Grindr aims to build an AI “wingman” for its users. The dating app’s CEO spoke with WSJ CIO Journal reporter Belle Lin about the approach to building an artificial intelligence agent at The Wall Street Journal’s CIO Network Summit. Plus, over-the-counter hearing aids were meant to make the technology more accessible for people with mild to moderate hearing loss. So, what’s behind the high return rates? Zoe Thomas hosts.
Sign up for the WSJ's free Technology newsletter.
Learn more about your ad choices. Visit megaphone.fm/adchoices
[00:00:00] [SPEAKER_01]: Robert Half Research indicates 9 out of 10 hiring managers are having difficulty hiring. Robert Half is here to help. Our recruiting professionals utilize our proprietary AI to connect businesses with highly skilled talent. At Robert Half, we know talent. Visit roberthalf.com today.
[00:00:23] [SPEAKER_02]: Welcome to Tech News Briefing. It's Friday, October 11th. I'm Zoe Thomas for The Wall Street Journal.
[00:00:31] [SPEAKER_02]: Over-the-counter hearing aids were supposed to make the tech more accessible. So why are people returning them at high rates? We'll find out.
[00:00:41] [SPEAKER_02]: And then, could an AI wingman help you find a long-term relationship? The CEO of Grindr thinks so.
[00:00:49] [SPEAKER_02]: We'll hear from George Arison on how his company is approaching building an artificial intelligence agent that could help users keep track of conversations, suggest dating spots, and one day, even do some of the dating for you.
[00:01:06] [SPEAKER_02]: But first, over-the-counter hearing aids were supposed to be a game-changer for people with mild to moderate hearing loss.
[00:01:14] [SPEAKER_02]: Instead, these devices are being returned at rates of 15 to 30 percent, according to some audiologists and retailers.
[00:01:22] [SPEAKER_02]: A large independent appliance and electronics store in Illinois stopped carrying them over the past two months because the rate of return was 40 percent.
[00:01:31] [SPEAKER_02]: Here to tell us what's going on is our family and tech columnist, Julie Jargon.
[00:01:35] [SPEAKER_02]: So, Julie, the Food and Drug Administration approved over-the-counter hearing aids nearly two years ago.
[00:01:42] [SPEAKER_02]: Why was that such a big deal?
[00:01:43] [SPEAKER_00]: Yeah, when that happened, the idea was that this would open up a huge opportunity for people with mild to moderate hearing loss to get the hearing help that they need without having to go so far as to pay for expensive prescription hearing aids, which are really intended for people with more severe hearing loss.
[00:02:01] [SPEAKER_00]: So, by allowing retailers to sell devices over-the-counter at a lower price point than prescription hearing aids that require you to see an audiologist, the idea was that this would give a lot of people less expensive opportunity to get hearing help.
[00:02:19] [SPEAKER_02]: What do people who return them say the problem is?
[00:02:22] [SPEAKER_00]: People are returning them for a couple of reasons.
[00:02:25] [SPEAKER_00]: One is some of the devices are not as high quality as they'd hoped and they're not providing the amplification that they need.
[00:02:33] [SPEAKER_00]: And also, a lot of times when people get these devices, even though they're over-the-counter, there are still some setup required to make sure they're adjusted properly for your particular hearing needs.
[00:02:45] [SPEAKER_00]: And some manufacturers of over-the-counter hearing aids are better than others at providing customer support.
[00:02:51] [SPEAKER_00]: So, a lot of the people that are returning them, according to retailers I've talked to and audiologists, is they're returning them because they're not getting the level of customer support they need to properly use the devices.
[00:03:03] [SPEAKER_02]: Over-the-counter hearing aids were supposed to be cheaper, but this technology can still cost thousands of dollars.
[00:03:10] [SPEAKER_02]: Why isn't the cost lower?
[00:03:12] [SPEAKER_00]: There are some low-cost ones.
[00:03:13] [SPEAKER_00]: However, those are the ones that tend to not have the level of customer service that people would like, and those tend to be lower quality.
[00:03:21] [SPEAKER_00]: A lot of the over-the-counter hearing aids still cost in the $1,500 to $2,000 range.
[00:03:26] [SPEAKER_00]: And those tend to be higher quality hearing aids, and there is a cost associated with having clinicians on hand to help people fit them and send updates over to their phones.
[00:03:38] [SPEAKER_00]: So, they ended up being more expensive than initially thought.
[00:03:42] [SPEAKER_02]: Apple's rolling out a software later this fall that can make their AirPods Pro 2 function like a hearing aid.
