👾 Web3 believer | 🧪 Always experimenting....Red devil....OP top 1
Joined November 2024
- Tweets44.5K
- Following2.1K
- Followers1.3K
- Likes377K
M1CHA3L retweeted
Earlier this year, @Aptos holders voted to make their own chain 10x more expensive to use, and outside the ecosystem it barely got any attention. Which is a shame, because what happened afterwards goes against how most of us think about fees.
Staking rewards were cut in half in the same package. Normally you’d expect activity to slowly bleed out after a move like that, but daily transactions actually doubled, going from ~4.5M to ~9M.
Fees were so low to begin with that a median transaction is still among the cheapest of any major chain, so users barely felt it.
The burn did, though. All gas on Aptos gets burned, so the same activity now takes a lot more aptos:native out of circulation than it used to.
I hold aptos:native myself, so I’ll be straight about it. The network keeps getting better and the tech has honestly never been the problem, but the price just hasn’t caught up with any of this yet.
That gap is exactly why I’m paying attention now. With issuance lower and investor and team vesting ending this October, usage is finally showing up on the supply side instead of only in the transaction count.
@blockworksres calls this a path to deflation, and I like that they didn’t oversell it. aptos:native isn’t deflationary today and the burn still covers only a small part of issuance.
The new activity is led by Decibel, the fully onchain perps order book incubated by Aptos Labs. Placing, adjusting and cancelling orders all hit the chain separately there, so a dollar traded on Decibel burns far more aptos:native than a dollar going through a regular swap.
The breakeven point in Sam Schubert’s report sounds far away at first, but it’s not unrealistic next to what onchain perp venues already trade daily. The chain has also handled much heavier load than that, so the real question is usage.
That’s why the number I’m watching from here is how much of issuance the burn covers each month. It’s where that usage shows up first, and it has already come a long way since March.
I’d rather hold a token whose supply story depends on people actually using the chain than on an unlock calendar, and for the first time aptos:native feels like that kind of token. If the burn keeps catching up with issuance, I don’t think the market can ignore it forever.
Long term, I’ve always been optimistic on aptos:native, and this gives me even more reason to stay that way.
Where $APT tokenomics stand after the updates first proposed in February.
Approved through onchain governance:
• Gas fees set 10x higher
• Staking rewards halved from 5.19% to 2.6%
• Supply hard-capped at 2.1B APT
Per @blockworksres, burns now cover ~10% of issuance, up from 0.2% under the previous gas rate, and daily transactions doubled after the change.
M1CHA3L retweeted
Our striker scored. Microbiology wanted the goal ruled out.
Last month at UNILAG, Biochemistry played Microbiology in an interdepartmental football match. We were already celebrating when their players shouted handball.
They insisted the ball touched our striker’s hand before he shot. We saw a clean goal. Same play, opposite accounts.
Now imagine the prize money sat in a smart contract.
It could hold the funds and pay the winner, but it couldn’t replay that moment. Someone would have to feed it a result.
That’s an oracle. The payment follows code, but the deciding fact comes from outside.
@GenLayer’s Intelligent Contracts can check accessible evidence instead. Written in Python, they can read live web pages and use a language model to interpret what they find.
If match footage and the rules were online, one validator could run the contract and propose a ruling. Others would run it independently with their own models, like separate referees reviewing a replay. Disagreement brings a new leader; anyone can appeal.
Even a room full of reviewers can be fooled by a bad camera angle.
Neither department had to be lying. We simply trusted our own eyes.
If five independent validators called handball but Biochemistry still swore the goal was clean, what evidence would change your mind?
M1CHA3L retweeted
The model I’d try first from LFM3 is the vision one
LFM3-VL-2B averages 71.4 across 62 vision benchmarks, on par with models 6x its size and it now takes video as input and reasons step by step
Open weights, runs on device, fine tunes on one GPU
🤝 Paid partnership
Introducing 𝙻𝙵𝙼𝟹: our new generation of Liquid Foundation Models. With a new tri-hybrid architecture and new multimodal encoders, the LFM3 family is the ideal model suite for specialized intelligence.
> 6 model instances, across text, audio, vision, and omni modalities, generative and decision models.
> 𝙻𝙵𝙼𝟹-𝟷.𝟽𝙱: text, 256k context window.
> 𝙻𝙵𝙼𝟹-𝚅𝙻-𝟸𝙱: vision-language model with video-input support for spatial understanding
> 𝙻𝙵𝙼𝟹-𝙰𝚞𝚍𝚒𝚘-𝟸𝙱: tool calls, voice chat, reasoning in 18 languages.
> 𝙻𝙵𝙼𝟹-𝙾𝚖𝚗𝚒-𝟸𝙱: an omni model combines vision, audio, and language.
