Jake

Helping people leverage data to reach their goals

[Jake @jakesdata]

Everyone is making functional representations. And those func’ repr’s with the most causal power are the target

[Jake @jakesdata]

Protocol = policy = gov’t++

[Jake @jakesdata]

Tribe v2 to see how my videos effect the brain. Maybe use for editing. This is wild

[Jake @jakesdata]

The most meta wins, with bounds on causal power at a given time. But ideas are antifragile

[Jake @jakesdata]

The world is a Markov chain and life is a Markov blanket, question mark

[Jake @jakesdata]

@jack I feel like “prompts” should be in there. “Essays will be shorter in words but longer in meaning.”

[Jake @jakesdata]

Tokens. NFTs. 🤔

[Jake @jakesdata]

Everyone needs their own TEE

[Jake @jakesdata]

The most meta wins

[Jake @jakesdata]

I built an AI/Python workflow app on IPFS https://t.co/WguvwEG9n2

[Jake @jakesdata]

how i built boost-ipfs https://t.co/F9yIFKmfxa

[Jake @jakesdata]

@balajis @a16z @a16zcrypto @peterthiel

[Jake @jakesdata]

just open sourced boost-ipfs — encrypted reactive state on IPFS with post-quantum crypto

entities subscribe to each other and run python triggers when upstream data changes. chain-agnostic, swappable between local and Base L2

https://t.co/WguvwEG9n2

[Jake @jakesdata]

As apps like ClawdBot/Moltbot/OpenClaw, N8N etc become more useful, there will be an increased demand by consumers for personal data control. E.g. people will want a Spotify API key. This is a force that drives decentralization/de-siloing of data (and therefore established moats)

[Jake @jakesdata]

It won't be long before we can vibe code within existing apps to modify them. It's easier for startups to do this because it's pretty hacky engineering -- it's harder for established companies, especially because it would move engagement metrics like crazy.

[Jake @jakesdata]

Boost Productivity: Human+AI Life OS — boost mode vs. vision mode

[Jake @jakesdata]

They need to be specific about the a/b testing during ranking. Which metrics are prioritized over others in model decisions

[vitalik.eth @VitalikButerin]

If done properly, this is a very good move.

I hope it can be verifiable and replicable (probably means publishing all tweets and anonymized likes with 4 week delay).

This would not solve all problems, but it would very effectively address concerns about algorithmic transparency that I and many other members of the public have been raising.

I actually think 4 weeks may be over-ambitious, because it means the algo would need to change that often to stay ahead of people gaming it; hence why I proposed 1 year in my recent post advocating algorithmic transparency. But this does really need to be verifiable and replicable though, so that someone who thinks they've been shadowbanned, deboosted, etc would be able to walk through the code executing the algorithm and see why their posts are not being seen by people.

https://t.co/Hlzhn78Bot

[Jake @jakesdata]

Underrated: researching consciousness (as seriously as non-expert could)
Why?
-Models of information integ.; modeled as "causal power" of neurons (attention economy-related, see Koch)
- Likely reliant quantum ( Hameroff, Penrose)
- Not fully understood, but effects all