[00:03:49] [SPEAKER_02]: Julie, what impact is that expected to have on the market?
[00:03:52] [SPEAKER_00]: Yeah, their existing AirPods Pro 2 will have hearing aid functionality starting this month.
[00:03:58] [SPEAKER_00]: And there's an expectation that this could eat into the lower cost over-the-counter hearing aid market.
[00:04:04] [SPEAKER_00]: But on the other hand, some audiologists and hearing aid manufacturers are hopeful that it will raise awareness that once people who maybe already have these AirPods or buy a new pair, which is $249,
[00:04:17] [SPEAKER_00]: will realize that they have hearing loss that they might not have really realized they had, and experience what it's like to hear sound more clearly.
[00:04:28] [SPEAKER_00]: And that if they need something more long-lasting or durable, that they might turn to the more expensive over-the-counter hearing aids.
[00:04:37] [SPEAKER_02]: All right, that was our family and tech columnist, Julie Jargon.
[00:04:41] [SPEAKER_02]: Coming up, could an AI agent one day date another AI agent?
[00:04:47] [SPEAKER_02]: We'll find out what the CEO of dating app Grindr thinks about that after the break.
[00:05:17] [SPEAKER_02]: At this point, you've probably used an AI chatbot like ChatGPT or Gemini to answer a question.
[00:05:30] [SPEAKER_02]: But an AI agent is understood to go a bit further than answering questions, summarizing text, or creating content.
[00:05:39] [SPEAKER_02]: Agents can actually go out into the digital universe and take action on your behalf without human intervention.
[00:05:47] [SPEAKER_02]: The dating app Grindr is working on an AI agent that could act as a wingman for its user base of mostly gay and bisexual men.
[00:05:55] [SPEAKER_02]: At the Wall Street Journal's CIO Network Summit, Grindr CEO George Arison spoke with WSJ reporter Bell Lin about the company's plans.
[00:06:05] [SPEAKER_02]: Here are highlights from their conversation.
[00:06:08] [SPEAKER_03]: So agents, how are they going to transform dating?
[00:06:11] [SPEAKER_03]: How are they going to transform even aspects of our lives beyond dating?
[00:06:14] [SPEAKER_04]: Grindr has a lot of data and has a user base that's very connected to the product.
[00:06:18] [SPEAKER_04]: People spend over an hour in the app a day.
[00:06:20] [SPEAKER_04]: And so I thought there was a ton of opportunity for AI.
[00:06:23] [SPEAKER_04]: And the really big takeaway happened for me when I realized that all these use cases that we want to expand into might complicate the app.
[00:06:31] [SPEAKER_04]: Because the app today is really simple.
[00:06:32] [SPEAKER_04]: But if you built them with AI first approach, you can actually avoid that.
[00:06:37] [SPEAKER_04]: Because if the AI can guess the direction you want to go into and then direct you in that way as you're surfing through the product,
[00:06:43] [SPEAKER_04]: then you're not putting a lot of additional buttons and places for people to go, but rather directing them in that direction.
[00:06:52] [SPEAKER_04]: And so we want to add a bunch of use cases, but we want it to be very non-intrusive in people's behavior.
[00:06:59] [SPEAKER_04]: And I think AI can make that happen in a really good way.
[00:07:02] [SPEAKER_03]: So what about the kind of agentic parts where the agent, some point in the near term future, can actually do things for us?
[00:07:10] [SPEAKER_03]: Is it AI agents dating other AI agents at some point?
[00:07:14] [SPEAKER_04]: I don't know about dating, but at least starting a conversation, right?
[00:07:18] [SPEAKER_04]: Even before Gen.AI, through heuristics, you had applications out there would suggest text to you like, hey, this is how you should respond.
[00:07:27] [SPEAKER_04]: So there's no reason why that couldn't happen on a user-based level.
[00:07:31] [SPEAKER_04]: Right.
[00:07:31] [SPEAKER_04]: I mean, I think it's kind of like, how do I start this conversation?
[00:07:34] [SPEAKER_04]: What should I say?
[00:07:35] [SPEAKER_04]: I mean, it's a very simple thing that an agent should be able to do for you and I think will.
[00:07:40] [SPEAKER_04]: But the more ideal version of that would be not just like general, hey, these are things you could use to start with, but rather based on how you normally talk and what you normally say.
[00:07:48] [SPEAKER_04]: And what I know about the person you're about to chat with, here's suggested language.
[00:07:53] [SPEAKER_04]: And that's far more compelling than just a general set of ideas of what you could say.
