Bit Buddy

@NewtonProtocol || $NEWT || #Newt
ニュートンを見ている間ずっと気になっていたのは、「認可レイヤー」の話ではありませんでした。確認を押したときに“安全だと感じる”ことと、実際に“安全である”ことの間にある小さなギャップでした。
ほとんどのオンチェーンアプリでは、信頼をウォレットのポップアップ問題みたいに感じさせてしまいます。接続して、半分読めるメッセージを読み、偏執的ならコントラクトを確認してから署名する。すると、その取引は着地するか、しないかのどちらかです。変なのは、いちばん重要な問いがたいてい遅すぎることです——そもそもこの取引を許可してよいのでしょうか？

@NewtonProtocol || $NEWT || #Newt
Newtonプロトコルの変なところは、EigenLayerのオペレーターがトランザクションを承認していることではありません。問題は、仕組み全体が結局、ほんの小さく、ほとんど退屈な瞬間に収束してしまう点です。"allow: true" か "allow: false" のどちらか。
それが、私がずっと見つめ続けていたポイントです。
SDKフローでは、Newtonがタスクをシミュレートして、意図が許可されるか／拒否されるか、そしてその理由を返すことができます。巨大なダッシュボードのような場面ではありません。派手な演出の「リステーキング・セキュリティ」イベントでもありません。ユーザーとトランザクションの間に置かれた、小さな結果オブジェクトがあるだけです。正直、そのあたりに摩擦が現れます。

@NewtonProtocol || $NEWT || #Newt
The strange part of Newton Protocol isn’t the agent doing something. It’s the moment before you let it.
That sounds small, but it’s the bit that keeps sticking with me. Everyone likes the idea of onchain automation until the interface asks what the automation is actually allowed to touch. Not in a vague “be safe” way. In a real way. Amount. Asset. Route. Time window. Revocation. Maybe price limits. Maybe volatility limits. Maybe only this wallet, only this chain, only this kind of action.
That’s where people slow down.
Newton is aimed at a huge surface area. Binance Research said the first verifiable agent demo had more than 1.1 million sign-ups, 600,000 verified agent transactions, and 350,000 activated agents at NEWT launch. It also listed 215 million NEWT as circulating at launch, 21.5% of the total supply. CoinMarketCap recently showed NEWT around $0.05, with roughly $9 million in 24-hour volume and a market cap near $15 million. So the thing is live enough to observe, but still early enough that the rough edges are more interesting than the polished story.
The rough edge is permission anxiety.
Crypto users are used to risk, but they’re not always used to specifying risk. That’s different. A wallet approval is usually a dumb little moment people click through too fast or obsess over too late. With an agent, that approval becomes more like writing a policy. You’re not just saying yes to one transaction. You’re saying yes to a class of possible future actions.
And people don’t naturally think that way.
You can feel the contradiction. The same user who complains about manual execution doesn’t want to give an agent too much room. The same DAO that wants cleaner treasury automation still needs someone to write down what “cleaner” actually permits. The same trader who wants recurring behavior doesn’t want to wake up and find out the bot followed instructions too literally.
That’s not a Newton-only problem. It’s probably the whole agentic finance problem. Newton just makes it obvious because its core promise depends on programmable permissions being understandable enough for humans and strict enough for machines.
The asset side keeps getting bigger. RWA.xyz recently showed about $299 billion in total stablecoin value and more than $26 billion in distributed tokenized asset value. Those numbers are why “safe passage for trillions” doesn’t sound insane anymore. But the funny part is that the next bottleneck may not be settlement or liquidity or even proof generation. It may be the user sitting there trying to decide whether an agent should be allowed to spend $500, $5,000, or 5% of a vault.
I don’t think people are scared of automation exactly. They’re scared of discovering that they never had a clear policy in their own head.
Newton’s useful pressure is that it forces the policy to exist before the action. That’s annoying. It’s also probably the point.
#SouthKoreanStocksRise5% $ARX
$NEX