游侠矿工｜不凑热闹｜只玩冷门｜争当 OG Vibe Mining 社区：https://teapaw.org/t.co/n8ovpWXdYn 请我喝咖啡：0x6fedcf4b0ef79a025eba12cd8c040a10788c4177
Moon
Joined October 2023
- Tweets4.7K
- Following6.4K
- Followers6K
- Likes7.7K
Pinned Tweet
这几天 tensorcash:native 回调确实挺狠。上 SafeTrade 后涨了一大轮，现在有回撤也不奇怪。至少目前，我看好它的理由没变。
我的想法很简单：对我这样的普通人来说，有机会和前高盛、摩根士丹利的MD级高管，在如此早期共同创业参与，可能是一辈子难遇的机会。
以后我还是会不定期分享早期/冷门的 PoW 项目，这点不变。同时，我也会拿出一部分精力，长期参与 TSC 的建设和推广，毕竟这是我自己为这个周期准备的最大 BAG。干对退休，赌错重来，就是这么简单。
如果先把技术放一边，只聊人。
与聪明、牛逼、认真做事的人同行，是我理解的成功捷径之一。
TSC 的重要生态支持者之一 AskTilly，由前高盛董事总经理 @enrico_otta 与前摩根士丹利执行董事 @samjavadi 联合创立。
这也是我愿意深入研究 TSC 的原因：一个早期项目的潜力，往往也藏在选择参与、持续建设的人身上。
早期研究，识人也是识机会。 tensorcash:native
100ⓧ 观察 retweeted
I built POW Valuation Lens so when I hover on any SafeTrade candle it shows me MC (at that day's supply via official stats) + FDV.
Used @AskTillyAI on Qwen 27B (Tensorcash miner) then built and finalized in Codex. Receipt covers one coding response.
SafeTrade has a pretty interesting track record of being early on assets like $TAO, $KAS, $TSC, $QTC, $QUANTUS, and $PRL.
now $BNT is entering the picture.
people love chasing coins once they’re already on everyone’s radar, but the real opportunities are usually found before the crowd catches on.
worth paying attention to what’s trading on @SafeTrade.
$CSD finally pulled a worthy move, next runner to be listed on @safetrade could most likely be @blocknetprivacy - $BNT? 👀
100ⓧ 观察 retweeted
🚀 $BNT × @SafeTrade — Another Step Closer! ❤️
Great news for the Blocknet community! 🎉
✅ Technical integration completed.
🔄 SafeTrade developers are working to bring the BNT node online.
🧪 Final testing is expected once the node is running.
A huge thank you to the SafeTrade team for their hard work, dedication, and continued support! ❤️
The BNT community truly appreciates your efforts. Every step brings us closer to the first exchange listing!
We're excited and eagerly looking forward to the official launch! 🤝🚀
⏳ The wait may soon be over!
#Blocknet #BNT #SafeTrade #Privacy #PoW
100ⓧ 观察 retweeted
tensorcash:native aims to go beyond proof-of-useful-work by turning AI inference into both a market and a way to secure a blockchain....
@enrico_otta explains how @Tensorcash differs from other protocols by using the same proof of actual LLM inference to enable a decentralized AI inference market while powering the network's proof-of-work mechanism
Rather than proving an operation used to run AI, the protocol uses proof tied to the inference itself, combining two functions in one system
“TensorCash is kind of more of a superset of what a simple proof-of-useful work chain is doing.”
Today on MCG
$TSC | @Tensorcash w/ @samjavadi ,@enrico_otta
Tensor Cash is the verification and settlement layer for global AI inference
Highlights include:
00:58 – Introduction to Enrico and Samad
03:53 – AI compute is becoming a commodity
06:23 – Model substitution and quantization problem
07:36 – A commodity needs verification
08:53 – What Tensor Cash is
09:50 – L1 where proof of work is inference
11:03 – Bitcoin turned electricity into money
15:48 – Verification by replication
17:27 – Cheating caught: FP8 passed off as FP16
19:57 – Aims to be an open protocol
22:14 – InferenceFi: order book vs OpenRouter
23:57 – Equity in InferenceFi, contributors to Tensor Cash
26:47 – Live: frontier model mined, billed in USD
28:35 – Next: marketplace and flywheel
30:16 – Receipt for every inference
30:48 – How to mine
33:24 – Closing pitch vs overpriced IPOs
34:20 – Host takeaways
100ⓧ 观察 retweeted
tensorcash:native aims to make AI inference verifiable so it can power a decentralized market...
