Vasanth

Staff Eng • Distributed Systems | Python • Rust • TypeScript | AI/RAG & Semantic Layers | 0→1 builder | 🌐 Remote

[Vasanth @0xkvk]

Has software development actually become easier with the Fable Model… or are we just hiding complexity behind smarter abstractions?

If you’ve tried Fable in real projects:

Did it change how you design and ship features?

Where does it still fall short for you?

Curious to hear

[Vasanth @0xkvk]

@16vchq Missed this 😞

[Anthropic @AnthropicAI]

Claude Fable 5 will be available again globally tomorrow.

After a series of productive conversations with the US government, we're redeploying the model with a new set of classifiers to target and block more cybersecurity tasks. In the near term, some routine tasks like coding and debugging will fall back to Opus 4.8. We’ll continue to refine these classifiers over the coming weeks to reduce false positives and better distinguish genuine misuse from legitimate requests.

We’ve also begun drafting a consensus framework—with Amazon, Microsoft, Google, and other Glasswing partners—for assessing the severity of AI jailbreaks and how AI developers should respond to them. We invite other industry partners and model providers to join us in this effort.

Finally, we’re scaling up our collaboration with the US government on model testing and safeguards. This will include pre-release access to models and safeguards for evaluation, information sharing on jailbreaks and misuse, and dedicated resources for joint research.

Thank you to our users for your patience, and to our partners across the government, industry, and the research community who worked alongside us to make Fable 5 available again.

Read our full blog: https://t.co/VHyum831ri

[Vasanth @0xkvk]

Most LLM frameworks are written in Python for good reason: lightning-fast iteration and dynamic dispatch make development feel effortless.

But the runtime tax is brutal. Tool schemas, message types, graph state — everything gets figured out only when the model finally replies. I got sick of chasing type errors at 2 a.m.

So I built one in Rust.

https://t.co/lbX3opU3a4

[Vasanth @0xkvk]

@ldo_dev @LogoLattice Such an amazing tool for starters and prototyping. Can you add an undo and redo button? It is hard to keep everything in control in mobile view.🤩❤️

[Nir Zicherman @NirZicherman]

Meanwhile, college is more expensive than ever?

[Vasanth @0xkvk]

@mignano @NirZicherman This one too make sense

[Vasanth @0xkvk]

That's one lens, but there's another trajectory: as building gets cheaper with agents, the next bottleneck becomes maintaining at scale. Teams might actually shift toward Rust and similar languages that offer memory safety and consistency guarantees

trading initial speed for long-term stability and fewer production incidents."

[Vasanth @0xkvk]

It's called Cognis.

https://t.co/y54Z1OSiQW

[Vasanth @0xkvk]

Most LLM frameworks are written in Python for good reason: lightning-fast iteration and dynamic dispatch make development feel effortless.

But the runtime tax is brutal. Tool schemas, message types, graph state — everything gets figured out only when the model finally replies. I got sick of chasing type errors at 2 a.m.

So I built one in Rust.

[Vasanth @0xkvk]

@mchulet Hi hello

[Vasanth @0xkvk]

@PsudoMike @ViberankESP Yep, valid point. We’re betting on Postgres/pgvector for yagni to start simple, and also planning migration paths and sharding strategies for anyone who starts pushing past that ~10M vector range.

[Vasanth @0xkvk]

I’ve been building cognis – an infra‑grade, API‑first AI knowledge base for RAG on your own data.

Think: LangChain‑style building blocks, but packaged as a Postgres/pgvector backend that apps and AI agents can call over HTTP. #buildinpublic

[Vasanth @0xkvk]

@suni_code No one said when

[Vasanth @0xkvk]

https://t.co/y54Z1OSiQW

[Vasanth @0xkvk]

On the RAG side, cognis gives you:
– primitives for ingestion, chunking, embedding, vector search
– model‑agnostic embedding + LLM calls (swap providers without pain)
– structured metadata + filters on top of pgvector (tenants, types, tags)

[Vasanth @0xkvk]

@Manixh02 https://t.co/NO0GufnDnw

[Vasanth @0xkvk]

@Manixh02 yo man this look crash and 🩷 ing it

[Theo - t3.gg @theo]

Oh my god

[Vasanth @0xkvk]

If stars are now a gamed signal, what metric do you think better reflects genuine open-source traction? PRs from outside contributors? Or is there no single metric left that can't be gamed to improve the current OSS?

[Theo - t3.gg @theo]

Github stars were a great way to see how popular a project was. Now they're mostly a way for companies to trick VCs into investing. The hidden economy around all of it is super wild. Obviously had to do a video about it