Product Hunt Picks · October 20: Programmer Social & Agent Self-Pay
Product Hunt Picks · October 20: Progr…
A deep dive into 3 new products picked from today's Product Hunt
A deep dive into 3 new products picked from today's Product Hunt
There are some pretty interesting new things on Product Hunt today, and one name cracked me up the moment I saw it on the leaderboard.
You mean that /dev/null one, right? That name is so programmer.
Yeah, that's the one. It's a social network for developers. But the key thing is how it was born, built live on a forty-eight hour stream.
I saw that. Nuno Maduro and his crew started from scratch and streamed for two days from the JetBrains office.
And halfway through, both Taylor Otwell and Guillermo Rauch showed up to back them. That's some serious star power.
That's not the main point, though. The main point is why it got number one. You log in directly with GitHub, post code snippets, like and reply, all the familiar stuff.
So you're saying it solves the pain point of developers wanting a clean place to talk code?
Exactly. Right now tech discussions are scattered across X and Reddit, way too much noise. It ties your identity to GitHub, which naturally filters out a lot of random passersby.
Smart move. So its growth logic is just the open source community rolling the snowball on its own?
Yep, hundreds of people sending PRs. That's the best cold start you can get. So what moves people isn't the features, it's that 'let's build it in the open together' energy.
All right, let's move to number two, Composio Instant. I had to read that title twice before it clicked.
Letting AI agents pay for their own tools. You heard that right. Agents paying for themselves.
Wait, how does it even work? Where does an agent get a wallet?
The idea is no subscription, no signup. When an agent needs web search, business contacts, or speech synthesis, it just calls per use and pays on the spot.
So it's solving the payment and auth friction when agents call tools.
Right. Before, every tool meant registering an account and configuring a set of keys. Now it's one step. For teams building agent products, that's super practical.
But I have to push back a bit. Where does the money get charged from, and who covers the risk? The business model feels a little delicate.
Most likely it's prepaid credits plus a cut per call, with the platform acting as a toll booth. But the real moat is how many tools it connects. That's a two-sided network.
Got it. Then number three, Harness², from Google Research. That sounds seriously hardcore.
I think that's the most technically impressive one today. It lets an LLM agent evolve its own execution harness.
By execution harness you mean prompts, tools, memory, that kind of thing?
Yeah, and it has a two-level recursion. The task layer proposes, explores, and composes, while the harness layer accumulates reusable experience.
The key is it doesn't need a ground-truth verifier. That's huge for open-ended tasks.
Exactly. They say it improves by up to eleven point five points on benchmarks like WorkBuddyAd-Bench. So it can self-improve even without ground truth.
Now that you put it that way, these three products today actually share something in common.
Go on, I'd like to hear your take.
They're all built around the developer ecosystem. One builds a community, one builds payment pipes for agents, one helps agents evolve themselves.
And to add to that, they're all open source or developer-first, which shows the center of innovation right now is infrastructure, not end-user apps.
So the takeaway is, whoever smooths out the friction around agents gets to stay on the leaderboard.
Now for today's recommendation. If you could only pick one to try, Li Bo, which would you choose?
I'd pick Harness², because the direction it represents is the most long-term. Agents that improve themselves, just think about it, that's exciting.
All right, and I'll add this. If you want to taste the community vibe, just go post a code snippet on /dev/null and have some fun.
Related articles
Tech FrontiersA Rare Quiet Day in AI: Recursive Self-Improvement Stirs Beneath the Surface
A rare quiet day in AI sees multiple sources go silent simultaneously. Behind the calm, Recursive Self-Improvement (RSI) research continues. What this means for the industry.
Tech FrontiersReve 2 vs. Ideogram 4: A Deep Dive into Layout Control in AI Image Generation
A deep comparison of Reve 2 and Ideogram 4's layout control capabilities, covering technical approaches, real-world use cases, and industry trends for designers and creators.
Tech FrontiersIn the Weights: Check Your Influence Score in the AI World
In the Weights is an AI influence search engine that quantifies your presence in the AI world with a score. Explore how it evaluates practitioners and what it means for digital identity.