DCHAIN

Data Intelligence, AI, blockchain
Some takes useful, some not so much. Always NFA, DYOR

[DCHAIN @blockAPYs]

@businessbarista Ai native

[Aakash Gupta @aakashgupta]

Boris Cherny built the first version of Claude Code in September 2024 as a side project to see what music he was listening to. No PRD. No Figma mocks. No product review meeting.

By November, 50% of Anthropic's engineering org was using it daily. By February 2026, SemiAnalysis estimated Claude Code was responsible for 4% of all public GitHub commits. They project 20%+ by end of year.

From "what song am I playing" to 1 in 25 commits on the world's largest code platform in 16 months.

The reason this matters for PMs: Cherny's team doesn't write specs. They build hundreds of working prototypes before shipping a single feature. Cherny said there's "no way we could have shipped this if we started with static mocks." When they built Cowork, the non-technical version of Claude Code, four engineers shipped it in 10 days. Using Claude Code to build it.

That's the "taste at speed" framework in practice. Taste means you can look at a prototype and know instantly whether the interaction feels right. Speed means you test that judgment against reality 50 times before lunch instead of debating it in a Google Doc for three weeks.

The old PM workflow was: research, spec, design, review, build, test, ship. Seven steps, eight weeks, one shot to be right.

The new workflow is: prototype, react, prototype, react, prototype, ship. The spec is the prototype. The review is using the thing. The research is watching what breaks.

Every PM skill except judgment gets automated in this model. Strategy documents, competitive analysis, user research synthesis, roadmap decks: all of it becomes commodity work an AI can handle. The one thing it can't replicate is the ability to use a prototype for 30 seconds and know it's wrong.

That's why taste becomes the moat.

[DCHAIN @blockAPYs]

@milesdeutscher I am struggling to justify perplexity computer when I already do the Claude code max tier . How should
I look at this ?

[DCHAIN @blockAPYs]

@itsolelehmann Cowork

[DCHAIN @blockAPYs]

Built a super simple app, a tribute to $TIBBIR
https://t.co/EwbO6SyqDR

[DCHAIN @blockAPYs]

7) Outcome
[https://t.co/qGNRGn8Shg]
It was fun to build. Fun to explore the base ecosystem, more to come *Building 26 experiments in 26 weeks — maybe 52 if my AI skills keep compounding. This was #1. Follow for the rest.*

[DCHAIN @blockAPYs]

1) How to deploy a simple app to Base and Farcaster

I deployed one HTML app to Base, Farcaster

No React. No framework. No build step. Just HTML, CSS, and vanilla JS.
#BASE #Coinbase

[DCHAIN @blockAPYs]

6)
Gotchas I hit, that took me a few hours

1. CoinGecko free tier rate limits or any in general, be respectful with your calls, appreciate the free share
2. Farcaster manifest needs the `accountAssociation` signed by your custody address — follow the Farcaster docs exactly

[DCHAIN @blockAPYs]

- **Volume/MC**: 74.85% - Extremely high daily turnover
- **Volume/Liquidity**: 1.89x
- **Total Liquidity**: $346.56K

*This report presents publicly available on-chain data and is for informational purposes only. It is not financial advice. Always do your own research (DYOR).*

[DCHAIN @blockAPYs]

# Token Scans
$PEON 0x14a35ba26f177a4b6768ee4266065413c995eb96d39091d143a1801a53158982

Started as a humorous/dev tool for AI agent sound alerts (nostalgic gamer vibes + practical use for dev

https://t.co/tyN8qQ6kWW

[DCHAIN @blockAPYs]

The dev/creator (linked to Gary Sheng, ex-Google, Applied AI Society founder, Forbes 30 Under 30) is somewhat doxxed/public

### Key Ratios

- **MC/FDV**: 100.0% - Nearly all tokens in circulation
- **MC/Liquidity**: 2.5x - Very deep liquidity relative to valuation

[DCHAIN @blockAPYs]

Time to learn structured thinking

Thought of the day - focus on results

[DCHAIN @blockAPYs]

Token scans

Blowfish
https://t.co/xyFBfk7pRu

- **MC/FDV**: 100.0% - Nearly all tokens in circulation
- **MC/Liquidity**: 10.6x - Healthy liquidity depth
- **Volume/MC**: 24.88% - Active daily trading

On-chain data, not financial advice. DYOR.

[DCHAIN @blockAPYs]

@jessepollak @baseapp @jessepollak - will the app allow trading across chains ?

[DCHAIN @blockAPYs]

@garrytan Can you share a hint on what you are trying to build ? Looks very interesting

[DCHAIN @blockAPYs]

@alanhoward Wise words. I am planning to explore, good to review it with this lens