

[inaudible 00:02:03].

Marcus Johnson:

It's not that complex, unfortunately. You guys got it. So password is second. So this is according to a report from SafetyDetectives, and the most hacked password is password. That's the second most common password. What the hell are we doing? Why even have a password? Just let people in. Rather ironically, let me in is... That was 22nd on this list. Gadjo, you had 12345. That would be the fourth most common. The most common is 123456.

Jacob Bourne:

[inaudible 00:02:36].

Marcus Johnson:

I know. Hello. Didn't see that coming.

Jacob Bourne:

[inaudible 00:02:39].

Marcus Johnson:

The third most common is 123456789. It's just variations of that. What is happening? All the ones, that's the eighth most. Dragon. Hmm, I didn't see that coming. Superman. Okay. Yeah, they're not good, is the point here. It's not great. But Gadjo, you just wrote a piece about how passwords might be going away.

Gadjo Sevilla:

Yeah, that's right. Google's hoping, and I think it's an industry-wide push, to move to passkeys, which means you'll use devices as a form of authentication rather than a password.

Marcus Johnson:

Thank God, because this is a disaster.

Gadjo Sevilla:

That can't come soon enough.

Jacob Bourne:

There's clearly a need for it with those kind of passwords.

Marcus Johnson:

Yeah, yeah. IT will be horrified to learn most of my passwords are on this list. Anyway, today's real topic, generative AI adoption and what Americans want from it.

Today in the lead, we'll cover generative AI. How many folks have adopted it? We'll look at what regular American folks are expecting from artificial intelligence and what 2024 holds for generative AI. And then we move to In Other News. We're talking about ChatGPT now giving real-time information. What does that mean? And then PCs with AI inside of them. What are they going to look like?

We start, of course, with the lead, and we're talking about AI adoption for marketers. A recent Adweek article from Stephen Lepitak explains how marketers have been adopting AI, citing a recent survey from Gartner. And so it's three stats in here, which I'll start with to set the table.

Number one, using generative AI for the funnel. That's what folks have been up to. Half of marketing leaders said they are already using generative AI tools within their funnels. 43% aren't right now, but plan to. Only 6% said, "No, thank you."

Number two, content marketing was the number one function using generative AI, with 73% of folks saying that. 43% product marketing, 37% customer experience. And then the third stat here, half of folks were using it, generative AI, for content production, drafting social media, and also ad copy. So gents, what stands out to you when it comes to how marketers have been adopting gen... I'll just say gen AI. I can't keep saying generative the whole episode. Gen AI at this point.

Jacob Bourne:

Yeah.

Gadjo Sevilla:

I think for me, the key thing here is that we're going to see a lot more content that is AI generated. So recently, Adobe released a new tool, Firefly 2. So that competes with DALL-E and Midjourney. So this creates ads and marketing images on the fly, and they're using thousands of stock photos and assets as the training models for these.

And so we can expect as marketers and content creators continue to adopt this, that they will be leaning on this technology. But the question here though is, will authenticity be an issue, especially with regards to marketing and honest marketing? The images do look very professional. They look very polished. But the fact is, they're not real. That's not a real product or a real model at that point. So.

Jacob Bourne:

Yeah, I agree with Gadjo's take. It's certainly not a silver bullet for the marketing industry. At the same time, it makes a lot of sense that adoption has been bullish among marketers. I mean, just the efficiency and creativity that general AI can give to the sector is... Personalization, I think, is one of the biggest areas.

I mean, you have AI that can, on the one hand, analyze heaps of consumer data and then turn around and use that consumer data to craft personalized content that targets an individual consumer. So it's a marketing dream, in a way, if you think about it. Really, it's ads, emails, images. Any kind of content can be personalized to an individual.

But there's a pitfall to that too, though, because that personalization can also come across as creepy to consumers that don't want to be analyzed. They don't want to be tracked. And that's exactly what this does. So it's really... It's walking a tight rope for marketers to do this in a way that instills trust and doesn't make people feel very uncomfortable.

Marcus Johnson:

One of my questions was going to be yeah, who's going to be the first advertiser to face major backlash for using AI in ads? Because in August, customer experience consultancy UserTesting found that folks are suspicious about AI being used in targeted ads. 70% of people were concerned about their privacy and personal data being used. Nearly 60% felt anxious about it being used specifically by marketers. So I'm wondering if we're going to have some big, splashy headline case where there is an advertiser who has used AI and they get a lot of blowback because of it.

Jacob Bourne:

Yeah, I think so. I think it's just a matter of time to be seeing that, for sure.

