Grok isn’t the best AI system on the market, but it might be the most situationally useful one for news/talk radio hosts, producers, and programmers. That sounds like a contradiction. It isn’t, and here’s why.
Every AI system seems to have its own personality, if that isn’t a dystopian word for it. ChatGPT can be a bit of a workhorse. Claude is more thoughtful, chatty, and down-to-earth. Google Gemini prides itself on its research capabilities, and it’s a great generative photo and video creator.
And then there’s Grok. If you didn’t know, it’s Elon Musk’s baby, and he’s poured plenty of time, effort and (most importantly) money into it. Frankly, the results aren’t great. Still, you can harness Grok to your advantage if you work in news/talk radio.
Grok is competitive, which is an interesting personality trait for an AI system. In my experience, that competitive streak pays off when you ask for a star-based ranking. Grok wants to give you the best results possible, so it loves to rank them.
News/talk radio lives on timing, credibility, and conversation. Grok, with its direct pipeline to X — the platform Musk also owns — can surface all three faster than a producer can scroll.
The Prompt That Puts Grok to Work
Like every other AI tool, what you put in determines what you get out. However, how you tell Grok to shape those results matters just as much. I use it to scrub X for content, for instance. If you’ve read a Barrett Media story that includes an X post, there’s a good chance I found it with this prompt:
Score every topic on these five criteria. Use only 0, 0.5, or 1 star per category. Borderline cases default to the more favorable score.
- Recency
- ≤ 24 hours old → 1 star
- 24–48 hours old → 0.5 star
- Older than 48 hours → 0 stars
- Credibility
- Accounts operated by major news media figures → 1 star
- News media-focused X account not affiliated with a legitimate outlet → 0.5 star
- Random person → 0 stars
- If a story is confirmed by a major source (or writers who work there — check bios), it qualifies for full credibility.
- Engagement
- Discussed by multiple credible sources and generating meaningful comments, reposts, quotes, etc. → 1 star
- Engagement mainly among non-credible/random X users → 0.5 star
- Little to no significant engagement → 0 stars
- Heat
- Breaking news involving a major talent or brand, hirings or firings, contract extensions, programming changes or clear hot-button conflict → 1 star
- Involves important topics but not top-tier or national brands → 0.5 star
- Routine update with little conflict or emotional charge → 0 stars 5. Shelf Life
- Likely relevant up to a week or longer → 1 star
- Probably lasts a couple of days → 0.5 star
- Likely dies within ~24 hours → 0 stars
Turning Stars Into Show Prep
Moreover, the star system sorts the noise for you. A five-star topic is fresh, credible, loud, hot, and durable. Consequently, a producer can scan a ranked list in two minutes instead of scrolling for an hour.
So, how can you use this to your advantage? If you’re a host or producer, start with the talent you already plan to discuss. Maybe that’s someone on MS NOW, or personalities like Don Lemon and Joy Reid. Have Grok search specifically for things they’ve said or content they’ve produced. Their highest-engagement posts are likely what your audience cares about, so ask for them.
Meanwhile, programmers can build a quasi-geofence. Ask Grok to highlight the most trending topics in your city, market, or state. What’s drawing the most attention? What’s flying under your radar? Those answers can shape a morning rundown, a promo strategy, or a daypart overhaul. Local relevance wins in news/talk radio, after all, and X chatter can show you what your market’s really talking about.
Twitter Isn’t Real Life, But It Still Matters
These are just a couple of options, and I know “Twitter isn’t real life” is a legitimate argument. Still, X is where news figures break stories, trade shots, and shape the conversation. Besides, a ratings book can’t tell you what happened at 7 a.m. today. A well-built prompt can.
Of course, verify everything. Grok can misread a post or surface a rumor, so confirm each story with a second source before it hits air. In other words, treat Grok like an intern with great scouting instincts.
Ultimately, more information is rarely a bad thing. Grok might not be the best AI system out there, but with the right scorecard, it can earn a spot in your prep routine. Give it a try, then tell me what you find.
Barrett Media produces daily content on the music, news, and sports media industries. Sign up for our newsletters to stay updated and get the latest information right in your inbox.
