Colin
/colin10188
Stream of consciousness thoughts about crypto, big tech, cooking, startups, or anything else on our collective minds. No topic is off topic.
Spent the past two weeks off-the-grid in Argentina for a wedding.
Looks like a lot's been happening on FC!
- starter packs
- new token meta
- anoncast meta
- frames v2
What else did I miss?
Looks like a lot's been happening on FC!
- starter packs
- new token meta
- anoncast meta
- frames v2
What else did I miss?
Spent the last two weeks traveling for back-to-back weddings. One in Toronto, and another in Hawaii.
Always great to disconnect, but eager to get back to work.
Always great to disconnect, but eager to get back to work.
📊 Matrixport：比特币九月意外走强，投资者看好十月潜力 https://www.chaincatcher.com/article/2143428
Sad I couldn’t make it to the dev day but thankful I can follow along with @dylsteck.eth’s live casting!
❌ optimizing for qDAUs
✅ optimizing for qBangers
✅ optimizing for qBangers
One thing I like about caster rewards in particular: no claiming.
Post good content and receive value straight into your wallet - no other action needed.
Post good content and receive value straight into your wallet - no other action needed.
Finally, native rewards for bangers
I think this is a very interesting problem space. It’s similar to some work I was doing at Google: labeling & predicting abusive behavior for millions of SMS messages across 1b phone numbers a month.
We would:
- manually label millions of spam reports
- create heuristic rules & train ML models using labeled data
- deploy heuristics & models to the device and on the server (bc different access to signals on each)
- rinse & repeat
Heuristics got us surprisingly far.
The vast amount of accessible onchain and FC data (metadata, behavior, content) should make it much easier to build decent classifiers.
We would:
- manually label millions of spam reports
- create heuristic rules & train ML models using labeled data
- deploy heuristics & models to the device and on the server (bc different access to signals on each)
- rinse & repeat
Heuristics got us surprisingly far.
The vast amount of accessible onchain and FC data (metadata, behavior, content) should make it much easier to build decent classifiers.
Curious to see how onchain identity plays out.
We now have a few different namespaces, with arguably not much current interoperability between them:
- ENS
- Farcaster
- base names
- XMTP
- others (unstoppable, lens etc)
We now have a few different namespaces, with arguably not much current interoperability between them:
- ENS
- Farcaster
- base names
- XMTP
- others (unstoppable, lens etc)
Provides multi-platform synchronization function to allow users to stay socially connected anytime and anywhere.
Discover the new standard of secure social media with blockchain.
Wow, can’t believe how spot on this frame is
—-
Colin-kun's waifu pillow is a giant Paragraph logo, uwu~ He snuggles it while dreaming of Farcaster integrations, nya!
Create a Waifu based off your FC profile with WaifuCast!
—-
Colin-kun's waifu pillow is a giant Paragraph logo, uwu~ He snuggles it while dreaming of Farcaster integrations, nya!
Create a Waifu based off your FC profile with WaifuCast!
Blockchain social networking makes me feel the authenticity and purity of social networks.