On the Radar: Online Prediction Markets, Feelings-Focused AI Models and Dead Internet Theory

Kalshi Working to Keep Kids Off Prediction Markets

What? Kalshi is the largest U.S.-regulated prediction market, primarily composed of sports betting. According to Axios, Kalshi recently announced it was “putting in place new tools to prevent minors from trading on the prediction market.”

So What? According to Common Sense Media, 36% of boys aged 11 to 17 reported gambling in the last year, including 12% who participated in sports betting. CNBC reports that nearly one-third of Gen Zers participate in prediction markets. Children are already banned from using Kalshi, but reports acknowledged that underaged users were bypassing those restrictions by using their parents’ IDs.

Now What? While new regulations will make it more difficult for minors to access the site, human ingenuity often finds ways to circumvent even the most stringent of restrictions. Likewise, there are other websites that host prediction markets and gambling features that teens may access. Parents will want to educate their children on the risks of gambling. For instance, one report found that over a period of six months, 70% of users on Kalshi never turned a profit. While news of winners gaining a quick fortune may be enticing, teens should know that these kinds of stories only make headlines because of how absurdly rare they are.

AI Models that Consider Users’ Feelings More Likely to Be Wrong

What? A study published in Nature discovered that “warm models” (AI language models trained to have friendly personas) “showed substantially higher error rates […] than their original counterparts.” Such models were significantly more likely to validate false beliefs, promote conspiracy theories and offer inaccurate information and medical advice.

So What? Science Alert reports that three out of four American teens have used relationship-based artificial intelligence—with many of these models classified as “warm models.” This makes it likely that your teen has encountered an AI model that has, based on its programming, sacrificed facts for feelings.

Now What? AI can be compared to a calculator. Yes, it’ll do a lot of the work for you, but it’s up to users to ensure the answer they get is actually correct. As AI models become a more prominent part of society, it is vital to learn how to verify truth from fiction, as AI is often unable to tell the difference. AI models can help in many ways, but they should never be used as a replacement for true research and work.

Dead Internet Theory Rises to Mainstream Concern

What? Dead Internet Theory, a concern that most of what we see online is primarily driven by AI-generated content and bots, is gaining traction among prominent voices. Reddit co-founder Alexis Ohanian and OpenAI co-founder Sam Altman both warned people of the theory. Ohanian stated on X that “At this point, it’s not *if* most of what we see online will be AI-generated. It’s that it already is.”

So What? Reports note that, in 2024, bots officially outpaced humans on the Internet, accounting for 51% of all global internet traffic. And a 2026 study indicated that more than one-third of all new websites were AI-generated or AI-assisted (with over 17% of these being entirely created by AI).

Now What? AI literacy and online discernment is needed more than ever. Chances are, you or your child have seen an AI-generated post and, without realizing it, assumed it was a human who made it. Teens should understand that AI is getting better and better at creating and mimicking human behavior online. With an influx of bots boosting certain posts’ popularity, parents should remind their children that interactions on an article or post do not equate to truth or authenticity.