Vine is back…and it’s Divine

gm legends. It’s Sunday.

This week: Vine gets a reboot, how to navigate a tougher climate for pre-seed fundraising, where to get better health insights from your fitness tracker, and a quick comparison of open-weight and frontier models. Plus, five of our favorite launches from the past week. 

Let this newsletter reboot your senses heading into the week, legend. Enjoy.

P.S. Launching soon? We’d love to hear about it → editorial@producthunt.co 🫶

Divine, a reboot of Vine, is now available for download on Apple and Android phones. 

The app restores access to half a million six-second looping videos that were lost when Twitter shut down the archive in 2019. Users can also upload new videos.

While Vine established a whole new category of short-form video content, it had trouble turning a profit because it was just a bit too short. You can’t do a whole lot in six seconds; a good bull ride is eight. Competitors like TikTok and Instagram won out by taking the basic concept and perfecting the length.

But Jack Dorsey, who ran Twitter when it shuttered Vine, has helped bring it back. Last year, Dorsey created a non-profit called “and Other Stuff” to fund open-source experiments for social. Divine, founded by Rabble, is its first project. 

With Divine, people can watch videos, but also make compilations and like/follow other creators. It’s social. 

And it’s not the only new social app that’s launched this month on Product Hunt:

• Couch Critic is a Chrome extension that does what Netflix won’t let you do: give user reviews. You can also insert timestamped comments and emojis.
• LifeOS uses inferences from your Claude and ChatGPT convos to match you with people the app thinks you should meet.
• Morsel calls itself “Strava for cooking”; see what your friends are making.
• Couch is what Zoom would be if Tom from accounting wasn’t there; it’s a virtual room where you can watch movies, play games, and chill with friends.
• Flipboard Surf is part of the socially curated magazine app’s evolution. Its “social websites” feature lets people launch their own sites that pull from both social media and established websites.
• IdeaBoard95 goes retro, giving you a public idea board that looks like a Windows 95 window.
• IndieEvent is a sort-of Meetup for makers, designed for transplants in a new city.
• Genzi, a social app built around shared music, hit the pause button a few years ago and is now back on track.

 

By Piotr Sędzik at Open Wearables

Something we noticed: LLMs are surprisingly bad at health data out of the box.

Give Claude or GPT a week of sleep and HRV data and you get "Your average HRV was 45ms, slightly lower than last week's 48ms." Cool, I can read a spreadsheet too.

The problem: raw LLMs don't reason about health. They don't get that declining HRV + poor sleep + high strain for 3 days straight is a pattern that means something.

So we built the Health AI Engine as an MCP server. Any LLM plugs in and gets actual reasoning tools: trend detection, anomaly flagging, cross-score patterns, personal baseline comparison.

Instead of "your HRV was 45" you get "strain exceeded capacity 3 of last 5 days, recovery down 23%, sleep consistency dropped since Wednesday, reduce intensity, estimated recovery to baseline: 2 days."

That's the difference between a chatbot and something actually useful.

 

By Julia Yu, founder of Unicorns Club

Once upon a time (in fact, as recently as two years ago), early-stage startups needed three things to get into investor conversations: a strong narrative, a solid team, and a prototype. Now, in most software categories, that’s no longer enough.

The advent of AI has shifted investors’ attention to actual metrics. Yes, even at the pre-seed level, investors don’t care so much about promise as they do about realized potential. They want monthly recurring revenue (MRR) and growth rate. (For a deep dive on the data, see my Substack post.)

I’m a cofounder of Unicorns Club, a global founder network that connects startups with investors. And I can tell you the data doesn’t lie. We’ve seen a shift in how our founders get funding. Unfortunately, many founders outside our network focus on out-of-date criteria because they haven’t updated their mental model yet. After reading this post, you will.

 

Hunter fmerian has been looking at the numbers between two recent releases — Kimi K2.6 and Opus 4.7 — and finds that while the latter wins for coding accuracy, the former is well ahead on price. 

So, in a battle between frontier and open-weight models, which one do you give more weight?