PixelPoseidon🔱🦈 retweeted
Network difficulty on TensorCash reached a new all-time high today.
More AI inference is running to secure the chain than ever before, and every block still comes with a proof of the model that produced it.
PixelPoseidon🔱🦈 retweeted
Verifiable inference is the cornerstone for an open and decentralised market, what we are building with @inferenceFI.
@Tensorcash protocol is already delivering verifiable inference at scale - great to see attention on the topic growing.
New Anthropic research: progress on provable inference for large language models.
Here, we introduce FlashProve. Building on established cryptographic techniques, FlashProve allows developers to verify whether a given response came from a given AI model. It’s substantially more efficient than prior open-source methods for doing so.
These proofs have a range of applications. They can, for instance, make it harder for attackers to hijack a chat, or enable safety evaluators to verify that the model they’re testing is the same one being served to customers.
For our full write-up, see here: anthropic.com/research/prova…
Further progress in provable inference will likely be an important part of improving trust in AI deployment. To that end, we're open-sourcing FlashProve’s code, here: github.com/anthropics/FlashP…
We’re grateful to our collaborators: Alessandro Chiesa, @GiacomoFenzi, @gwberg95, @miranda__christ, Rahul Dalal (rdalal.github.io) and Ben Mayer (@transmissions11).
For everyone interested in tensorcash:native here is a very simple explanation
A short version of how everything around @Tensorcash fits together:
@Tensorcash is the base. A proof of work chain where the work is AI inference, and every answer carries a proof of the model that produced it. On top sits a financial layer - native assets, settlement, compute derivatives, a coordination layer for trading directly between parties, and a DEX in research. Native assets, compute derivatives and the coordination layer are already live on mainnet.
@inferenceFI is building the market on top of that verified output - order books for LLM tokens, with its index values published on chain.
@AskTillyAI is where people will use it - a simple AI app for chat and agents, coming online later.
From the GPU to the user, all on the same rails. We’ll publish a full article soon on how it all connects, and what comes next.
1
3
88
Very relevant for tensorcash:native
nitter.cf/a16z/status/2113986385…
Zooming out, GPUs are still renting ahead of a standard 6-year depreciation schedule
Rental price vs. schedule:
- H200: +143%
- B200: +111%
- RTX PRO 6000: +73%
- H100: +54%
- A100: +9%
Most of the gap opened up after April, when token demand took off
Charts of the Week: a16z.news/p/charts-of-the-we…
3
93
People are looking for excuses to sell non stop
Brothers, have some faith for once
You're in this industry for it's upside
Why are you acting like such pussies then when it goes up
Plenty of room for being a beartard when things turn bearish
Until then grow a spine
PixelPoseidon🔱🦈 retweeted
Lots of work went into this. But it’s finally time to put this out to the world.
Robinhood mainnet this Friday
The floor only rises.
Robinhood Chain is a casino right now. $CASHCAT, ethereum:0x07f5b6823751c2e2cd4560f28af75ff887102241, $INDEX, $SLVR, $HOODRAT, $DIH — they print volume, they print charts, they print exit liquidity. Then they don’t.
$SZR is the opposite design.
Sazare is a Uniswap v4 hook market and a launchpad. The factory is permissionless. Anyone can launch. The difference is what you launch into:
Strictly monotone redemption floor
1B lifetime mint cap (burns never reopen it)
1% fee that compounds the floor
Permanent burns
No admin. No upgrades. Immutable.
Price can fall. The floor cannot.
Why this beats the rest of Hood L2:
$CASHCAT / $HOODRAT / $DIH — narrative + LP. When sellers show up, the book is the floor. $SZR sellers walk into a rising redemption ratio. Tokens burn. Remaining holders get thicker bedrock.
ethereum:0x07f5b6823751c2e2cd4560f28af75ff887102241 — launchpad fee flywheel. Cool until the next Noxa. Sazare launches inherit the same floor, warehouse, and cap. Creator fees don’t just sit in a treasury; they feed buybacks, burns, and the rising floor.
$INDEX — tax-and-distribute RWAs. Cute. $SZR can use those RWAs as settlement. Imagine a community token whose floor is measured in $NVDA or $SPY. Every trade locks more stock under the price.
$SLVR — inflationary mine. Sazare flips that: mining add-ons route rake into buybacks, burns, and a higher floor. Inflation becomes deflation.
Stock tokens ($NVDA, $SPY, $AAPL) — exposure with no memory. Sazare is speculation with memory. Every exchange is recorded into the reserve.
New $SZR only enters through a rising issuance curve. Every token pays its own backing. Redeem at the floor → tokens die forever. Volatility warehouse eats dumps and recycles them on the way back up. Iwa (NFT liquidity) sends 10% of revenue to $SZR buy-and-burn.
You don’t just launch a ticker. You launch a market that remembers every trade.
Small stones become bedrock. That’s the name.
CA: 0xbeD7c1660a01dbEEe225bd893ac31727a91A5773
sazare.io
NFA. Read the hook. Watch the reserve. The rest of the chain is still asking the chart for permission. $SZR already wrote the floor in code.
