Sabitlenmiş Tweet
Solana Ecosystem Call
11.6K posts
Solana Ecosystem Call
@EcosystemCall
Solana's Late Night Show. Comedy. Interviews. Coverage. Last Thursday of Every Month. 2PM UTC.
SOLANA Katılım Ekim 2024
561 Takip Edilen6.1K Takipçiler
The least sexy problem in crypto got solved by @altitude!
Connect your accounting software and every transaction arrives already classified.
Together with Altitude's automated card expense management, the books now essentially do themselves.👀
English
The Cloudmen over at @sanctumso went all-in on mobile last month and launched @sanctumapp!
If you remember Sanctum's legendary Wonderland from back in the day, chances are high, you'll love Sanctum App!
Solana Staking, but cute and gamified. We like!
English
@HOONIGALL these clips hoon is posting as "fun visual" to accompany the actual tweet, are better than 99% of AI video out there
English
Straight from AI video production #14
Already 2 weeks of sharing my workflow and creating a video every single day.
Congrats to everyone who applied these tips 🔥
Generation resolution
720p is the sweet spot. That's my default for EVERYTHING.
Below 720p the quality collapses, faces mush, details die and no upscaler can invent what was never generated.
Above 720p the price per generation explodes and in a workflow where you reroll, iterate and test constantly, that cost multiplies across every attempt.
Here is my advice : iterate everything at 720p until the scene is exactly right then, only if that shot truly demands more, regenerate that single validated scene in 1080p.
You pay premium price once on a shot you already know works
HOON@HOONIGALL
Straight from AI video production #13 Negative prompting. Describing only what you want is good but the model still has a thousand doors left open, and it will happily walk through the wrong one. The majority of video models don't need a separate negative prompt field. You write the negatives INSIDE the prose, right next to what they protect. The rules that make negatives work: - Place the ban right next to the element it protects (The weather is sunny, no rain) - Pair it, positive first and negative to seal it. Use the negative prompt to prevent unwanted outcomes, but also to fix bad gens ✅
English
INTRODUCING THE FIRST SUPERHERO IN THE SQUIG UNIVERSE
Muse - a.k.a. The Needle (la aguja)
@soy_muse
English
Big W for @MeteoraAG last month: their Referral Staking Program is live.
Gotta admit: the team is thinking their stuff through!
You can tell they are extremely clever by simply looking at what they ship: spotless, clean, gud fee tek.
don't forget: rockmakes.money
English
The Solana Ecosystem Call airs every last Thursday of the month! Sign-up and submit your wins on ecosystemcall.com.
English
OH MY GOODNESS! THE MONKES ARE ON THE LOOSE!
@MonkeFoundry kicked off Cohort 2 in July.
@tabdotmarkets, @swishdotcash, @otus_finance, @surfcashx and @nuketrade are sprinting towards perfect PMF, now guided by Monkes!
Congrats!
English
Solana Ecosystem Call retweetledi
Spending the day reflecting on the most recent GTM for @EcosystemCall
Before even looking at the numbers, this theme felt pretty disjointed and didn't have the same cohesion as previous ones.
Lot to learn. Will post my thoughts here once I've gathered them (and run them through Claude so I sound like I know what I'm talking about)
English
Straight from AI video production #13
Negative prompting.
Describing only what you want is good but the model still has a thousand doors left open, and it will happily walk through the wrong one.
The majority of video models don't need a separate negative prompt field.
You write the negatives INSIDE the prose, right next to what they protect.
The rules that make negatives work:
- Place the ban right next to the element it protects (The weather is sunny, no rain)
- Pair it, positive first and negative to seal it.
Use the negative prompt to prevent unwanted outcomes, but also to fix bad gens ✅
HOON@HOONIGALL
Straight from AI video production #12 Report your generations back to Claude. Everybody does the first half.. Your video fails, you go back explain what broke then you get a fixed prompt. That's a normal workflow. But almost nobody does the second half : when a generation comes out PERFECT, tell Claude that too. Why it matters ? Claude has excellent memory within your project, every win you report becomes a locked convention. Do both consistently and something magical happens. You describe a new scene in two sentences and the prompt comes back already carrying every lesson you've validated together. Fixes teach Claude what to avoid but good generations teach it what to repeat ✅
English
English
One of the largest Solana communities in the world is in Ireland, and @SuperteamIE rented a literal castle (@cstldao) for 150 builders in August.
I'll be there for a few days and would love to see you there as well.
Join us: castledao.ie
English
Solana Ecosystem Call retweetledi
Solana Ecosystem Call retweetledi
Solana Ecosystem Call retweetledi
English
types of marketing these companies are good at (imo):
> content distribution & localization: @JupiterExchange
> community led marketing: @solflare
> user acquisition through partnerships at scale (Pudgy card collab): @KASTxyz
> ecosystem & amplification: @MetaDAOProject
> growing through engaging with community members no matter the influence: @world_xyz
> b2b marketing & sleek brand: @altitude by @thenouxnoux
> community programs: @MeteoraAG by @vesper792
> funny team led content: @elfa_ai & @decircusmaster
> product narrative matching brand identity (dinosaurs as childhood lore): @JurassicFi
English
Solana Ecosystem Call retweetledi
