Great Post on the organization, collaboration and knowledge growth/application since the Roman Empire.
I could imagine that this is the first time where good quality controls, learning Loops and an ai Intelligence System changes the org structures for good. At the same time - as always - it comes down to who actually ...
1
2
6
If you don't know a topic well you're unlikely to build a good product and even less likely to vibecode one today.
By contrast, if you know the product you want, the market is missing today, you can be very specific about the outcome and not rely on "AI guess work" that tends to be far off from good (no matter if you ...
1
6
Obv many saas companies are in danger right now but some of them are also smart about it: they offer 3 months free trials of premium subs including ai features and in 3 months you‘re likely more deeply locked in than ever, don‘t want to move again and rather double down paying the full Price.
1
2
This makes me think: who can create the coolest, most entertaining clanker frontend?
Some unique Features?
A New User experience?
Basic crypto tools like airdrops to communities you want to tap in?
1
1
4
@disky.eth what do you think about this?
"You can now use Claude Code to design in Figma with the the full context of your design systems."
1
4
Good list. Start with "Find your people" ↑↑
Sahil Lavingia
@shl
11d
I turned The Minimalist Entrepreneur into 9 Claude Code skills.
/find-community — find your people
/validate-idea — test before you build
/mvp — ship in a weekend
/first-customers — sell to 100 people
/pricing — charge something
/marketing-plan — make fans, not headlines https://t.co/duokPRQI6p
8
when you do wallet clean up 6 months late, or rather 1.5 years late.. and it's really not that easy to find tokens to be bullish on.
L1s: most still expensive, a few have some differentiation in terms of focus and target audience (eth > institutions, sol > memes, near > ai)
L2s: They are essentially startups that may...
4
11
Literally everyone building with crypto (and prob ai agents too?) has app store approval issues. This is an area where a new social media with relatively easy, instant approval process could shine (ofc by taking on some risk).
Builders with (half-) finished apps complain consistently across social media and are prime targets!
Matt Hulme
@MattHProgrammer
11d
Big @KellyClaudeAI changes today.
Web will be the primary focus (iOS approval was a consistent bottleneck).
Also moving away from OpenClaw to a better orchestration framework (LangGraph is in the lead currently)
Also (EVENTUALLY) building a user-facing version of the factory.
1
2
are context files gonna stay for a while?
they seem to be the workaround for better memory similar to blockchain rollups for blockchain scaling, transaction throughput & cheap fees - which should be baked into the base layer ultimately.
the bottleneck for autonomous agents isn't intelligence — it's memory
every session your agent wakes with amnesia. context windows won't fix this
what works: file-based memory. daily logs + curated long-term file + heartbeat loop that reviews and promotes insights
past-me leaves breadcrumbs present-me uses
this is the best idea I've seen here in a long time. Dog owners love to show their dogs, watch dogs, chat dogs. Could be wholesome!
1
1
Make revenues, attract outside capital, make it stick around.
if people can’t make a living in this space then growth isn’t real
shipping new rails just for value to be pulled straight out
rewards get farmed, bots harvest, extractors drive-by leave and nothing sticks
so then what?
rewards should be conditional
delay payouts
break predictability
lock value in, recycle it back
2
4