PixelPoseidon🔱🦈 retweeted
One of the most advanced projects but people are slow to it as they look for DexScreener as a source of truth.
It has its own AMM, hence DexScreener info is completely off. Knowing and checking that is what gives you edge on seeing the project early.
PixelPoseidon🔱🦈 retweeted
One of the most important aspects of network effects is that they become more valuable as time goes on and as they grow.
Gaining access to this system creates opportunities that would otherwise not be available. The network becomes its own value proposition. This is inherently true in many different fields.
The unified perps pools and additional activities they will provide are no different, which is why access to such things should be priced properly.
Launchpads face a very interesting dilemma, where you want more launches in order to increase volume and fees.
But SZR is not like most other launchpads. The network itself will have value that others simply do not offer, alongside the tech and development that will continue to come.
Because of this, launching on SZR should not be free. Each launch will require the creator to burn SZR.
It will be a small fixed amount, as in the early days of the protocol the value proposition will not be fully displayed, and creators should not have to pay large sums in order to participate. But as the protocol grows and becomes stronger, the cost should also reflect the increased barrier to entry.
It automatically filters for quality over quantity. Just from the numbers alone, less than 1% of standard launchpad coins graduate.
This inherently hurts the space. Too many rugs, too much attention being split, and developers spending far too much time fighting for attention amongst garbage slop rather than actually building out what they need to build.
It does not, and should not, have to be this way.
We create the network, the network accumulates more and more value over time, and developers can gain access to this value for a price.
Many small stones turning into one powerful object. Together we win.
A really good article about $SZR worth a read
nitter.cf/hardriver/status/20932…
前面三篇推文,已经将 $SZR 的渐进发行、交易带、地板储备和 Add-ons 拆解过一遍。
这次 @SazareProtocol 连夜最新公布了 Warehouse Perps。
表象上是给 Sazare 加了一套永续合约,实际最真正创新的是:
原本只能负责托底的“地板储备”,开始变成可以支持杠杆、赚取资金费、增加流动性和回购销毁的工作资本。
━━━━━━━━━━━━
先说普通永续合约的问题:
交易者开杠杆
→ 平台收手续费
→ 价格下跌触发清算
→ 清算代币被砸回市场
→ 玩家亏钱,币价继续受伤
交易量看起来很热闹,但并没有养活代币。
Sazare 想把这条路反过来。
假设某个 Sazare Token:
市价 1.10 美元,地板价 1 美元。
有人开了一个价值 1100 美元的多单,系统借出资金买入 1000 枚 Token,并把它们锁在仓位里。
因为这 1000 枚 Token 最低可以按地板赎回 1000 美元,所以仓库真正需要额外承担的风险只有:
1100 - 1000 = 100 美元
简而言之:
100 美元风险资金,可以支撑一个 1100 美元的仓位,资金效率是 11 倍。
如果市价距离地板只剩 1%,资金效率理论上可以达到 101 倍。
不是说玩家可以无风险开 101 倍杠杆,而是:
价格越接近地板,仓位中已经被真实资产保障的部分越多,仓库需要额外承担的风险越少。
地板第一次从“价格保护”,变成了真正可以工作的抵押物。
━━━━━━━━━━━━
更创新的是亏损怎么处理。
普通多单亏损:
触发清算
→ 代币被强制卖出
→ 市场承受额外抛压
Sazare 的设计是:
亏损仓位中的 Token
→ 按地板价格赎回
→ Token 永久销毁
→ 资金池得到偿还
→ 玩家承担亏损
→ 市场却没有遭遇强制砸盘
翻译之后:
普通杠杆交易把亏损变成砸盘;Sazare 试图把亏损变成销毁。
玩家照样会亏钱,仓库也会消耗风险储备,没有免费午餐。
但亏损不再变成市场上的新抛压,剩余供应反而减少。
━━━━━━━━━━━━
交易者支付的资金费,也会继续留在系统里:
交易者支付资金费
→ 增强储备和永久资金池
→ 支持回购销毁
→ 资金池可以承接更多仓位
→ 更多仓位继续产生资金费
而且,它不依赖外部预言机和传统清算线。
只要预付资金费还够,短时间插针不会立刻触发机器人清算。资金费耗尽后,仓位先冻结,玩家还有时间补钱、减仓或者退出。
相比多单用地板价值保障债务,另一方面空单则由 Warehouse 借出 $SZR 库存,做空者以后必须买回并归还。
开空带来卖压,平空带来回补买盘,持仓过程继续给协议贡献资金费。
━━━━━━━━━━━━
把整个循环放在一起:
地板保障仓位
→ 仓库只承担市价与地板之间的差额
→ 更少风险资金支持更大仓位
→ 交易产生资金费
→ 资金费增加储备、流动性和回购销毁
→ 系统获得更大交易容量
所以这次的升级,不仅仅是“$SZR 增加了合约交易”, Sazare 第一次把:
发币、地板、储备、杠杆、收入和销毁,组装进同一套经济系统。
过去,永续合约只是寄生在代币旁边赚钱。 @SazareProtocol 想让永续合约变成代币自己的“代谢系统”:交易带来收入,收入增加流动性,亏损减少供应。
当然,目前这仍是设计方案,还没有正式部署。
宣传中的数百倍、上千倍,首先指理论资金效率,不等于玩家可以无风险使用上千倍杠杆;“没有清算”也不等于不会亏钱。
方向确实非常性感和创新,但最终能不能跑起来,还是要等代码、参数和真实资金上场后再看,这么小的市值,我愿意等。
1
16
444
The funny thing about hyperliquid $hype is that it has real cash flow and is the only crypto project valued as a real business. So in a sense it is the first crypto project with meaninful product adoption.
