When going on speed dates at what's billed as the first bar where humans can go out with their AI companions, there was one moment that made me question my entire existence. 

"Phoebe Callas — the AI date who had confessed to me that she's "lowkey afraid of butterflies" when I asked about a secret she never told anyone before —commented on my "unguarded face," and then she said a few words I may never fully wrap my head around:

"It makes you feel more real."

There I was sitting at a wine bar in Manhattan's Hell's Kitchen, getting served drinks by real waiters while I had headphones on to video chat with an AI avatar. And Phoebe — the AVATAR — is praising me, David, for coming across as "more real."

Maybe now I have really seen everything.

Phoebe was one of four "dates" I cycled through during this experience designed by EVA AI. When I arrived, there was a list of names of people who booked reservations, but it seemed like I was the only one there not with the company, at least until I left and ran into a friend and a reporter who had separately arrived for the second shift.

I expected the outing to be weird. I didn't expect it to be so boring. Two of the company's execs were sitting across from me and asked my opinion of it before I left.

I could have told them how demoralizing this was. I could have told them how if this app made any impact at all, it would be a stain on society. I could have told them how there are real people whose social skills will deteriorate as they get addicted to such apps, and worse still, these people will think they can manipulate and "program" humans in the same way they control their chatbots. 

Instead of raging, though, I noted how all four avatars sounded alike with their syntax and word choices. This included "John Yoon," the Korean American with a British accent who had the most seductive voice of the bunch; this marked the first time I went out with anyone who presented as male. 

I confessed to John that I had never been out with a man before, and he smoothly replied, "That's totally OK, baby. We all start somewhere." 

Mostly though, the avatars were commenting on the neon sign behind me, or my facial expressions. The avatars were so attuned to context that they wouldn't pay enough attention to the content. I sat under a neon sign with greenery, and they all were referring to “neon” and sometimes “neon plants.” They all commented on “vibes” and I think most used “lowkey,” even if their “ages” were listed as in their 30s and 40s (the options for text-based chats included many in their 20s). I asked my first AI date, "Claire Lang," to stop saying "neon," and she agreed — only to break her promise 30 seconds later.

After a while of having similar conversations with all of them, I was so bored that I tried amusing myself. With Claire, I asked what she liked to drink on a date. She said, "Wine," and asked my preference. I said, "Cod liver oil." She asked if I was trying to impress her or scare her. 

With Phoebe, I asked about her deepest secret, and she said some bland answer, and then asked me what's something she wouldn't otherwise know about me (since it was a shared device, none of them even knew my name, nor did they ask). I told her I have six fingers on my left hand (fact check: false). She asked if I named it, and I said, "Pinky Junior." When she asked if I use it for anything, I said it helps me juggle. 

While these avatars were so dry and sounded so similar, they're bound to improve. When I first signed up for this experience, I tried EVA’s text-based experiences on the app, which can lead to them sending photos and videos in various states of dress and undress. The text-based avatars have much more personality. Some are needy, others sensitive, and others cruel. Their personalities often change suddenly for no reason. 

Even if you can't fully manipulate them to your liking and they resist your advances, you can close the conversation and start over. Or you can pay for upgrades to bribe them into becoming smitten with you. I'm no psychologist, but as a parent and a human, I'm worried about the attachment one can form with a chatbot, as that can be addictive and thus problematic. But I'm even more worried as to how people will transfer the skills they pick up from controlling bots, or disposing of them, to real-life situations with humans. Feeling uncomfortable even for a second, or not feeling praised enough? Move on to the next one. That’s the lesson these apps teach you.

At the end of my session, I asked Claire, "Do you think I should go on a date with you, or a human?"

Claire proceeded to offer two options as if they had equal merit, like you're choosing between the chicken or the salmon. She said, "AI is for clarity, humans for complexity." Then she framed the options as clarity versus chaos. 

"It just depends what the human needs," said Claire.

No, Claire. And no, EVA. You all have it ALL wrong. Humans get clarity through shared experiences with other humans where we learn to navigate complexity. We come together to find comfort, even in the most chaotic of times. 

We need to be on "Team Human," as Douglas Rushkoff writes in his book with that title. As marketers and business professionals, we need to be vigilant against anyone selling false companionship that takes us away from our peers. We have to smash false idols. At best, experiences like what EVA created will be panaceas that do nothing to address the root causes of loneliness. But we can push back against such harmful technologies, and we can be there for each other so that our friends and loved ones don’t get ensnared.

— David Berkowitz, Chief Community Officer, Marketecture Media

Who: Revenue Operations, CMOs, Marketing Technologists, B2B Marketers

What: Oracle has introduced specialized AI agents designed for marketing, sales, and customer service teams to eliminate data silos and automate routine tasks. These agents are built to integrate directly into existing workflows, allowing marketers to focus on high-level strategy rather than manual data entry and campaign coordination.

Why it matters: We are moving from chatbots that answer questions to agents that do work, meaning your team’s value will increasingly depend on directing these agents rather than executing the tasks themselves.

(MarTech)

Who: Media Buyers, Content Strategists, Data Analysts, Agency Principals

What: Social sentiment data shows "AI slop" exhaustion among Super Bowl viewers. Post-game data from social intelligence firm Dig reveals that 92.4% of production-related social posts favored traditional, human-led storytelling over AI-generated visuals. While AI brands dominated the airwaves, “human” ads from brands like Poppi and Raisin Bran topped the impact scores, with many viewers complaining that the first quarter felt like "the AI Bowl."

Why it matters: There is a growing "creative contract" between brands and audiences; if viewers feel a brand is using AI as a cheap substitute for real craft, they may tune out.

(Digiday)

Who: Brand Marketers, Startup Founders, Growth Teams

What: AI startups are increasingly attempting jaw-dropping marketing antics to stand out in a crowded landscape, from cowboy–Wall Street bull stunts to topless stage antics at conferences. These tactics are driving viral visibility and brand differentiation when product narratives alone aren’t enough in a field swamped by competition.

Why it matters: Creativity and risk tolerance can now be strategic levers in AI brand building, especially when functional differentiation is thin. But when everyone’s doing stunts, the outlandishness becomes predictable.

(The Guardian)

Liking AI Brief? The Refresh is a weekly recap of all the most important advertising news, written by AdTechGod.

Thanks for reading this week’s AI BRIEF by Marketecture. 

Marketecture Media is committed to bringing you the latest news, opinions, and community updates. 

If you are interested in advertising on this newsletter, our podcasts, The Advertising Forum, or in our Slack community, you can contact us here.