Marcus Johnson:

Mm-hmm. So folks adopting generative AI, a lot of folks experimenting with it, whether that's experimenting and then rolling it out, just experimenting, planning to experiment. It seems like there are fewer, a small number of folks who haven't really done anything with it at all. When it comes to marketers, the main benefit folks expected to see from their generative AI deployment was improved speed to market, followed by being able to create more contents in the same amount of time. So that's what marketers were hoping to get out of their generative AI deployments. The main barriers were not having the expertise needed and unforeseen security threats.

Jacob Bourne:

And I would add just a reduced need for overhead.

Marcus Johnson:

Mm, to the benefits, yeah?

Jacob Bourne:

Cost cutting, essentially.

Marcus Johnson:

Mm-hmm, mm-hmm. So there are some concerns, though, about generative AI, and I'm wondering how skeptics' concerns are going to be alleviated. One of the main concerns is there's too much choice. There are currently over 360 generative AI companies according to CB Insights. Do we see that continuing to go up in the next six months or to a year, or is that on a very steady or a very expedited downward trajectory at this point?

Gadjo Sevilla:

I think with all the investments pouring in, there might still be a peak to this. And we probably haven't seen that slowdown happen yet. But what is happening, on the other end though, is people currently have to go out of their way to find generative AI solutions. But by next year, a lot of these will be integrated into the tools they use already. Big tech is already co-opting a lot of these features into various tools and apps. And so whether they want to or not, it'll be at their disposal, so I think that will just make it more commonplace. And that should push up the usage as well as level off all those startups that are basically one-trick pony AI companies.

Marcus Johnson:

Mm-hmm.

Jacob Bourne:

Mm-hmm, mm-hmm. Yeah, I mean yeah, we haven't seen the slowdown yet. There's still a lot of hype and excitement around this technology. But at the same time, I think the skeptics' concerns are very valid.

There's a lot that goes wrong with generative AI that could impede the trillions of dollars in economic value that's expected from becoming a reality. I mean, it hallucinates. It makes up stuff that gets things wrong. People can bypass safeguards. There's bias in the output. There's all kinds of things that make the tool less useful than what it's sometimes advertised as, and the problem is that it's really difficult for tech companies to address and overcome these challenges.

And even within tech companies, there are people that are really concerned about this, including at Google. And of course, we're talking about the leading tools like ChatGPT, and people are also concerned about sometimes what's perceived as a varying level of performance. You don't log in and get the same type of a performance every time you use it, which is an issue for people that are really relying on it to do their jobs on a daily basis, right?

Marcus Johnson:

Right.

Jacob Bourne:

Yeah.

Marcus Johnson:

Do you think it's going to change though? Because you have to see if they work properly before you can start using them.

Jacob Bourne:

[inaudible 00:10:48]-

Marcus Johnson:

Trishla Ostwal-

Jacob Bourne:

Yeah.

Marcus Johnson:

... of Adweek was giving an example, saying Digital marketing agency Winclap received 40 new pitches for generative AI tools in Q3 alone. After testing, the agency discarded over 60% of them due to, as you mentioned, issues like hallucinations. We're going to talk more about the rules that we can expect to see from AI in an episode coming out this coming Monday, but one of the things that we're going to talk about on the episode is the idea of, do these companies need these tools to be approved before they hit the market? Because it doesn't seem like they are, and that seems like it's a problem with adoption or worse, things going bad.

Gadjo Sevilla:

And to add to that, I mean AI regulation is proving to be extremely complex. The government has tried to make suggestions, but they can't even find a starting point. And it's been offloaded to the companies themselves, which, as we all know, they will regulate as best suits them, meaning you can't really expect an industry-wide initiative to come from that at this point.

Marcus Johnson:

Mm-hmm.

Jacob Bourne:

Mm-hmm.

Marcus Johnson:

Yeah.

Jacob Bourne:

Yeah. I mean, it's a sticking point when AI companies can't ensure a consistent product on a daily basis because it's such a volatile technology. At the same time, because it's difficult to fix that and because it doesn't seem like generative AI is going anywhere, well then what happens? Where do we go from here? I think tech companies [inaudible 00:12:16] be investing a lot of money in trying to overcome these challenges, but let's not hold their breath.