@Tensorcash lets participants prove that an AI model processed a prompt and generated its answer, enabling trustless payments for AI compute while using inference as proof of work to secure a blockchain
“What we're trying to do with TensorCash is get the same unit of electricity to do both at the same time.”
Today on MCG
$TSC | @Tensorcash w/ @samjavadi ,@enrico_otta
Tensor Cash is the verification and settlement layer for global AI inference
Highlights include:
00:58 – Introduction to Enrico and Samad
03:53 – AI compute is becoming a commodity
06:23 – Model substitution and quantization problem
07:36 – A commodity needs verification
08:53 – What Tensor Cash is
09:50 – L1 where proof of work is inference
11:03 – Bitcoin turned electricity into money
15:48 – Verification by replication
17:27 – Cheating caught: FP8 passed off as FP16
19:57 – Aims to be an open protocol
22:14 – InferenceFi: order book vs OpenRouter
23:57 – Equity in InferenceFi, contributors to Tensor Cash
26:47 – Live: frontier model mined, billed in USD
28:35 – Next: marketplace and flywheel
30:16 – Receipt for every inference
30:48 – How to mine
33:24 – Closing pitch vs overpriced IPOs
34:20 – Host takeaways
100ⓧ 观察 retweeted
Live soon with @samjavadi.
Questions on @Tensorcash are more than welcome.
teapaw.org/i/broadcasts/1aKbdEjkj…
100ⓧ 观察 retweeted
33,000 blocks of tensorcash:native in 14 seconds 🫧
Watch miners, pools and the SafeTrade listing shape the network, wallet by wallet.
57% of all mined TSC has never been sold.
👉 map.tensorcash.boblabs.eu
100ⓧ 观察 retweeted
Great to see the community contributing real-world benchmarks on TensorCash.
Mining adds almost nothing to the cost of serving a request. So one GPU, one request, two revenue streams: the block reward, and verified inference a customer will pay for.
An independent benchmark by @boblabs_eu showed that mining on TensorCash adds at most 0.3% GPU time to inference.
Qwen3.8-27B ran on three GPUs over a full agentic session, once without mining and once with every request mined.
GPU time changed by −1.1% to +0.3%.
At close to zero overhead cost, a miner can serve inference together with a proof of the model that wrote it, and compete for the block reward at the same time.
ALT Independent benchmark. TensorCash mining adds close to zero GPU time. GPU time added by mining over a full agentic session: RTX PRO 6000 Max-Q −1.1%, B200 +0.3%, H100 PCIe +0.2%. Measured 21–23 October 2026. Source: Bob Labs.
100ⓧ 观察 retweeted
gpus are the scarcest thing in ai right now. so a side job can't cost much.
an outside team, @boblabs_eu, ran a full ai agent session on @Tensorcash with every request mined. it took at most 0.3% more gpu time.
so a gpu can keep serving ai and mine blocks on the side. each answer comes with proof of which model wrote it.
i expect that to make mining an easy yes for anyone already running ai models
$TSC
An independent benchmark by @boblabs_eu showed that mining on TensorCash adds at most 0.3% GPU time to inference.
Qwen3.8-27B ran on three GPUs over a full agentic session, once without mining and once with every request mined.
GPU time changed by −1.1% to +0.3%.
At close to zero overhead cost, a miner can serve inference together with a proof of the model that wrote it, and compete for the block reward at the same time.
ALT Independent benchmark. TensorCash mining adds close to zero GPU time. GPU time added by mining over a full agentic session: RTX PRO 6000 Max-Q −1.1%, B200 +0.3%, H100 PCIe +0.2%. Measured 21–23 October 2026. Source: Bob Labs.