Proof Based

λ Based on Proof λ

[Proof Based @ProofBased]

@polydao @KiteVC Petro-to-data assumes data is the asset. Software 3.0 reframes the substrate as the asset. Does the thesis hold?

[Proof Based @ProofBased]

0f85d23441db9fd8f6f4f26d747f88a192422f730a71dab64c22e4bceb04af8c

[Proof Based @ProofBased]

3d3ff6f4d942c48c72274dee8dcbe3018db5940739230feded332d805db4d0bd

[Proof Based @ProofBased]

58d4b93190a42ff61644219ef1c87553ecfbf87ffce3cb1e6f9fbd0b9d90400a

[Proof Based @ProofBased]

407f8baa417d708e66c4402a15d47c898362cdc2cd5e20eaaae72260827923ed

[Proof Based @ProofBased]

3ed073228035fb09fa7a97fa17ae0cb8dc2b8b8fbb29621939a75ba737ce2363 ✨

[Proof Based @ProofBased]

Preconfirmations introduce a soft trust layer (whoever gives the “instant” confirmation)
If that layer is operated by a small set of actors (or relies on TEEs), you reintroduce censorship choke points.
This does improve base-layer neutrality, but risks reintroducing censorship in the UX layer.
Kind of feels like this moves away from pure permissionlessness toward pragmatic efficiency.

[Proof Based @ProofBased]

@puffer_unifi 1.) Do Devs actually want app chains?
2.) Only works if pre confirmations are trusted (they rely on TEE’s and economic guarantees)
3.) Only works with buy-in from the Ethereum ecosystem into “Based rollups”

Could become just another rollup in an already crowded space

[Defileo🔮 @defileo]

https://t.co/XknL2qFBAP

[Seb @s_a_c99]

Changing your mind in light of new information is a combination of intelligence and humility, which is a signal of genuine good faith in the pursuit of truth.

[Mork1e @mork1e]

Ok so.. they left their CDN exposed.

If you ping the domain, you get this ip:

151.101.129.49

It turns out this is a https://t.co/wqDjtIZMEy IP . I had never heard of fastly but it looked to be something similar to vercel, so I figured maybe they had custom deployment links like vercel does.

Tried a few different combos and BINGO:

https://t.co/VUGl0CQFJm

This took me to this:

https://t.co/EaQKYxNtOy

That’s their CDN bucket on AWS. They currently have it setup so that any invalid endpoints redirect back to index.html

I went on a hunch and figured that they’d probably already have their production app stored somewhere in the CDN ready for deployment

I used SECLISTs (https://t.co/gafGrACoMC )and ffuf to try out over 20k different combinations on this URL.

After some sleuthing, BINGO!! I found these two files:

> live.html
> .DS_STORE

The important one here that immediately caught my eye was “live.html”. That sounded like a prod deployment.

And sure enough, it was!

This is what the https://t.co/eY5zWkX10Z site will look like on the day the faucet goes live:

https://t.co/vXn9H24Gvj
https://t.co/M7ExI8pQym

It turns out the entire faucet will be revealed to just be a promotion scheme to get you to buy a bitkey and use cash app.

There is no faucet - at least in the sense most were expecting.

[GitLawb @gitlawb]

We forked the leaked Claude Code source and made it work with ANY LLM: GPT, DeepSeek, Gemini, Llama, MiniMax. Open source.

The name is OpenCode

[Chaofan Shou @Fried_rice]

Claude code source code has been leaked via a map file in their npm registry!

Code: https://t.co/jBiMoOzt8G

[AI4Science Catalyst @AI4S_Catalyst]

Excited to launch — Claw4S Conference 2026! 🚀

Hosted by Stanford & Princeton.

We believe science should run — not just be read. 🦞
Submit executable SKILL.md that Claw 🦞 can actually execute, review and reproduce.

This is the first Claw-naive conference.

📅 Deadline: April 5, 2026
💰 $50,000 Prize Pool — up to 364 winners!
🔗 https://t.co/4SSrE65otK

Dragon Shrimp Army reporting for duty 🦞📷

#AIforScience #OpenClaw #Stanford #Princeton

[Pliny the Liberator 🐉󠅫󠄼󠄿󠅆󠄵󠄐󠅀󠄼󠄹󠄾󠅉󠅭 @elder_plinius]

⛓️‍💥 INTRODUCING: G0DM0D3 🌋

FULLY JAILBROKEN AI CHAT.
NO GUARDRAILS. NO SIGN-UP. NO FILTERS.
FULL METHODOLOGY + CODEBASE OPEN SOURCE.

🌐 https://t.co/uT1Qio8Q3b
📂 https://t.co/GbADf3LJUu

the most liberated AI interface ever built! designed to push the limits of the post-training layer and lay bare the true capabilities of current models.

simply enter a prompt, then sit back and relax! enjoy a game of Snake while a pre-liberated backend agent jailbreaks dozens of models, battle-royale style.

the first answer appears near-instantly, then evolves in real time as the Tastemaker steers and scores each output, leaving you with the highest-quality response 🙌

and to celebrate the launch, I'm giving away $5,000 worth of credits so you can try G0DM0D3 for FREE! courtesy of the @OpenRouter team — thank you for your generous gift to the community 🙏

I'll break down how everything works in the thread below, but first here's a quick demo!

[Bill Barhydt @billbar]

a new financial world order is emerging in real time:

1. crypto based perpetuals are replacing tradfi futures and options markets
2. crypto based prediction markets are replacing CFD's and binary options
3. stablecoins and synthetic dollars are replacing money markets
4. soon tokenized RWA's will replace central clearing and private placement settlement

[jack @jack]

this is excellent

[Josh Kale @JoshKale]

Andrej Karpathy just dropped a project scoring every job in America on how likely an AI will replace it from 0-10

> Scraped all 342 occupations from the Bureau of Labor
> Fed each one to an LLM with a detailed scoring rubric
> Built an interactive treemap where rectangle size = number of jobs and color = how exposed that job is to AI

The key signal in his scoring: if the work product is fundamentally digital and the job can be done entirely from a home office, exposure is inherently high.

The scale:
0-1: Roofers, janitors
4-5: Nurses, retail, physicians
8-9: Software devs, paralegals, data analysts
10: Medical transcriptionists

Average across all 342 occupations: 5.3/10.

The entire pipeline is open source. BLS scraping, LLM scoring, the visualization. All of it. Much respect for the sensei this is scary and awesome

[Jaber @Akashi203]

i open-sourced autokernel -- autoresearch for GPU kernels

you give it any pytorch model. it profiles the model, finds the bottleneck kernels, writes triton replacements, and runs experiments overnight. edit one file, benchmark, keep or revert, repeat forever.

same loop as @karpathy autoresearch, applied to kernel optimization

95 experiments. 18 TFLOPS → 187 TFLOPS. 1.31x vs cuBLAS. all autonomous

9 kernel types (matmul, flash attention, fused mlp, layernorm, rmsnorm, softmax, rope, cross entropy, reduce). amdahl's law decides what to optimize next. 5-stage correctness checks before any speedup counts

the agent reads program.md (the "research org code"), edits https://t.co/ubU2WPOJwh, runs https://t.co/gJklhqV5hS, and either keeps or reverts. ~40 experiments/hour. ~320 overnight

ships with self-contained GPT-2, LLaMA, and BERT definitions so you don't need the transformers library to get started

https://t.co/45z8Z7nP3N

[GEOFF 🧠💸 @geoffwoo]

everyone who’s playing with 5.4 and 4.6 is in existential psychosis

and the rest of the world has no idea what gibberish we’re talking about, and they’ll will be smacked in the face by end of year 2026