Google conducted a rather large global survey of approximately 1,000 citizens per country across 17 countries to find out how people are feeling about AI. The differences per region are stark and one topic bubbles to the top across the board. 

This chart shows the positive impact respondents expect to see from AI in the next 25 years. The US lags emerging economies by a lot.

Education and healthcare are the top areas where respondents expect to see benefits from AI. And for healthcare specifically, it is the top category for all but the emerging economies.

Fewer respondents think that current AI systems are showing positive impact on health currently but its in the running.

Head over to the Google article and grab a copy of this survey of over 17,000 global citizens. You have to scroll to the bottom to get the PDF.

Here's a view into the data we have "behind the curtain." We have made statements before that X/Twitter's ad revenue was down but the number of users was essentially stable. Finding public information on this was difficult to say the least. Today, in a Wire exclusive, we have permission to share some Pathmatics data with our audience. 

When Elon Musk took over Twitter revenue fell by 51% the next month (reporting is offset by one month). To be fair it's easy to see an annual pattern in the data but the 51% drop was unprecedented.

That is, until further shenanigans helped the number drop 46% from June to July which put revenue at only 26% of it's pre-Musk high. Some returned before the holidays but the visual speaks for itself.

It's a bit hard to see on this small chart but the fifth largest category is Health & Wellness. The drop at first follows the general trend but in Q3 2023 it recovers and is actually up year-over-year by 20%.

Health and wellness isn't pharma, however. 

These are the top 10 health spenders on the platform which are mostly pharma and health devices. We blurred out the names as per our policy but left the industry advocacy group, PhRMA. 

We can see throughout 2022 Twitter was having trouble keeping these advertisers because of ad risk issues. When Musk took over is was a nearly total drop with one company hanging on for a bit and then another returning through 2023.

We have said this many times, X/Twitter users aren't leaving. This is especially true for HCPs who follow hashtags around annual meetings. If the company fails from lack of revenue these users will be searching for an alternative, which is certainly what Meta's Threads platform is banking on.

There is a downward slope on the user numbers, -17% from when Musk took over the company, but it's not a collapse like you may have read previously. The brand risk is still high, however. 

The one company that is still on the platform essentially has the whole place to themselves. 

True story, Peter Flaschner and I (don't use first person, Peter's a cool guy though -Ed.) were talking about podcast ideas just last week. YouTube came up and it was a bit of a blind spot for us but it turns out it's a podcast powerhouse.

The top platform for podcast listening, and watching, isn't Spotify, it's YouTube. This, according to two different studies.

It's more like 1.25 years ago, but the tipping point for YouTube was in Q3 2022 and it hasn't looked back. When looking at this "favorite platform" chart note that 45% of people don't consider on of these three top platforms their favorite. The podcast market is quite distributed with many smaller platforms serving the market.

If you're owned by the world's largest search engine you might be good at discovery... and in YouTube's case that's true. Respondents put it solidly at the top of the chart.

Wrong! This study also included data on what marketers are doing. We really have to give it to Cumulus, yes they have an agenda but they bring data.

60% of marketers are using podcasts and 80% are discussing them in meetings. Don't suffer from professional FOMO ... PFOMO? Combine podcasts with your influencer plans (don't assume things -Ed.).

Meta has committed to building a "general artificial intelligence." Technically Mark Zuckerberg, CEO of Meta, said "general intelligence" but any difference will be a technicality. The software will also be made open source, but again, Zuckerberg said "we should open source and make it as widely available as we responsibly can" (emphasis ours). 

This statement was made as a short video on Instagram which you can view directly:

As part of the initiative Meta has announced it is purchasing 350 thousand more Nvidia H100s. To put that in context, a deal with India for 16 thousand H100s was estimated at $500 million. At those prices, the Meta deal would be worth upwards of $11 billion. It's likely less because of discounts but still. That is why Nvidia, historically a gaming GPU maker, has a stock price that looks like this:

It seems to be standard practice when talking about stocks to mention that we're not financial analysts and this is not investment advice. So there you go (prudent -Ed.).

The World Federation of Advertisers—no, we hadn't heard about them either—conducted a survey on AI use with their members. The results are directional at best being based on 69 senior respondents across 55 WFA member companies.

The full results are for members only but we took some of the data in the press release and created a couple to top results charts. 

When asked how they were approaching generative AI, most marketers are on board.

When asked how they were using, or planning to use, GenAI marketers started with content creation.

There is more raw data in the press release on concerns about GenAI and other topics.

Last Friday the FDA had a webinar to discuss its reorganization. The extent of the changes are pretty substantial:

The yellow highlight is ours, and emphasizes that CBER, CDRH, and CDER are all shown as directly impacted. Having said that we didn't hear anything in the webinar that sounded like it would fundamentally change pharma's relationship with the agency.

If you're interested, the webinar portion is only 30 minutes and maybe you'll pick up on something we didn't. (I happen to know the author was multitasking -Ed.).

It looks like US hospitals are recovering from a bad start to the year, but are still not at pre-pandemic levels. Any stability in the healthcare market should be cause for increased optimism, especially for marketers looking to talk with payers.

Regardless of what the news-manufacturing sites tell you, it's a recovery year for the economy. Ad spend is tracking to that and is set for a 13.2% increase in digital.

Meanwhile, social is gaining more of those digital dollars with a full quarter of spend predicted in five years. Also note the pandemic bump as search and social proved their responsiveness to the lockdowns and changes in consumer behavior.

A post on LinkedIn highlights an all-AI-generated TV ad that ran in Poland. This looks like a first for actual, in-market ads as it was 100% developed using generative AI.

Something may have been lost in translation, however, as at first blush it looks like an HR violation waiting to happen. Plus, it's a bit on the nose for North American audiences.

Watch the ad at the link to see what we mean. This product is OTC in Poland in case your MRL alert just went up (don't talk to the audience -Ed.).

According to comments on the post, the ad cost 80-90% less to make than a traditional ad and it tested above benchmark in at least some metrics. This is likely because of the unprecedented nature of the ad. Two things will happen in tandem:

One thing is for certain, marketers as storytellers will have new powerful tools to help them.

It's telling that the headline of this story is that AI is not the top issue for marketers, but it's good to remember that all of our other issues haven't disappeared. 

This story contains a lot of data from a survey of 300 marketers but Brandwatch could have done a better job of presenting it. The numbers are embedded in the text and you have to read it carefully to extract the data. We charted the top concerns from this text.

A mobile app data broker has been banned from selling location data by the FTC. The location data included sensitive places such as "medical and reproductive health clinics, places of religious worship and domestic abuse shelters."

The customers of this data were more than just advertising targeting companies. 

Contentsquare surveyed 2,751 digital professionals across 48 countries who work in marketing, user experience, and product management. They asked these folks an open-ended question:

They then used ChatGPT 4 to pull out the themes from the mass of data. So, technically AI didn't say AI was important, your marketing peers did (don't talk to the audience -Ed.). The top 10 trends were: