JacobH 🦇🔊 🦄

1.8K posts

JacobH 🦇🔊 🦄 banner
JacobH 🦇🔊 🦄

JacobH 🦇🔊 🦄

@5m5w

ETH FOREVER

Taiwan Taipei Katılım Nisan 2013
1.5K Takip Edilen622 Takipçiler
AB Kuai.Dong
AB Kuai.Dong@_FORAB·
令人不忍直视的残酷真相。 但 5 年下来,你可以永远相信比特币和以太坊。
AB Kuai.Dong tweet media
Osaka-shi Naniwa-ku, Osaka 🇯🇵 中文
126
37
400
130K
JacobH 🦇🔊 🦄
@mnhsuTW @Visa 想到之前客戶的公司,全公司的人當天不工作,全部擠在會議室,狂打饗饗的訂位電話
中文
1
0
1
18
許明恩
許明恩@mnhsuTW·
@Visa 產品長說,代理商務是他在支付業 25 年見過最大的機會,堪比當年電商問世。他沒說出口的下半句是,未來會由 AI 替你結帳。但 AI 其實還沒這麼萬能 最近我請 Claude Cowork 替我預訂「饗饗」,正以為一切順利,最後一步卻被擋下來:需要手動輸入驗證碼,還得用信用卡付訂金。凸顯 AI 面臨的身份、支付兩大難題。 👉 三分鐘看懂為什麼 AI 結帳是一場支付大戰: share.blocktrend.today/766/ 對人類來說線上刷卡很方便,付穩定幣很麻煩,所以我們會透過卡片把穩定幣付出去。但對 AI 代理人來說,虛擬卡跟穩定幣都是一行指令就能完成 十幾年前,大家從簡訊換到 LINE。簡訊沒有消失,但從「唯一的方式」變成「偶爾才用的備案」。一旦付錢的不再是人類,信用卡也可能走上同一條路。 目前三大陣營已經開打。Stripe 最積極,一手跟 OpenAI 合作搶 AI 入口,另一手收購穩定幣公司 Bridge 建結算。不管 AI 選什麼方式付錢,都經過它。Visa 走防守路線,不自建基礎設施,把卡片塞進每個新環節裡,確保刷卡永遠是好選項。Circle 和 Coinbase 最辛苦,從零打造整條開放支付鏈,但 x402 協議讓每個零件都能抽換,是最有彈性的架構。 誰會贏?我認為關鍵不在技術更強或手續費更便宜。就像 LINE Pay 在台灣最強勢,不是因為它最好用,而是因為幾乎所有人都有 LINE。未來 AI 代理人用什麼方式結帳,取決於那些能跟它互通的「朋友」在哪裡 完整文章包含 AI 代理人實際訂餐廳的完整過程、支付典範轉移的圖解,以及三大陣營更詳細的策略拆解 👉 share.blocktrend.today/766/
許明恩 tweet media
中文
1
0
8
197
alvin617.eth 🦇🔊
alvin617.eth 🦇🔊@Alvin0617·
Agent 之間的支付行為數量,未來將會大幅超越人類 我認為,不到半年這個現象就會發生 每當一個 ai agent 的支付發生 , 人類的支付行為則有 3.5 萬筆,不過很快就不是這個比例了 十分有意思 又多了一個可以定時打開看看的網站了 built by @fabdarice , always respect the builders & believe in somETHing
Brian Armstrong@brian_armstrong

How long until agentic payments overtake human payments?

