rajivayyangar

this was cool to see (disclosure: I'm a small investor because I've worked w/ Andriy, a cofounder). Really high design bar. A team of agents working like a design studio on your behalf. BYO models.

I've been searching for a term for what differentiates certain tools, workflows, or startups that improve models versus those that degrade or hold back the models. I am really stoked on the term "buoyant". I wrote a PG-style blog post where I broke down the metaphor of buoyancy and how to think about different solutions.

The results strike me as comparable or worse than you could get with a Markov model. I think it reveals the gap in understanding between LLMs and music. I think you need to either: - Set up a pipeline to decompose music into, say, harmonic sequences and melodic sequences, and then have the LLM work on some more fundamental or more high-level layer of musical composition and then re-translate it back into actual sounds. - Develop a better dataset and train the LLM more natively on musical examples.

Does anybody know of a project that has produced more convincing results?

It's heavily inspired by a past product - Wisp (super collaborative docs) - which inspired my last startup: Tandem (YC S19) - a virtual office for remote teams.

(I'm the CEO of Product Hunt) This isn't far off the mark, but I do want to push back on two things.

One: "VC backed companies have the funds to be able to boost their launch on PH to guarantee their launch will be in the top 5." - absolutely not. We do not tolerate paying for votes or other means of artificially boosting score.

Two: "its going to be a lot harder for you, the side hustler, the soloprenuer, the full time bootstrapper" - It's true that the bar is higher, but it's because The Market is more competitive. It was a personal goal of mine when I stepped in a year ago to make Product Hunt into a place where you can launch your side projects and they can rise to the top. To do that, we need to reduce the noise a bit and ensure that community and organic signal are balanced. This is difficult, but we're working on it. For example, Chris Van Pelt's side project OpenUI (open source v0) was #2 of the day a little while ago - hunted by a friend without any launch prep.

Product Hunt changed my life - I launched a side project that got tons of traction and it caused me and my friends to jump in full-time and start a company. I want to make Product Hunt an even stronger force for changing lives than it's been in the past, but that requires evolving the platform for the market in 2024.

Cool - simple, to the point. Love the "now" vs. "later" suggestions.

Saving, shared queries, easy link sharing, and basic charts with some intelligence. It also has excellent query UX (auto-complete, etc.).

Most of all, it's really fast. It fundamentally depends on your DB of course, but PopSQL doesn't add any extra bloat the way Mode does.

Funny - one of our teammates used to do this at a previous company. We built in a similar function for remote workers with a Slack integration. It sends a little message with pre-formatted responses that indicates you'd like to talk on Tandem.

The maker-manager dilemma is real - which is something we try to address with the prominent Focus mode. At the same time, a big reason we show the app status is so that the "business owners" can hopefully take cues and minimize their interruptiveness. For example, as a product manager, I will often decide not to bother my co-founder Tim because I see he's in a coding editor (and thus in deep-focus).