Marcus Johnson:

Yeah. Gadjo, to your point about self-regulation, there was a study done by AI Policy Institute, and over 80% of people were saying that they do not trust AI companies to self-regulate. And that's 56% of them supporting federal regulation. The thing with federal regulation, though, it's interesting because the AI companies, they say they also want regulation, but you do have to watch the wording of their statements. And I thought Sarah Myers West, who's managing director for the research center AI Now Institute, says, "Look at the very, very delicate phrasing that OpenAI uses when they make their calls for regulation. There's always a qualifier, like saying, 'We want regulation for artificial general intelligence,'" which Jacob, you've pointed out many times is the next version. After generative AI, we would get to AGI, artificial general intelligence. So that's something they're saying. "We want regulation for something that doesn't exist yet or for models that exceed a particular threshold," thus excluding everything that they've already come out with-

Jacob Bourne:

Mm-hmm, mm-hmm.

Gadjo Sevilla:

Yeah.

Marcus Johnson:

In commercial use use.

Jacob Bourne:

Mm-hmm.

Marcus Johnson:

So I thought it's interesting that you got to pay attention to the wording of these companies when they're saying, "Yeah, we also want regulation." What do Americans want? I mean, 56% of them say they support federal regulation. 72% of American voters say they want a slowdown in the development of AI. That's versus just 8% who prefer speeding it up, according to new polling from think tank AI Policy Institute and YouGov.

So Americans seem to want to slow things down here. Back in March, there was a letter, the open letter co-signed by Elon Musk and thousands of other people in the AI community, demanding a pause in AI research. British newspaper The Guardian was reminding us of that. So it seems like the industry had called for a bit of slowdown. People are calling for a bit of slowdown. Any chance of that at all, or is it just full throttle at this point and that's really just PR speak for the companies and something that's going to be ignored for the masses?

Jacob Bourne:

Yeah, I don't see an intentional type of slowdown happening. I see an indirect slowdown potentially happening just because of these challenges and also because of the chip shortage, for example, is having an impact on the pace of model training at this point. So those types of things could slow it down, but I don't see it as being an industry or regulatory-driven initiative where "Okay, this technology's advancing so fast, let's slow it down." I don't think there's any sign of that. Not that public opinion doesn't have a sway, but I think at this point it's important to remember that a lot of people don't even really know about generative AI. There's, I would say-

Marcus Johnson:

Great point.

Jacob Bourne:

... a significant number of Americans that aren't really familiar with ChatGPT, haven't used it. Maybe they've heard the name, but haven't dove into what it's all about.

At the same time, I think what we might end up seeing is that... The enterprise and tech companies are, of course, very.... And investors, especially, very excited about AI because of the economic value. But I don't think the average person is going to learn about generative AI and think, "Oh, I'm going to get rich off this technology." And so I think we're going to see public opinion definitely sway towards this end of really wanting to see a slowdown, wanting to see the most responsible kind of AI development possible from companies, and that eventually could put pressure in the government to do more than it's doing now.

Marcus Johnson:

You make a good point. Most people haven't used it, so how do Americans know what they want from AI or how can they understand what they're talking about enough to say, "We do want to a slowdown and we do want it to be sped up"? You pointed out overall, not many people have used it. But half of those under 40 have used an AI chatbot, but just one in 10 over 60 have, according to a June BlueLabs poll. So I think that's a really good point.

Finally, folks, Jacob, you recently wrote that the generative AI frenzy could face a 2024 reality check. How so?

Jacob Bourne:

Well, some are saying that the sector is way overhyped, and there's some evidence that this could be the case. I mean, the tools are out, they've been commercially deployed, but it doesn't mean that companies are actually making a profit off of it. For example, Microsoft has been losing money on its GitHub Copilot, its autonomous coding software, on average $20 a month per user and up to $80 per month per user. So significant.

Marcus Johnson:

Losing. Wow.

Jacob Bourne:

Losing money. Google has had to raise its prices on AI tools it's making available through the cloud for similar reasons. AWS, Amazon Web Services, has gotten complaints from customers that its prices are too high. And it's a difficult thing because the cost of training models, deploying models is very high for companies, and so there's a threshold at which people want to pay for tools that haven't been perfected yet.

I'm not completely convinced that there's going to be this crisis for the sector come 2024. What this all really depends on is high-performance, efficient, low-power-consuming AI chips, and that's where a lot of investment is going into. I mean, Nvidia has plans for new chips coming out next year. Even the AI companies themselves are developing those chips. More players are getting into the AI chip race. And so I think what we're going to see next year is a lot of new chips come on the market to try and bring down the cost of model training and deployment to help the sector become more profitable.

Marcus Johnson:

So we're going to skip the halftime report because I want to end with a question for you, Gadjo. And it's going back to marketers using generative AI and this idea that the creative, at some point, might start to feel quite, for lack of a better word, samey. And so what I mean by that is, so the reason we have different creative in marketing today is because different companies employ different people with different ideas. Whereas if marketers eventually end up using one of, say, a handful of generative AI tools, will everyone's output start to look similar?