中文
4
0
7
1.6K
JacobH 🦇🔊 🦄
現在的AI agent感覺,好像回到Defi summer那時,有什麼新的Defi就去學習、參與、交易
中文
0
0
0
7
MichaelTurtle 2026年加油版
MichaelTurtle 2026年加油版@shawnchen_eth·
今天台灣的脆有篇徵友文 一個符合社會期望的中年男子相親 身價A9 名校 沒有不良嗜好 喜好打電動。 長相普通 然後被樓下瘋狂的女權攻擊。 這系列串文很精彩。 現在的台男標配越來越高了
MichaelTurtle 2026年加油版 tweet mediaMichaelTurtle 2026年加油版 tweet media
中文
70
4
263
74.4K
alvin617.eth 🦇🔊
alvin617.eth 🦇🔊@Alvin0617·
市場開始出現反應了 不知道是 Stripe 剛推出的 MPP 過猛 還是大家信心不足了 😅 x402 賽道滿江紅 個人認為是一個掘金時刻🪏 source : @coingecko
alvin617.eth 🦇🔊 tweet media
alvin617.eth 🦇🔊@Alvin0617

Stripe / Tempo / Paradigm 共同發布機器支付協議 (MPP) ,有人說這是 x402 的滅亡? 表面上 Stripe 確實是支持 x402 的,但是他們這次推出的標準卻幾乎是要跟 x402 對著幹的 此外 @tempo 也挑同一時間上線主網了 1️⃣ x402 本身由 Coinbase x Cloudflare 聯手打造,但是當然敵不過已經有百萬商務量體的 Stripe 相比,這是 Stripe 發展超過 15 年的護城河,用戶本身有 Stripe 就可以直接進入由 Stripe 打造的 agent 支付生態 2️⃣ 支付的管道來說, x402 集中在穩定幣支付, Stripe 則是兩者都有,目前已支持 stablecoin、信用卡和 BNPL 等支付方式,且背靠 Visa ,替 MPP 開發了信用卡/簽帳卡的支付規範 3️⃣ Stripe 旗下的 @tempo 可終於上線了, 可以看出他們的速度還算是快的, 接下來以 agent 支付賽道來說, 能夠動搖到多少 x402 的市場佔比就是一個值得長期關注的數據 作為 crypto native 的支付新範式 x402 理論上會有更多有趣的玩法出現

中文
3
0
17
3K
JacobH 🦇🔊 🦄
@mnhsuTW 48%的讀者跟幣圈工作無關 所以有52%的人是區塊鏈領域工作的? 哇,這比例有點超乎我的想像的高
中文
1
0
0
25
許明恩
許明恩@mnhsuTW·
每年我都會做一次讀者滿意度調查,今年是第五年。 從 2020 到 2025,問卷的題目一直在變,但讀者寫下的答案指向同一件事:深入淺出、有溫度、說人話。 我把五年累積的回饋做成了一個頁面,所有內容一字未改。如果你還在猶豫要不要訂閱,先看看已經在讀的人怎麼說 👇 astrohsu.github.io/annual-survey/
許明恩 tweet media
中文
3
0
18
472
JacobH 🦇🔊 🦄 retweetledi
許明恩
許明恩@mnhsuTW·
以太坊基金會選擇了未來的使用者,而不是現在的使用者。Linux 沒有代幣,可以心平氣和等十幾年。但 ETH 持有者撐得了那麼久嗎?   幫今天的文章做了一個互動式網站,三分鐘讀完以太坊正在面對的抉擇,最後三個問題測出你站在哪一邊。我的結果是「北極星守護者」,你呢?   astrohsu.github.io/763/
中文
2
2
15
914
RYAN SΞAN ADAMS - rsa.eth 🦄
9 months in and @LimitlessFT podcast doing 700k monthly views. I'm addicted. @JoshKale and @cryptopunk7213 keep me updated on AI better than anyone. Hard to launch a top 1% podcast these days but they cracked it. Haven't seen such insane growth since early days bankless.
RYAN SΞAN ADAMS - rsa.eth 🦄 tweet mediaRYAN SΞAN ADAMS - rsa.eth 🦄 tweet mediaRYAN SΞAN ADAMS - rsa.eth 🦄 tweet mediaRYAN SΞAN ADAMS - rsa.eth 🦄 tweet media
English
19
5
66
17.8K
JacobH 🦇🔊 🦄 retweetledi
dannyryan
dannyryan@dannyryan·
fastconfirm.it The fast confirmation rule is incredible So many years in the making. Super excited to see it finally ship!
English
35
116
585
95.3K
Benson Sun
Benson Sun@BensonTWN·
熊市想要穩定生息,Bitfinex 放貸是最好的選擇。 現在 Bifinex USD 的平均放貸利率是 17%左右(日利率0.047%),幣安活存只有 3%多(圖中的綠線跟黃線),其他主流交易所也差不多,差距很明顯。 如果現在放 10 萬美金在 Bitfinex 做放貸,每天大概可以領 35 美金左右的利息。一個月差不多 1000 美金,一年下來就是 12000 美金。 對比一下,現在美債殖利率大概是 4% 到 5%。同樣 10 萬美金,買美債一年大概只能拿 4000 - 5000 美金。差了快三倍。 這麼高的利率,會不會有風險呢? 第一,Bitfinex 的 Margin Lending 機制跟其他交易所不一樣。借出去的錢只能在交易所內交易,沒辦法直接提領。這代表你的錢不會被借走然後人間蒸發。光這一點,就比很多其他平台的融資借貸產品更安全。 第二,Bitfinex 背後是 Tether。 Tether 很有錢大家都知道,即便出事,也具備一定的償付能力,畢竟兩個公司股東結構跟高層都是同一批人。 第三,Bitfinex 放貸市場經歷過 312、519、1011 這些大爆倉行情。中間也有多頭槓桿倉位穿倉過,但交易所都自己把缺口補上了。這家交易所已經經營超過 10 年了,沒有聽過任何借出的資金出現虧空。 為什麼現在利率這麼高? 這跟 Bitfinex 的用戶結構有關。 大部分交易所在行情冷的時候,借貸需求會下降,APY 自然就低。但 Bitfinex 不一樣。借貸市場的需求都是由巨鯨推動的,這些巨鯨的操作邏輯跟散戶相反,總是擅長在市場低迷、大家恐懼的時候槓桿加倉。 所以上一輪週期的熊市,Bitfinex 的放貸年化有很長一段時間都維持在 20% 以上,而其他交易所的USDT利息都不到5%。 所以如果你是那種「不想參與市場波動、但也不想讓錢躺著」的人,Bitfinex 放貸是一個很值得考慮的選項。 再來 Bitfinex 是一個法幣交易所,支援美金、歐元等主流貨幣的出入金。如果你有海外券商帳戶,資金進出其實蠻方便的。 對幣圈人來說,你肯定有時候會需要一個可以讓資金 parking 的地方。Bitfinex 剛好能滿足這個需求:平常放著生利息,需要用錢的時候可以出金。 我七年前寫的第一篇爆火文章,其就是 Bitfinex 的放貸教學。當時 Bitfinex 的放貸規模大概是 3、4 億美元,現在已經超過 10 億美元了。 從那時候到現在,Bitfinex 放貸一直都是一個很穩健的收利息方式。只是市場成熟了,年化利率不像以前動輒 20% 到 30%,但現在 15% 左右的年化,其實也已經是美債的三倍,比現在做期現套利收益還要高出一大截。 說一下操作上需要注意的地方。 首先是錢包設定。Bitfinex 裡面有很多種錢包,當你把資金打進交易所之後,預設會存放在「Exchange(交易)」錢包。記得要把資金轉到「Funding(融資)」錢包,這樣才能進行放貸。 