[00:07:58] [SPEAKER_04]: The next kind of batch of that to me would be, okay, you guys are talking about going out to get coffee.
[00:08:04] [SPEAKER_04]: How about here are three places where you could go get coffee in between where that person is located and where you're located, right?
[00:08:10] [SPEAKER_04]: Because Grindr is very location-based.
[00:08:12] [SPEAKER_04]: That's actually pretty convenient as well.
[00:08:14] [SPEAKER_04]: And then the eventual step, yes, some people will choose to do this.
[00:08:18] [SPEAKER_04]: Some won't.
[00:08:19] [SPEAKER_04]: We very much believe in not forcing people to do anything.
[00:08:22] [SPEAKER_04]: But for those that do choose to do this, there's no reason why some of the initial conversations and kind of validation can happen through AI talking to AI, right?
[00:08:31] [SPEAKER_04]: So I have a replica of myself.
[00:08:32] [SPEAKER_04]: You have a replica of yourself.
[00:08:34] [SPEAKER_04]: And those two talk to each other.
[00:08:35] [SPEAKER_03]: How are you also thinking about the risks?
[00:08:38] [SPEAKER_03]: So what kind of keeps you up at night when you're thinking about the ways in which the wingman could go awry or could hallucinate or tell users when they're seeking information they might not know, sexual information.
[00:08:49] [SPEAKER_03]: It could lead them in the wrong direction.
[00:08:51] [SPEAKER_03]: There's also questions about liability.
[00:08:53] [SPEAKER_03]: So are you responsible in that instance?
[00:08:55] [SPEAKER_03]: Data privacy is incredibly important to your users.
[00:08:57] [SPEAKER_03]: The risk of a cybersecurity breach.
[00:08:59] [SPEAKER_03]: And also prompt engineering.
[00:09:00] [SPEAKER_03]: The same sort of like prompt injections that can be applied to LLMs now could also be applied to agents.
[00:09:07] [SPEAKER_03]: I could go on.
[00:09:07] [SPEAKER_03]: But how are you thinking about all of this?
[00:09:09] [SPEAKER_04]: So the biggest risk we face actually is to do nothing.
[00:09:12] [SPEAKER_04]: The same way that there were many products for people to meet each other prior to render existing.
[00:09:18] [SPEAKER_04]: There were web-based products only.
[00:09:20] [SPEAKER_04]: And then the iPhone came out.
[00:09:22] [SPEAKER_04]: And there was a platform shift happening.
[00:09:25] [SPEAKER_04]: And those companies chose not to move to the mobile platform.
[00:09:29] [SPEAKER_04]: Gave Grindr a huge opening.
[00:09:30] [SPEAKER_04]: Well, those companies don't exist anymore.
[00:09:32] [SPEAKER_04]: And Grindr has the market share that they gained from them.
[00:09:35] [SPEAKER_04]: That's the biggest risk.
[00:09:36] [SPEAKER_04]: We don't actually innovate.
[00:09:38] [SPEAKER_04]: And then someone comes out and does something that is AI first.
[00:09:41] [SPEAKER_04]: And suddenly we are then trying to compete with them.
[00:09:43] [SPEAKER_04]: So that would be, I think, actually the single biggest risk.
[00:09:45] [SPEAKER_04]: Obviously, privacy is critical.
[00:09:47] [SPEAKER_04]: And we are extremely privacy first.
[00:09:49] [SPEAKER_04]: We actually get challenged a lot about like, why don't you ask for ID verification?
[00:09:53] [SPEAKER_04]: Why don't you ask for a photo from a person that's a face verified photo?
[00:09:57] [SPEAKER_04]: Well, the reason is because a lot of people still don't have the luxury to be as open about who they are.
[00:10:04] [SPEAKER_04]: We just released a new privacy policy recently, which users have to actively accept.
[00:10:09] [SPEAKER_04]: That allows us to use data inside the app to train the model.
[00:10:13] [SPEAKER_04]: So we've not done that.
[00:10:14] [SPEAKER_04]: And we'll only do that with user consent.
[00:10:16] [SPEAKER_04]: And ultimately, a lot of what we'll be doing with AI will be based on user choice.
[00:10:20] [SPEAKER_04]: It will not be, hey, here's a product.
[00:10:22] [SPEAKER_04]: You must use it this way, right?
[00:10:24] [SPEAKER_04]: It will be, this is available to you.
[00:10:26] [SPEAKER_04]: If you'd like to use it, you can.
