6 month old company (not solo; vibe coding means what here??) sold for 80M. Impressive still.

It's a product that offers a vibe coding service to users. You enter a text prompt for what you want to build (no-code), and it tries to build it.

I don't know if these 8 people vibe coded the vibe coding product itself -- they may have. But in the title the reference to "vibe coder" refers to their product.

Isn't this what lovable and that tool provided by vercel?

> vibe coding means what here

It means you tell the product what you want and it gives you the code.

Any model can provide perfect JSON according to a schema if you discard non-conforming logits.

I imagine that validation as you go could slow things down though.

The technical term is constrained decoding. OpenAI has had this for almost a year now. They say it requires generating some artifacts to do efficiently, which slows down the first response but can be cached.

Expect this is a problem pattern that will be seen a lot with LLMs.

Do I look at whether the data format is easily output by my target LLM?

Or do I just validate clamp/discard non-conforming output?

Always using the latter seems pretty inefficient.

You can avoid this by using curl. Half kidding.

Curl, by default, tells sites exactly what version of Curl you are using.

Then pass a random user agent string on each request

The browser is a sandbox with a bunch of discoverable features. Those features exist for the user but a side effect is they leak data which individually is probably not interesting but collectively is a fingerprint.

To be less of a fingerprint you'd need to remove JS from the entire web.

which would be amazing, a good 85 per cent of js on the web is both pointless and useless.

Neat idea. This probably has the disadvantage of coupling deployment to a service. For example how do you scale up or red/green (you'd need the thing that does this to be aware of the push).

Edit: that thing exists it is uncloud. Just found out!

That said it's a tradeoff. If you are small, have one Hetzner VM and are happy with simplicity (and don't mind building images locally) it is great.

For sure, it's always a tradeoff and it's great to have options so you can choose the best tool for every job.

Ever see the military choppers doing their exercise? Amazing and scary.

Yes! We lived in a penthouse apartment in Surry Hills (the old hat factory) for 2 years from 2010 and the first time the choppers came in and dropped troops off on the rooftop opposite I was terrified! Became a regular source of entertainment, to sit on the terrace watching them be dropped off

I worked in that area a while back. It is (was then) a quiet part of Sydney just commercial offices and business that supports it. Despite being a fairly central location. It's a beautiful place to hang out you can walk to harbour views.

Looks like you are putting a derivative behind a paywall though, no? I think quid pro quo let pudiklubi publish your work too? Some kind of open license?

Let alone productionize it! And god forbid maintain it. And have support that doesn't crap out.

Nope. Just that 1. Is better than people. 2. Isn't better than people. Pick one!

If the former then yes singularity. The only hope is it's "good will" (wouldn't bet on that) or turning off switches.

If the latter you still need more workers (programmers or whatever they'll be called) due to increased demand for compute solutions.

> Nope. Just that 1. Is better than people. 2. Isn't better than people. Pick one!

That's too coarse of a choice. It's better than people at increasingly large number of distinct tasks. But it's not good enough to recursively self-improve just yet - though it is doing it indirectly: it's useful enough to aid researchers and businesses in creating next generation of models. So in a way, the recursion and resulting exponent are already there, we're just in such early stages that it looks like linear progress.

Thanks. Your nuanced version is better. In that version I can still ignore most of LinkedIn and Twitter and assume there will still be a need for people. Not just at OMGAD (OpenAI...) but at thousands of companies.