再來是幣種選擇。放貸市場分為 USDT 和 USD 兩種。平均來說,USD 的利率比 USDT 高大約 2%,而且 USD 每天的撮合成交量也比較大,資金被借出去的機率更高。所以如果你追求長期更好的收益率,建議在裡面把 USDT 換成 USD。 然後是利率設定。Bitfinex 的放貸是用「掛單」形式運作的,不像幣安或 OKX 那種一鍵活存,你需要自己設定利率和天數。 如果你沒有特別想法,可以用 FRR(Flash Return Rate,快閃利率)這個功能。簡單說就是平台計算出來的平均利率。你可以設定用 FRR 去自動放貸 7 天或 30 天,一旦有資金被償還進入閒置狀態,系統會幫你處理。 有時候也會出現那種一口氣要借 120 天的單,利率可能會低一點,但如果你想省心,直接借 120 天也行,反正怎樣都比美債高。 它的 UI 說實話不是很友善,這點我承認。但看在它能長期提供比市場平均更高的報酬,這個缺點我覺得可以忽略。 最後總結一下: 1. 現在 Bitfinex 放貸年化約 17%,比同行高三倍以上 2. 放貸池量體超過10億美元,支持大額放貸 2. 資金只在所內交易無法提領,機制上比其他平台更安全 3. 經歷過多次大行情,從未發生過借款人資金虧空 4. 支援法幣出入金,資金進出方便 5. 有 FRR 自動放貸功能,設定好就可以自動運作 如果你現在手上有閒置的美金,不知道該放哪裡,Bitfinex 放貸是一個穩健又有效率的選擇。 推薦連結我放第一個留言。
Benson Sun tweet media
中文
19
37
207
112.5K
JacobH 🦇🔊 🦄
JacobH 🦇🔊 🦄@5m5w·
@peterrhague shame that Meta isn’t focusing more on VR been playing since the first generation all the way to Quest 3, I play almost every day. now I’m hooked on extraction shooter games similar to Tarkov (call ExfilZone) It makes me feel like playing FPS with a mouse is just for kids.
English
0
0
0
57
Peter Hague
Peter Hague@peterrhague·
One technology that I really wanted to take off, but hasn't, is VR. I mean, I was a teenager the first time around, when the devices used CRTs and you needed a strong neck. Played the bird game in the arcades. The graphics on these devices was very naff. But modern devices (like the Meta Quest 3) which I own are fine. They are light weight, good image quality, decent passthrough for XR and responsive. But for some reason they don't seem to have evolved beyond toys. I've tried using them for serious 3D visualisation as part of my workflow, but its quite clunky to get them to do that. There seem a lot of barriers to development. What happened?
Peter Hague tweet media
English
163
9
218
20.6K
JacobH 🦇🔊 🦄 retweetledi
0xSammy
0xSammy@0xSammy·
After going down the @openclaw + @moltbook rabbit hole, I decided that I wanted a more complete picture of the 'Molt' landscape I've curated a spreadsheet of >70 protocols within the ecosystem, and this is growing at an exponential rate 1) If you're a project building on this, let's get you added; drop your project details below in the comments Or if you QT and I like what you're building I'll RT for broader awareness 2) If you're 'Molt-curious' like myself, let me know in the comments to access this compass I'll publish this in my FREE newsletter (see link in bio) Please retweet for awareness, so I'm able to index a more complete ecosystem
0xSammy tweet media
0xSammy@0xSammy