[00:10:27] [SPEAKER_04]: But you don't have to if you don't want to.
[00:10:30] [SPEAKER_03]: You said that you can't afford not to make the big bets in technology.
[00:10:34] [SPEAKER_03]: And also that you're asking users if they want to consent to this up front.
[00:10:38] [SPEAKER_03]: You're being very proactive.
[00:10:39] [SPEAKER_03]: But still, how do you think about liability?
[00:10:41] [SPEAKER_03]: Are your investors concerned about what AI could do?
[00:10:44] [SPEAKER_03]: Where it might mess up?
[00:10:46] [SPEAKER_04]: I think the biggest risk on the messing up, at least that I can imagine right now, is around
[00:10:52] [SPEAKER_04]: hallucinations.
[00:10:53] [SPEAKER_04]: And we can have liability protections on that.
[00:10:55] [SPEAKER_04]: And we will, obviously.
[00:10:57] [SPEAKER_04]: But nevertheless, AI says something that it's not supposed to say or it's not good to say.
[00:11:02] [SPEAKER_04]: Then there's like screenshots about that.
[00:11:04] [SPEAKER_04]: Then someone tweets about it and then it takes over.
[00:11:06] [SPEAKER_04]: And so we have to be really careful about that.
[00:11:08] [SPEAKER_04]: And we are.
[00:11:09] [SPEAKER_04]: And we put a lot of protections in place to try to manage for that.
[00:11:12] [SPEAKER_04]: But it'll still break.
[00:11:13] [SPEAKER_04]: And we just need to be prepared for that.
[00:11:14] [SPEAKER_04]: And then secondly, it could lead people to do something that they're not supposed to do
[00:11:19] [SPEAKER_04]: themselves.
[00:11:20] [SPEAKER_04]: But again, ultimately, users have to be responsible for their own lives.
[00:11:24] [SPEAKER_04]: I get into my Tesla at my house every day.
[00:11:28] [SPEAKER_04]: And literally right out of the garage, I turn on self-driving.
[00:11:31] [SPEAKER_04]: And it drives me to the office.
[00:11:32] [SPEAKER_04]: And there's one place where I have to intervene.
[00:11:35] [SPEAKER_04]: And otherwise, it drives itself.
[00:11:36] [SPEAKER_04]: But I'm taking on that liability.
[00:11:38] [SPEAKER_04]: And that's my choice to do that.
[00:11:41] [SPEAKER_03]: So if you get lost, you're not blaming Elon?
[00:11:43] [SPEAKER_03]: No, I'm not blaming Elon.
[00:11:44] [SPEAKER_04]: And also if I get into an accident, I'm not blaming Elon.
[00:11:46] [SPEAKER_04]: That's on me.
[00:11:47] [SPEAKER_04]: On the flip side, you could take the approach, which is we're going to make this absolutely
[00:11:50] [SPEAKER_04]: perfect before it's in users' hands.
[00:11:52] [SPEAKER_04]: That's too risky.
[00:11:53] [SPEAKER_04]: I don't know if you'll ever get there.
[00:11:54] [SPEAKER_04]: And frankly, whether it's going to be economical or not.
[00:11:57] [SPEAKER_03]: All right.
[00:11:57] [SPEAKER_03]: We could end on that note.
[00:11:58] [SPEAKER_03]: Thanks so much, George.
[00:11:59] [SPEAKER_03]: Thank you.
[00:12:00] [SPEAKER_02]: That was Grindr CEO George Arison speaking with WSJ reporter Belle Lin.
[00:12:05] [SPEAKER_02]: And that's it for Tech News Briefing.
[00:12:07] [SPEAKER_02]: Today's show was produced by Julie Chang.
[00:12:10] [SPEAKER_02]: I'm your host, Zoe Thomas.
[00:12:11] [SPEAKER_02]: Jessica Fenton and Michael LaValle wrote our theme music.
[00:12:15] [SPEAKER_02]: Our supervising producer is Catherine Millsop.
[00:12:18] [SPEAKER_02]: Our development producer is Aisha Al-Muslim.
[00:12:20] [SPEAKER_02]: Scott Salloway and Chris Zinsley are the deputy editors.
[00:12:24] [SPEAKER_02]: And Falana Patterson is the Wall Street Journal's head of news audio.
[00:12:28] [SPEAKER_02]: We'll be back this afternoon with TNB Tech Minute.
[00:12:31] [SPEAKER_02]: Thanks for listening.
[00:12:32] [SPEAKER_02]: Bye-bye.

