Pokémon cards are up 1,350% since 2020 and crypto money is fueling the next leg higher.
Logan Paul flipped a single Pikachu card from $5.3 million to $16.5 million. Now every 20-year-old with $50 thinks they're the next one.
People are stampeding stores, running bots to clear out online inventory, and huddling in parking lots to split card hauls.
Analysts are already using words like "frothy" and "bubbly."
The Pokémon Company keeps dropping new sets every few months and just launched 30th anniversary products, so the hype machine isn't slowing down.
Somewhere between nostalgia and speculation, a children's card game became a legitimate asset class.
Source: CNBC
“未来的世界属于懂 Token 的人。” —— 出自一位10岁博主之口。
刚换了 Mac Studio 的他,不是为了打游戏,而是为了“养龙虾”(跑多个 AI Agent 协同工作)。他把复杂的 AI 产业链比作一个大蛋糕,从能源层到应用层,层层剖析。
别觉得小孩在过家家,他讲的“Token 是 AI 时代的硬通货”这个观点,可能比很多专家的报告都接近本质。
这届小孩的 AI 认知已经 Next Level 了,建议大人们反复观看,治治我们的“算力焦虑”。👇
Fantastic to spend time with many of the leaders in the Bittensor $TAO community at Yuma's first ever summit. The vibes, enthusiasm and camaraderie reminded me of DCG's first Summit in 2015
https://t.co/ritf6lTYGK
Truly an inspirational group. Big things are happening
Bought another ~$50k in $tao
Current exposure is ~$750k
My price target is $500 in 2026
My thesis is subnets are shipping product
This is not investing advice, this is me asking you to savage my trade so I learn
#jaytrading
"But Bittensor doesn't really do anything..."
Most companies or crypto projects struggle to be the world's best at 1 thing.
In just over a year, Bittensor has become the world's best at over a dozen things. And I've most likely missed a few subnets here.
Bittensor is the Nasdaq of distributed decentralising AI... For me, it's literally a candy shop of some of the brightest minds out there building cool stuff.
And all it takes is just 1 out of 128 subnets to achieve Product Market Fit and make hefty revenues and it will drag the price of TAO up. Because it will attract buyers into that subnet, but the only way to buy that subnet token is to first buy TAO and then swap it for that subnet token.
Ethereum missed a trick here. ETH was never needed (other than gas fees) or linked to an ERC-20 project that mooned. With Bittensor, when you buy a subnet token, you're not selling TAO, you're temporarily swapping it...
Subnet 64 @chutes_ai just confirmed in their latest interview with @JesusMartinez that they have the capability to run 1 trillion parameter models.
let's put that into perspective for a second:
templar was making waves for running 72b models and rightfully so, that was genuinely impressive. but @chamath, when he brought up templar in a recent interview with Jensen Huang from @nvidia, thought they were running 4b models. not 72b. 4b. and he was still bullish on it.
now imagine when he finds out what chutes confirmed they can do..
1 trillion parameters. on a decentralised subnet. no AWS. no Google Cloud. no centralised infrastructure pulling the strings. just Bittensor doing what it was always supposed to do - incentivising the best compute in the world to show up and compete.
Jensen Huang talks about whoever is pushing the frontier of compute. that's the conversation chutes just walked into.
this is the kind of milestone that doesn't stay quiet for long. when people outside of crypto start connecting the dots that a bittensor subnet is running inference at a scale that rivals or exceeds what the top AI labs are deploying, the narrative around $TAO changes completely.
Chamath gets it wrong on the model size and is still impressed. Jensen sees who's actually moving compute forward. and chutes just put a number on the table that neither of them can ignore.
the noise hasn't started yet. but it will.
Sam Dare just dumped 37,000 TAO on his own subnet holders.
Roughly $10 million. In two clips. After publicly accusing Bittensor of being centralized.
Then he walked.
This is the Covenant AI exit nobody is telling you the full story about.
The setup.
Sam Dare runs Covenant AI. They operated three subnets on Bittensor. Templar (SN3), Basilica (SN39), Grail (SN81).
Templar was the crown jewel. The team behind Covenant 72B. The model Jensen Huang shouted out on All-In. The reason TAO ran 90% in March.
Yesterday Sam posted a long thread accusing Bittensor founder Const of "decentralization theatre."
Hours later he sold his bags.
The allegations.
• Const suspended emissions to Covenant's subnets
• Const removed Sam's moderation rights over his own community
• Const deprecated Covenant's subnet infrastructure
• Const dumped TAO at strategic moments to apply economic pressure
• A three-person multisig that's actually controlled by one guy
Sounds damning. Until you read Const's response.
Const's actual response.
• "I do not have the ability to suspend emissions." He sold 1% of his alpha holdings on three Covenant subnets that were on near-100% burn code. Same buy/sell rights every TAO holder has.
• He didn't remove Sam as a Discord mod. He gave Sam a temporary timeout for deleting honest criticism from his own community channel. Mod role was reinstated.
• "Deprecating infrastructure." Const doesn't even know what that means.
• Less than 1% of what Const had personally invested in Sam's teams. Visibility is impossible to avoid in his position.
Const previously gave Sam a 2,000 TAO bonus and invested thousands of TAO into his subnets.
Const's "dump" was around 200 TAO.
Sam's dump was 37,000+ TAO.
The rug.
That's $10 million pulled from holders who believed in the Covenant 72B story.
Subnet holders are catastrophically down. The "noble dev exposing the evil founder" framing falls apart the second you check the wallet.
What Const is actually building next.
Const answered a DM from @TaoOutsider right after the exit. He sounded calm. Called it a "spring forward moment."
Here are some suggestions @0xSammy gave for TAO:
• Lock-based subnet ownership. Subnet ownership tied to a team's long-term economic commitment.
• Vesting schedules visible onchain so investors see unlocks in advance.
• Founders won't be able to sell emissions as they receive them.
• Exit cool-down periods so leaving teams can't dump 100% at once.
• Headless subnet architecture so no single team can pull the plug.
• Protocol-level IP retention so models trained on Bittensor compute stay with the network.
• Onchain revenue sharing flowing back to alpha holders.
These are the changes Bittensor needs right now. I hope this $10M rug will force these changes to happen sooner, rather than later.
The contrarian take.
Most people are watching this and concluding Bittensor is broken.
I'm watching this and concluding Bittensor is finally maturing.
Const built the team. Const funded it. Const did most of the research on Templar. Sam was the face. The compute, the emissions, the incentive design that made Covenant 72B possible. None of it leaves with him.
Someone else can take that model and build a bigger run on the same network.
The subnets to watch right now per Stillcore Capital's Mark Jeffrey:
• SN64 Chutes. Inference at 1/6th the cost of competitors.
• SN62 Ridges. AI coding assistance beating Claude and Cursor on SWE-bench at 1/770th the cost.
• SN4 Targon. $42M valuation.
• SN9 IOTA. Decentralized training infrastructure.
(Also SN50, 68, 65, 44, 18, 85, 17, 48, 93, 54)
The TAO bull thesis didn't need Sam Dare. It needed proof that builders can build on permissionless infrastructure.
Covenant 72B was that proof.
It still exists. The network that made it possible still exists.
Sam left. The thesis didn't.