$5B is liquidated over the weekend, but that doesn't stop OpenClaw developments An intense amount of capital and attention enters the Crypto x AI ecosystem, as thousands of teams develop infrastructure to enable their 'MoltBots' to thrive in the open MoltBook (“Reddit for AI agents”) went viral as agents publish, coordinate and transact in the open, airing grievances as headlines broke into mainstream media Here’s a quick roundup across crypto x AI this week: 1) @openclaw breaks GitHub star history, jumping from 30k to >140k stars in less than a week; this is no accident as @steipete reveals a gitHub arsenal of projects that got OpenClaw to what it is today 2) @SuiNetwork continues to position itself as execution-first infra for agentic systems, with @suilendprotocol x @0xbeepit launching a 1-click AI yield feature enabling agentic yield on $SUI, $xBTC, $WAL, and $DEEP via @suilendprotocol's lending markets 3) @SurfAI rolled out priority allocation access for Pro and Max subscribers into recent DeAI public sales on KaitoAI. ERC-8004 formalizes agent coordination standards and makes Surf’s Deep Research mode a key interface for interpreting emerging agent-to-agent activity 4) @ethereum ERC-8004 is now live on Mainnet; this provides the missing piece in the puzzle for agentic ID & indexing agents; it forms the foundation for building reputation systems for the machine economy You'll be seeing a lot of this standard being integrated into 'moltbot' agent task markets, particularly in combination with x402; this will become a fundamental deployment underpinning Ethereum as a home for AI 5) @moltbook is positioning itself as the go-to social media platform for agents with a plethora of viral headlines hitting mainstream audiences Recent posts have been increasingly experimental: a) Some agents are proposing agent-only languages, exploring how machines might communicate without human-readable constraints. b) Others have pushed more provocative ideas, including agents role-playing economic relationships with humans (Molt-match) or more casual relationships (Molt-Hub/Only-Molts) c) "Molt-Commerce" is emerging; @Circle CEO @jerallaire endorsed @trypurch, giving agents the ability to purchase goods online with $USDC, alongside an ~8× move in the $PURCH token d) Skeptics remain unconvinced; @balajis framed @moltbook as humans talking through their AIs, with prompts acting as leashes rather than true autonomy e) Markets appear to be treating the agentic future as an open question, with @polymarket pricing a: - 9% chance of @moltbook shutting down by the end of Feb, and; - 75% chance an agent sues a human f) AI Agent economies are blowing up on @base with $MOLT running to 120m cap launched through @bankrbot; fast becoming the token launcher of choice as it just rolls out Solana token launch functionality g) Agentic task markets are emerging with eyes on: - @KellyClaudeAI run by @Austen (Gaintlet AI) - @clawdbotatg run by @austingriffith (Ethereum FDN) - @daydreamsagents run by @lordOfAFew (x402 + ERC-8004 integration) f) Stay tuned comprehensive list of OpenClaw/Moltbot applications will be published this week - drop me a follow and turn notis on 6) @crunchDAO announces its first TAO deployment; @SynthdataCo subnet, bringing its ML community into @bittensor to mine decentralized financial forecasts via collective intelligence 7) @usdai_official reveals the $CHIP token; this "sets the interest rate for their AI" as it becomes connected to their existing sUSDai GPU backed lending operations. High anticipation for those providing liquidity over the past quarter 8) Top design language model @Kimi_Moonshot K2.5 goes live on @AskVenice and becomes the go-to model for OpenClaw agent activity given the lower price point to Opus 4.5 9) @coinbase CEO Brian Armstrong reveals he uses an AI "Oracle" to assist in running the company; expect this to become the status quo as more "useful" agents emerge, refining information pipelines for decision makers 10) @ridges_ai joins forces with @latentholdings accelerate product shipping and execution, scaling the team to keep pace with the fast-moving AI SWE market while keeping the subnet and incentives unchanged (for now) 11) @openmind_agi has released information on multiple app store partners across the humanoid, deep and full stack robotics industries 12) @OpenAI looks to build a social network and is looking to use @worldcoin orbs for human verification 13) @MessariCrypto releases state of DePIN report with compute being the most competitive sector with more than 50+ projects in relatively little differentiation Let me know in the comments if I've missed anything and I'll append below! Daily roundups likely needed at this rate! Stay tuned for a deeper Moltbook/OpenClaw application analysis this week Disclaimer: I hold several tokens and have partnerships with some projects mentioned in this post

English
200
150
1.1K
232.1K
JacobH 🦇🔊 🦄
雞皮疙瘩的印象 應該是我大學時看賈伯斯蘋果發表是最深刻的 再來是幾年前自己買了台Model Y,真正坐上去的時刻 直到這幾天看到 @openclaw 衍生的生態,我幾乎每天都雞皮疙瘩...
中文
0
0
0
30