669 Photos and videos
Built this from zero on Solana:
-Full Governance Engine
-Gamification system
-XP economy
-Quest engine
-Sprint framework
... and so much more
The most complete governance platform in web3 and nobody knows it exists yet.
April 1, that changes.
The trenches are about to get a new way to verify who the real bag workers are.
$ORG π±
Mar 30
Web3 communities are about to level up π
No more dead chats. No more ghost voters.
Every contribution counts, Every member matters.
3 months building in silence. April 1 we go LIVE π
You're either early or you're late.
ORG πΆ
@bonkfun
9
10
45
1,348
Last week Elon become the world's first trillionaireπ€―
While waiting for Fable 5 to come online, I built a page to make you feel how insane $1,000,000,000,000 actually is.
30-second game. Try to get rich. You will fail. It's just math.
Beat my score π
thetrilliondollarpage.com
6
2
20
241
Wanted to post my Fable 5 hands-on verdict. Then i ran /model and realize its currently unavailable.
$200/mo Claude max and it's still behind a gate
Aparently the US government pulled it. national security directive. they think someone found a jailbreak
Anthropic says it's a misunderstanding and they're working on it.
4
1
13
255
Most people using Fable 5 are hitting their limits.
Fable is so overpowered that its low effort already beats opus 4.8 max on most tasks
This is what im doing:
- Fable medium = orchestrator (plans, delegates, reviews)
- Opus xhigh = reasoning-heavy decisions only (or Fable low)
- Sonnet high = execution
- Haiku low = everything else
Same subscription. No codex plugin (you should tho). No second account.
Anthropic 20x Max is for building, not for burning Fable on grep commands.
6
2
17
361
Claude Mythos reproduced more real Firefox exploit PoCs than any other model. That's the chart below.
Rumored pricing: $16/M input nd $80/M output.
One builder did the math at ~$600 to ship a single app.
Vibecoding gave normal people permission to dream. An idea, a few bucks, no degree, no team. I don't want the most powerful tools turning into a club for people who already had money.
Maybe competition saves us. A cheaper Mythos is rumored, and GPT 5.6 leaks point to nearly the same quality for less.
Genuine question. When the best model is this capable and this expensive, who actually gets to build with it?
3
18
511
Been thinking about Mythos all day.
Anthropic built a model that finds thousands of zero-days and scores 72% on expert hacking. they're too scared to ship it. red teamers already have it. public in weeks.
What i keep wondering: what happens to crypto when this lands?
Half of DeFi's "value" is contracts nobody audited hard enough. Point a Mythos-class model at every token on solana and a lot of them don't survive the scan.
Does that kill the casino, or finally make it safe enough to build on?
I don't know. But i'd want it aimed at my own positions first.
3
15
179
that chart is from anthropic's own Mythos preview, not hype. the model before it (Opus 4.6) was under 1%. red.anthropic.com/2026/mythoβ¦
3
73
Since gpt 5.6 was delayed i dont know but..
"claude-oceanus-v1-p" appeared in the console and a checkpoint was handed to red teamers yesterday. Red team access usually starts ~7 days before launch.
Leaked pricing: $16/M input, $80/M output. 1M context.
If the rumors hold, this is Mythos the model Anthropic's own draft blog called "by far the most powerful AI model we've ever developed."
2
2
18
398
I want to start making videos testing AI tools.
Spent an entire day on my first one. a proper honest test of PewDiePie's Odysseus... script, OBS, teleprompter, edit plan, the works.
Full day has passed and im thinking: everyone's already onto new Hermes Agent Desktop, and i'm sitting here with a full day of prep and zero footage.
What's even the point of posting a review now?
Genuinely though, how do people ship a video hours after a tool drops?? baffles me. any tips? π
5
2
22
240
Been playing with Odysseus, PewDiePie's self-hosted AI workspace. Feels unfinished but hey, its open source maybe we can make it better!
Running a blind model comparison right now: same prompt, two local models, side by side, you vote before the reveal.
Making a full video on it. first impressions are way better than i expected for something you run on your own box.
5
2
27
1,182
June 15, Anthropic starts charging for `claude -p`.
I have automations firing it all day. Never tracked it because it was free, why would I?
Added a counter yesterday. The number is embarrassing. I was so wrong about my own usage it's actually funny.
If you build on top of Claude Code: check your usage before June 15. You might be surprised.
How do you track per-call AI costs in production? π€support.claude.com/en/articlβ¦
6
1
17
170
Tough markets don't make me want to slow down. they make me want to help more.
The whole point is to make it worth it for the people who believed first.
Jun 2
The market's been rough lately. doesn't change what we're building.
To everyone still holding: we see you, and we're keeping count.
Starting next week we begin rewarding every wallet that stayed, and opening some of our tools to holders inside Organic Hub.
Anyone can ship in a green market. You find out who's real when it's red. Thanks for being here π
2
2
23
190