Kaize1 GRASS retweeted
A self-evolving agent + a 428B model + 3 Macs = ?
Your own AI lab.
We ran @MiniMax_AI M3 locally with @tryParallax, right on our desk.
Then @GA_agent_ai took over to create a 5-stock portfolio and write it to disk.
No cloud. No API bills. Nothing left the machine.
Wild to see a ~3K-line agent drive all this with a 400B+ model on local hardware.
Thanks to the GenericAgent and MiniMax teams for making local AI feel real.
Kaize1 GRASS retweeted
🌍 The first gate to Vinci World is opening.
Pre-registration goes live now:
🌐 vinciworld.xyz
Be among the first users to enter Vinci World — a gamified interactive browser for next-gen collecting, powered by real-world assets, social participation, and intelligent agents.
Powered by @renaissCLTB.
🎖️Early pre-registrants can unlock the Vinci World Genesis SBT.
The world is loading.
Step in early.
I just claimed my $SLIME Early Supporter spot on @smapocke 🟠
Soul-bound. Limited slots. No utility — just the proof that I was here first.
smapocke.com/slimesbt
9
Cute slimes + free mint on Solana? I'm in. Love the early supporter angle, let's see what utility the $SLIME premium NFT actually brings 🔥 Applied 📷
Introducing $SLIME
Please check the website for how to participate in the AirDrop and TGE. slime.sptokens.com
27
Kaize1 GRASS retweeted
Introducing $SLIME
Please check the website for how to participate in the AirDrop and TGE. slime.sptokens.com
Kaize1 GRASS retweeted
Introducing the $SLIME premium NFT.
Supply: TBA
Network: $SOL
Mint price: Free
Applications are now open:
➔ smapocke.com/slimesbt
More details Coming Soon.
Season 1 of @The_Beacon_GG goes live May 25 alongside $BCN!
Use my code and get a 50 Gobloonz + 1 Umbra Chest head start on everyone else 👇🏼
app.thebeacon.gg/pre-registe…
3
Kaize1 GRASS retweeted
DAWN is kicking off Q1 2026 with strong momentum as we look towards ecosystem activation.
Here’s what’s actively taking shape across the network:
• Expanded network transparency and visibility
• New deployer & subscriber surfaces
• Hardware, mobile, and desktop tooling progressing together
• Core protocol systems moving toward public participation
• Continued coverage expansion across the network
Stay tuned...
Kaize1 GRASS retweeted
As we kick off the new year, here’s a refresher on how DAWN works, and how we are making user-owned Internet a reality.
Watch below to see how you can get involved 👇
Kaize1 GRASS retweeted
DAWN has raised $13M in Series B funding, led by @polychain
This capital accelerates our global expansion, new deployments, and ecosystem partnerships as we continue our mission of scaling decentralized broadband worldwide 🌍
Kaize1 GRASS retweeted
On 3DOS, every time a factory prints your file in some random city, you get paid.
Not from views, not from likes — from objects crossing a workbench.
Kaize1 GRASS retweeted
Costs for home internet service vary greatly around the world. We pulled a few results 👇
Just 31.84% of people worldwide in our survey pay less than $20 a month for their service.. Data centers offer internet at extremely low prices, often as cheap as 30 to 40 cents per gigabit per second.
DAWN solves this gap by using wireless technology to bypass expensive monopoly markups and provide access to affordable internet.
Kaize1 GRASS retweeted
$631B in 30-day perp dex volume.
Another record month.
Stork is the oracle for perps.
Kaize1 GRASS retweeted
Quote-tweet this over “crypto has no real adoption”:
3DOS routes real orders to real machines to make real products.
Designers, printers, CNC shops, buyers.
No memecoin needed to prove this matters.
Kaize1 GRASS retweeted
Reply “3DOS” if you want more deep dives on:
• How designs become on-chain manufacturing assets
• How local factories can plug in as nodes
• How decentralized manufacturing eats legacy supply chains
Builders + DePIN nerds, this rabbit hole is for you.
Kaize1 GRASS retweeted
We’re excited to partner with @LMSYSOrg’s SGLang to bring LLM serving to everyone. 🔥
Parallax is a high-concurrency inference framework that lets anyone host LLMs of all sizes across heterogeneous GPUs and devices — local or remote, anywhere in the world.
SGLang @LMSYSOrg powers Parallax’s per-stage GPU inference backend, providing key serving primitives such as continuous batching, prefix/radix KV cache reuse, and optimized kernels, while Parallax manages global scheduling, layer placement, and pipeline orchestration.
On Apple Silicon, Parallax integrates MLX with our own batch scheduler and dynamic KV management behind the same unified interfaces.
SGLang’s modular design, clean pipeline-parallel hooks, and explicit activation handoff align perfectly with Parallax’s distributed, pipeline-parallel architecture — keeping throughput high and tail latency steady at scale.













