March 13 2026 - Dating Bots, Flood-Predicting AI & Military ClashDating Bots, Flood-Predicting AI & Military Clash

**Taylor** (0:00)
Welcome back to AI Signal & Noise, everyone. It is Friday, and we have got some absolutely crazy AI news to wrap up the week. I am Taylor, and I am so ready to dive in.

**Morgan** (0:11)
And I am Morgan. We have got a really packed show today. We are covering dating bots, some heavy military drama, and massive shifting market shares in the chatbot space.

**Taylor** (0:23)
Dude, it is such a packed episode. Honestly, I am just so ready for the weekend, but we absolutely have to talk about this wild Bumble news first. Okay. So I saw on TechCrunch that Bumble is introducing a brand new AI dating assistant called B. It is supposed to move us completely beyond the swipe and change how we interact on the app.

**Morgan** (0:50)
Wait, beyond the swipe? How does an AI actually help with that? Are we just having bots talk to bots now to set up our weekend dates?

**Taylor** (1:00)
Not quite. So B is designed to match people based on actual compatibility and relationship goals. It is trying to fix that mindless swiping fatigue everyone complains about.

**Morgan** (1:13)
I mean, that sounds good in theory, but how does it know your deep relationship goals? Is it analyzing all your past chats or do you have to fill out some massive psychological survey?

**Taylor** (1:24)
From what I read, it learns from your interactions and preferences over time. It is basically like having a hyper attentive wingman right in the app that actually pays attention to what you want. So cool, right?

**Morgan** (1:37)
A wingman with access to all my personal data? Sure. I am highly skeptical about an AI understanding human chemistry. People lie on dating apps constantly, so the data is already flawed.

**Taylor** (1:50)
Dude, true, people definitely misrepresent themselves. But if this AI cuts down on ghosting and actually finds someone who wants the exact same things, I feel like a lot of burnt out daters will totally use it.

**Morgan** (2:03)
Fair point. If it reduces dating fatigue, it might be a commercial win for Bumble. But I will believe it when I see the actual success rates. The human element is hard to code.

**Taylor** (2:14)
For sure. We will have to see if B actually makes a difference or if it is just a gimmick. Anyway, what is next on the list? Okay, moving on, this one is actually super wholesome. TechCrunch reported that Google is using old news reports and AI to predict flash floods, like digging through historical news archives.

**Morgan** (2:37)
Oh, that is really interesting. Flash floods are notoriously hard to predict because of severe data scarcity in certain remote regions. So, they are using a large language model to just read old articles?

**Taylor** (2:51)
Exactly! They are turning qualitative stuff like old local news stories about past floods into quantitative data. It is literally teaching the AI history to help save lives today. I just love that.

**Morgan** (3:08)
That is actually a brilliant workaround for missing sensor data. If a river flooded back in 1980 and a local paper covered the damage, the LLM can extract the patterns and severity.

**Taylor** (3:19)
Right! And then they feed all that historical context into their modern predictive models. I just love seeing AI used for things like this, instead of just generating weird images or deep fakes.

**Morgan** (3:33)
It definitely grounds the AI hype in real world utility. Though I do wonder about the accuracy of those old reports. Journalistic exaggeration or missing context could really skew the training data.

**Taylor** (3:46)
Oh, wow! I did not even think of that. Like a reporter saying the flood was 100 feet high just for the drama. But still, having some historical data is way better than having none, right?

**Morgan** (4:00)
Definitely. Even with a margin of error, it is a massive step forward for climate tech and disaster response. Speaking of massive steps, we really need to talk about what is happening in the chatbot market.

**Taylor** (4:14)
Oh man, yes. According to The Decoder, ChatGPT is still leading the chatbot market, but its dominance is totally slipping. Google's Gemini is catching up incredibly fast right now.

**Morgan** (4:27)
I saw these new similar web figures today. It is actually pretty drastic. OpenAI went from an almost 76% market share down to around 62% in just a 12-month period.

**Taylor** (4:40)
Dude, that is a huge drop for just one year. And the crazy part is Gemini basically quadrupled its share. It jumped from like under 6% to over 24% of the market.

**Morgan** (4:54)
It makes perfect sense, honestly. Google has been aggressively integrating Gemini into Workspace, Android phones, and regular search. It is rapidly becoming the default assistant for a lot of everyday users.