The tooling is totally replicated in open source. OpenCode and Letta are two notable examples, but there are surely more. I'm hacking on one in the evenings.
OpenCode in particular has huge community support around it- possibly more than Claude Code.
> This is how infrastructure works, and supposed to work
No, infrastructure doesn't have to work this way. This is a very old-school mentality.
Sign the content with a key that you control. Back up the content locally. And boom- your server is easily replaced. It only helps copy data around and performs certain conveniences.
I've been working on this full-time for a few years. If we succeed, we solve link rot (broken links) on the web.
Dumb question: whats the difference between "low-code" and "libraries+frameworks"?
Usually the point of a library or framework is to reduce the amount of code you need to write. Giving you more functionality at the cost of some flexibility.
Even in the world of LLMs, this has value. When it adopts a framework or library, the agent can produce the same functionality with fewer output tokens.
But maybe the author means, "We can no longer lock in customers on proprietary platforms". In which case, too bad!
LLMs are a game-changer, but they’re only half the story: probabilistic and fuzzy. The missing half is a universal formal language: something precise enough to translate cleanly between human languages and let anyone communicate with computers, without learning programming!
My dream is to remove barriers everywhere: culture, science, medicine, law, diplomacy. You don’t erase ambiguity, you encode it! Dialects, jargon, puns, inside jokes, social context.. we can build everything in.
Maybe this isn't possible, but now that we have language models to help... it might be!
Your alternative is... what exactly? A unique and baroque file format for each application (see: Git)? Folders of JSON or markdown files which are slow, easily corrupted, and lack indexing? Depend on some memory-heavy external DB service like Postgres?
In most cases, embedding SQLite is the best solution. And that is exactly what it was designed for.
Thanks for spawning many interesting topics. A dose of cynicism is great, in moderation!
> P2P is cool if you have a desktop, but you cannot host from laptop or phone that spends most of the time sleeping (unless you want your battery to die real fast). The solution is hosting providers - which are already decentralized (and federated, if you squint hard enough)
Yes, most people will rely on servers because phones are terrible p2p nodes. When identity is properly owned by the end users, the servers have nearly zero lock-in, unlike traditional hosting providers. A community's server can go down for some reason and the community can easily transition to other server(s), keeping their conversations and knowledge intact. Sadly this is not the case with Mastodon or even Bluesky.
> _Cryptographic_ identities have huge problem of it's own - there are many people who don't have any persistent data on their PC
This is probably the single biggest problem we are facing, because it impacts UX. There are several tools available to mitigate this issue, but I don't believe there is a perfect solution. Keys can be linked across devices with cross-signing, there are mechanisms that can enable key rotation: DNS, social media connections, and social/manual rotation in the worst case. The plan is to leverage existing tools that are used to keep secrets safe for regular people: system keychains, password managers, passkeys, smartphone "wallets".
> Turns out, other than piracy, there are no legitimate uses. The existing technologies are good enough.
People become very comfortable in their virtual prisons, and most people won't change unless they have a reason to. Maybe they have legitimate work or content that is stigmatized and censored by other platforms. Maybe they live under an autocratic regime. But I think most people want better control over their content moderation and feed algorithm.
> People has been proposing those things forever. No one needed them back then, and no one needs them today.
I'm not laughing at your exaggerated use of "no one". Decentralized and censorship-resistant technology is society's fail-safe. Maybe your social media oligarch isn't abusing their power too much today. Maybe your government actually supports free speech today. What about tomorrow, the next decade, and the next century?
The solution is to build on the traditional web. How does anybody find anything new on the web? Basically: hyperlinks!
People will create links from social media. With some basic SEO, your content can be indexed by your favorite search engines. Increasingly these "web4" sites will link to themselves, leveraging the built-in social features that are portable across sites/servers/peers.
As somebody working in this "future-web" space, I see HUGE issues with the legacy web stack:
- It requires a server to publish, which is expensive and difficult for regular users with a laptop or a phone. This can be solved with a mix of p2p and federation
- There is no decentralized trust system- only DNS+HTTPS, which requires centralized registration (TLDs). A domain may be cost-prohibitive for somebody who just wants to write comments and a few documents on the web. This can be solved by forming a social graph of cryptographic identity validations (aka, the "web of trust")
- There is no versioning system. This can be solved by making chains of immutable signed content, like we do with git.
- There is no archival system that allows you to "back up" the content of a website in a trustless way. Look at IPFS and BitTorrent for the solution there.
I believe these are the main reasons the web has failed as a social publishing system. Aside from companies and technically skilled individuals, everyone publishes on centralized social media platforms. This is a dangerous consolidation of power.
We hate to admit it, but the open web has taken the "L". The good news: these are solvable problems and I'm not giving up anytime soon!
> Honestly there kinda is a new web, they call it web 3 and it's only crypto scams.
To distance ourselves from crypto scams, we strongly avoid the web3 label, despite some similarities.
Utilizing p2p tech is not illegal. It is illegal to redistribute copyrighted content without authorization- and we are working to build this into the protocol so that peers will respect copyright by default. People can redistribute at their own risk. I'll be the first to admit that this is complicated, and we have a long way to go in this regard.
Plus, the vast majority of people will just use the web frontend, with a peer on the server. Most peers can be hosted by content creators and tech-savvy friends+family.
I get freaked out when I consider the future of archive.is. Thanks to the nature of the web today, it is incredibly fragile.
As the co-creator of a censorship-resistant publishing platform, I really wish we would migrate to a peer-to-peer technology. We could develop network effects on a decentralized platform with a cryptographically-provable network of trust. Most people don't realize it is possible to handle media distribution in a robust way.
I'm not just trying to shill my solution! I wish there were more competitors using these techniques to try and save the web.
I'm confused how you would "build something on ChatGPT" in the first place. Does ChatGPT have an API?
Of course OpenAI's GPT-5 and family are available as an API, but this is the first time I'm hearing about the ability to build on top of ChatGPT (the consumer product). I'm guessing this is a mistake by the journalist, who didn't realize that you can use GPT-5 without using ChatGPT?
> You must not use any Output relating to a person for any purpose that could have a legal or material impact on that person, such as making credit, educational, employment, housing, insurance, legal, medical, or other important decisions about them.
If somebody forks this project, I dare you to name it "Open Open Web UI". If they threaten you, just rename to "Open WebAI", "Open UI Web", and other permutations, until their legal budget runs dry.
Clearly this company is following OpenAI's playbook- start with lofty OSS ideals, put "open" in your name, then fall down a slippery slope.
And sometimes the best way to contribute is by forking. Depends on the situation for sure, but this encourages an acceleration by the maintainers of the original library. If they don't have the time to do so, they can comfortably ignore the fork.