> What if Lovable's mission became: Pay out $1 billion to vibe coders.
This reframes everything. Customer success means revenue success. Platform improvements mean better monetization tools. Growth means more builders making more money.
LOL.
Companies want their revenue numbers to "line go up" and if that means success for the customers, that's a nice side-effect. It is absolutely not the goal, however.
Paying for outcomes is a nice idea but you can make WAY MORE MONEY selling picks and shovels to people who you know will never strike gold. (Plus, you can have a nice little side-hustle selling them maps of where treasure might be).
LLM companies are incentives for the opposite. They don't want you to get the best outcome as quickly (or with as few tokens) as possible. They want to keep you on the slot machine. They want to maximize token use while giving just good enough results to not drive you to a competitor.
"Oh man, you're so close! If only you weren't out of tokens. Too bad you settled for the 20 dollar plan instead of the 200. One more prompt would do it, then everything would work perfectly"
It’s basically a royalty model. That’s common in some industries and with some products. I haven’t looked lately but both Unity and Unreal Engine had royalty models; game devs would pay either a fixed per-unit fee or a percentage of revenue after a certain volume of sales.
To be viable as a business plan, this requires that a certain percentage of your customers have viable products.
Here’s the thing though: anyone who has a high volume of sales will want to shed the royalty. This could be by negotiating different terms or just rewriting to avoid the component or service that wants the royalty.
For Unity and Unreal, it’s pretty common knowledge that AAA studios have separately negotiated licenses, presumably to reduce or eliminate the per-unit royalty. Some studios write their own engine, though that has its own costs.
For vibe coding I have real doubts about this model. There’s effectively no moat and no defensive IP (ie: patents), so anyone making enough revenue to pay $$$ on royalties will probably end hiring SWEs to rewrite their software to avoid royalties.
The difference is obvious: it doesn't cost Epic anything if you download their engine flail around for 5 years and release a buggy bomb. 5 years of tokens would cost a lot.
> What if Lovable's mission became: Pay out $1 billion to vibe coders.
This reframes everything. Customer success means revenue success. Platform improvements mean better monetization tools. Growth means more builders making more money.
LOL.
Companies want their revenue numbers to "line go up" and if that means success for the customers, that's a nice side-effect. It is absolutely not the goal, however.
Paying for outcomes is a nice idea but you can make WAY MORE MONEY selling picks and shovels to people who you know will never strike gold. (Plus, you can have a nice little side-hustle selling them maps of where treasure might be).
LLM companies are incentives for the opposite. They don't want you to get the best outcome as quickly (or with as few tokens) as possible. They want to keep you on the slot machine. They want to maximize token use while giving just good enough results to not drive you to a competitor.
"Oh man, you're so close! If only you weren't out of tokens. Too bad you settled for the 20 dollar plan instead of the 200. One more prompt would do it, then everything would work perfectly"
Once defined as tools, the outcome depends on skill of using the tools ...
This is easy to say but not always easy for the majority to do if they haven't priced according to value or outcomes before.
Would he extend the same deal to his customers? He won't charge them unless they make money, then he just gets a percentage of that?
Like where does that end?
It’s basically a royalty model. That’s common in some industries and with some products. I haven’t looked lately but both Unity and Unreal Engine had royalty models; game devs would pay either a fixed per-unit fee or a percentage of revenue after a certain volume of sales.
To be viable as a business plan, this requires that a certain percentage of your customers have viable products.
Here’s the thing though: anyone who has a high volume of sales will want to shed the royalty. This could be by negotiating different terms or just rewriting to avoid the component or service that wants the royalty.
For Unity and Unreal, it’s pretty common knowledge that AAA studios have separately negotiated licenses, presumably to reduce or eliminate the per-unit royalty. Some studios write their own engine, though that has its own costs.
For vibe coding I have real doubts about this model. There’s effectively no moat and no defensive IP (ie: patents), so anyone making enough revenue to pay $$$ on royalties will probably end hiring SWEs to rewrite their software to avoid royalties.
The difference is obvious: it doesn't cost Epic anything if you download their engine flail around for 5 years and release a buggy bomb. 5 years of tokens would cost a lot.
Yes. Some companies I take equity and become advisors.
That could also be contingency, if you take a percentage of the improvement or savings.