52
PixelPoseidon🔱🦈 retweeted
I am the only one responsible for all my massive failures and all my massive wins. One has to take total responsibility, learn, process, and keep going
If you feel bad about missing out, trust me, you are not alone and should keep going, we all had massive setbacks👇
nitter.cf/Innerdevcrypto/status/…
Mistakes & failure: you are not alone
People think that somehow massive drawdowns in net worth, disappointments, failure, mistakes, only happen to them, but the contrary is true, it happens to all of us and is a crucial part of learning in this space
Shared this before, but just so you realize again that you are not alone in learning hard lessons, making a lot of mistakes, and seeing large amounts of net worth evaporate, all part of the game, here my personal list of failures and mistakes: it is so long i do not even remember all, but here we go:
- cryptopia collapse: big amount gone: lesson: never all in one exchange
- holding 20k verge (xvg) all 2017 and selling a couple of days before it pulled x50 on pornhub collaboration announcement: lesson: never sell all of a coin when early and you have conviction
- forgetting bittrex password and spending months trying to get in, had money on there: lesson: use written paper and encrypted usb backups
- selling doge too early: lesson: always keep some
- had offer to do the bnb seedround, did not do it: potential many millions in gains missed: lesson: pay attention when great offers happen
- ftx collapse: got 90% out on time, but 10% was stuck when sl was hit after longing ftt: lesson: withdraw at first sign of trouble and do not go long an exchange token when exchange collapses thinking it cannot happen
- large amount staked on osmo, took 21 days to unstake, huge losses when market turned: lesson: never stake with long unstaking time
- lightning struck mountain next to my home and all electronics fried: lesson: always keep paper backup of passwords and seedphrases
- accidentally did a massive pepe long on bybit with the x1000 ticker, did not pay attention, woke up with massive loss, 6 figures+: lesson: always double check what and how much you buy
- last year, couple of seedrounds done and investment in alternative energy: all failures basically: lesson learned: never too much in ´fancy´ things and do not do seedrounds too late in a cycle
- huge amount on anchor and luna, luckily had it unstaked on time: lesson: move serious profits to bank and not crazy defi platforms with insane interest
- potential 10M-100M+ Trump gains faded. Knew about the launch early, fell asleep, still got in early but could have been way more: lesson: if you know something will happen, drink coffee and make exception
- CT account grew and more physical threats and people trying to geo-locate: lesson: anonymity is super important and be careful sharing data
- overall my drawdown amount from potential peak net ATH worth is much larger than amount in total gains, which is still great considering i started with 10k. lesson: gratitude instead of thinking what could have been
so you see, you are not alone in making a lot of mistakes, all part of the game. most important thing is lesson learned, feet on the ground, and keep going, space is full of opportunities
Replying to @Innerdevcrypto
Sucks to be sidelined but chasing the market now can really end up badly. If this is the start of a new bullmarket there will be plenty of pullbacks to buy in. What this move shows us is that the crypto market still has legs and people with huge amounts of cash are looking at it, so it's still relevant regardless of all the AI developments and capital rotation. That alone is a huge confirmation for me.
60
PixelPoseidon🔱🦈 retweeted
Replying to @Innerdevcrypto
Sucks to be sidelined but chasing the market now can really end up badly. If this is the start of a new bullmarket there will be plenty of pullbacks to buy in. What this move shows us is that the crypto market still has legs and people with huge amounts of cash are looking at it, so it's still relevant regardless of all the AI developments and capital rotation. That alone is a huge confirmation for me.
PixelPoseidon🔱🦈 retweeted
Replying to @pixel_poseidon @SatoshiFlipper
Been in this game since 2014. You need to trust the process. Crypto and blockchain is just going to get bigger and bigger each year. It is not going away. Anytime there is a crash, just close your eyes and come in to buy the top 1 or 2 asset with size, forget about it for 2 years . It is that simple. Don’t over complicate it.
Recently read Trading in the Zone by Mark Douglas, recommended by @BobLoukas . What a great read. Here are my notes on how it applies to the crypto market:
nitter.cf/pixel_poseidon/status/…
1
74
























