p.enthalabs

Fable open sourced NanoClaw's agent factory. It cost $800

twitter.com · Read Story HN original

Comments

Since its behind an account wall, I can't tell if they are saying that Opus made the clone for cheaper or that Opus failed? It appears that Fable did far more "work" based on output tokens.

Aside: We should just ban Twitter links. Nobody should be required to make an account to view content on here.

On the aside, where's the difference between someone posting a Twitter link and an Economist link? Both are locked behind a wall, whether it account-wall or pay-wall.
We can ban those too.
if we would ban everything that is not 100%: free, ad-free, tracker free

what would we be left to read? couple of snarky remarks and maybe 2-3 paragraph of coherent text.

journalism is a job! money has to come from somewhere

Well, restricting posts to those actually accessible isn't unreasonable. Allowing them is HN's worst policy, in my humble opinion.
The Economist isn't owned by someone who cheats at video games, subverts the government, and thinks Nazi salutes are funny.
Same with medium. Why?
It's really unusual that developers would rather write behind a twitter authwalled link than to write to their own site and just tweet about it.

If you can spend $800 on Fable in a day, you can launch a fantastic looking/performing personal site in 15 minutes ¯\_(ツ)_/¯

Perhaps Hacker News could ban this style of link to help encourage that?

But then you couldn’t engagement-farm like Twitter /s
I guess if you enjoy engaging with an echo chamber
nowdays people get their reading on social media.

social media, also started heavily demoting posts with external links.

thus if you want reach, you need to post on platforms...

I’m confused. The post is a long description of merging one branch into another in a clean way, done by agents, at a scale never seen before. Bravo.

Does it do anything now that it didn’t before?

Who’s this for?

Why did it matter?

Is the world better now?

Unfortunately the post doesn’t discuss outcomes so I’m having a hard time seeing how the team’s $800 was best spent on this vs. some other priority.

This reads like AI-generated slop, flagging.