And so Tinuiti's practice lead of emerging technology, Nirish Parsad, he was saying that, quote, "If these AI systems become ubiquitous, there's a looming danger of brands all having highly similar creative, which leads to a homogenous brand landscape." Is that a possibility here? If you have all these marketers using a handful of tools, that small collection of tools, that the output is all going to start to look quite similar, or is it pulling from so many places that that's not really possible?

Gadjo Sevilla:

I think at the very start, you might see the similarities. Once they do refine these tools, however, and get more input maybe by sourcing original work from photographers, artists, things that have a certain look and a certain soul to them, then it's only going to be as good as the source material that it's copying from, I think. But you do see that now. When you do certain prompts, the results are predictable, in a sense. You get to see similar elements across different tools.

Marcus Johnson:

Mm, all right. Interesting. All right, gents. We'll leave the first half there. We'll move straight to the second. Today in other news, ChatGPT can now give real-time information. And what is a PC with artificial intelligence inside of it capable of?

Story one, ChatGPT can now give real-time information, putting publishers on high alert, suggests our briefings analyst, Daniel Konstantinovic. The chatbot ChatGPT can now access up-to-date content from the internet whereas before, it could only access information from 2021 and earlier. Now, it can pull from what happens just a second ago. ChatGPT can now provide real-time answers and a link to its sources, but the concern is that users may be unlikely to click through after getting the answers they were looking for. So you ask it something, it just gives you the answer. And here's a link to the source, but I don't need that link because you already told me. The feature's available to paying users, but will eventually be available to all. Jacob, how much out of 10 do you think ChatGPT will hurt publishers?

Jacob Bourne:

Yeah, at this point I'm going to give it about a six-level threat. It really depends on performance. I'm not giving it higher than that because I think there's an element of trust here. I think people do want to get their information from trusted sources. And generative AI at this point, I don't think it's risen to that level of trust in people's minds yet. However, six is a pretty significant threat level, and I think it is an existential threat.

What companies face here is a potential reduction in overall website traffic. For publishers in particular, a loss of ad revenue, loss of subscription revenue. When they get less foot traffic, they're getting less user data. There's the potential for spread of misinformation by the chatbots. So there's a lot of threats here and I think it's a serious one, but it's not a very near-term threat because of the performance issue and then because of the trust issue.

Marcus Johnson:

Okay. Story two. Gadjo, you recently wrote that the PC industry, personal computers, is betting on AI features to overcome its sales slump. You explained that as global PC sales continue to... I don't know why I just explained what PC stands for. I'm so sorry. Let's move past it.

You explained that as global PC sales continue to decelerate, HP, Lenovo, and Dell are adding generative AI features to the next generation of laptops, desktops, and servers. HP and Lenovo, you point out, both said that AI PCs for general availability could hit shelves July 2024, so next summer, or early 2025, according to The Register. But Gadjo, what can we expect AI PCs to do differently?

Gadjo Sevilla:

So I think the biggest advantage of on-device AI is reduced latency. The PCs running neural networks onboard will be able to do things in real time without having to access the cloud. There's also a huge security upside for this because we know that companies [inaudible 00:22:37]-

Marcus Johnson:

We know everyone uses the same password. Yes.

Gadjo Sevilla:

That and they always say, "You can use AI, but don't use our official documents or any business information on the AI because you can't put it out there."

Marcus Johnson:

Right.

Gadjo Sevilla:

So that's going to change because these AI PCs will be able to be locked down behind company firewalls. And they will have access to all the company's data which, again, it's going to be huge for privacy and I think huge for adoption.

Marcus Johnson:

Yeah. That's all we've got time for, unfortunately. Gents, thank you so much. As always, thank you to Gadjo.

Gadjo Sevilla:

Thank you very much. Glad to be here.

Marcus Johnson:

Yes, sir. Thank you to Jacob.

Jacob Bourne:

Thanks, Marcus. Thanks, Gadjo.

Marcus Johnson:

And thank you to Victoria who edits the show, James who copy edits it, and Stuart who runs the team. Thanks to everyone for listening in to the Behind the Numbers Daily, an eMarketer podcast made possible by Intuit Mailchimp. You can tune in tomorrow to hang out with host Sara Lebow on the Reimagining Retail Show as she speaks with VP of content, Suzy Davidkhanian, and senior analyst on the retail team, Zak Stambor, all about the genius of Costco.