> 𝚍𝟷-𝟸𝙱 and 𝚍𝟷-𝙾𝚖𝚗𝚒-𝟸𝙱: decision models with typed outputs and calibrated probabilities in a single pass. Omni adds audio input.
> fine-tune on 1 GPU.
> a new tri-hybrid architecture
> architecture optimized for 14 processor classes across GPUs, CPUs, and NPUs.
> 𝙻𝙵𝙼𝟹-𝚅𝙻-𝟸𝙱 is a beast. It performs competitively with Gemma 4 12B (6x larger), and Gemini 3.5 Flash Lite, and GPT-5.6 Luna, with 1.9B parameters.
> 𝙻𝙵𝙼𝟹-𝟷.𝟽𝙱 is #1 on the Pareto frontier of device intelligence 𝚙𝚒𝚙𝚎𝚝𝚝𝚎.𝚕𝚒𝚚𝚞𝚒𝚍.𝚊𝚒
Every modality for specialized intelligence. Pick the one that fits your use case, customize it, and run it on-device.
1/ 🧵
M1CHA3L retweeted
Just tried Cannonball on @Duelbits and the multipliers are wild. Aiming for those hoops feels like the move right now
Who’s already hitting big?
💣 𝗖𝗔𝗡𝗡𝗢𝗡𝗕𝗔𝗟𝗟 𝗚𝗜𝗩𝗘𝗔𝗪𝗔𝗬 💣
$500 up for grabs. 10 winners. ⚔️
Our new Original is live, and we're giving away $50 to 10 lucky winners! 👀
𝗛𝗼𝘄 𝘁𝗼 𝗘𝗻𝘁𝗲𝗿:
✅ Follow @Duelbits
👥 Tag a friend and drop your User ID
💣 Tell us your best Cannonball strategy
10 winners will be randomly selected. Good luck! 🍀
M1CHA3L retweeted
how do gmgnai contribute to trading platforms?
jesse was asked to describe it details the usefulness
check out what he said below and the importance of gmgn to the trading world and what they do
FT: @jessepollak x @gmgnai x @base
M1CHA3L retweeted
Grvt 2.0, Turning Yield Into Trading Power
What if your capital didn't have to choose between earning yield and supporting your trading activity?
That's the idea behind Yield as Margin, one of the features planned for @grvt_io in November.
The concept is straightforward, eligible Grvt Earn balances could be used as trading margin while continuing to generate yield.
Instead of keeping earning and trading completely separate, Grvt wants to connect them within one self custodial ecosystem.
For me, this is one of the more interesting parts of the Grvt 2.0 roadmap. Capital efficiency isn't just about finding more opportunities. It's also about making better use of the capital you already have.
Of course, the feature's actual mechanics and eligible assets will matter when it launches.
With the first 1.1M ethereum:0xad29f2723fcdbcf665f210f25e06f97477e417cf buyback and burn already completed, the next thing to watch is how Grvt turns its broader vision into a working product.
M1CHA3L retweeted
Already completed Level 1 KYC on InterLink? According to the Day 82 announcement, you will not need to repeat the process in this round.
Eligible users can proceed directly to the waiting list for the distribution of transferable ITLG. The amount received will depend on account metrics, so completing KYC does not mean everyone will receive an identical allocation.
This is an important distinction for the community. Verification establishes eligibility, while account metrics help determine distribution.
✓ Check your account status and follow official updates instead of assuming you need to submit KYC again.
Know your status before taking action.
#InterLink #ITLG #ITL
M1CHA3L retweeted
the ability to stop is an underrated part of an AI trading strategy.
a model can identify opportunities all day, but without clear limits, one bad stretch can turn a promising strategy into a serious risk.
that’s what makes the risk controls behind @agenticscredit interesting to me. stop losses, tighter limits as drawdowns increase, and trading halts when hard thresholds are breached put risk management directly into the execution process.
it’s a different approach from simply displaying warnings and hoping the agent responds correctly.
i think the real challenge for autonomous trading isn’t just making agents more capable. it’s making sure they operate within boundaries, especially when market conditions turn against them.
more trades don’t always mean better decisions. sometimes, the smartest move is to stay out of the market.
This post shares project information from a personal perspective and is not investment advice.
🤝 Paid partnership
M1CHA3L retweeted
Bro just casually dropped a yearly salary in one spin on jack.com 😭 @N3onreload
M1CHA3L retweeted
In this conversation on @gmgnai Live, @jessepollak CEO of @base shares his perspective on the current state of crypto, highlighting how onchain sentiment has shifted and why builders must keep pushing forward despite market uncertainty.
His message is clear: market hype comes and goes, but lasting growth comes from building useful protocols, solving real problems, and creating content that captures people's attention.
Instead of focusing only on token prices, the real opportunity lies in building things that bring value to the onchain ecosystem.