🏦 Take control of your digital assets!Join CoinDesk for a deep dive into secure, flexible digital asset custody. Explore how institutions can define rules, manage keys, and orchestrate token operations with precision, preparing for the accelerating world of tokenization.🗓️July 10th⏰10 A.M. ETGet expert insights into scaling from pilot to production: 🔗 ⤵️ https://t.co/RZv88gwXHk[Sponsored by @Ripple] ...

🔥 LATEST: Galaxy Research says Robinhood’s tokenized stocks shift NYSE liquidity on-chain, challenging TradFi’s dominance. https://t.co/jGjWUfL90T ...

RT Token Terminal 📊 (@tokenterminal)◼️ According to @BlackRock, over 80% of companies that generate $100m+ in annual revenue are private.⛓️ A subset of these companies will likely seek liquidity through tokenization vs. IPO -- especially if tokenization becomes labeled as an 'easy IPO'.🔎 Without proper reporting to investors, the 1st wave of private equity tokenization is likely to resemble the 1st wave of SPACs. ...

Stock tokenization on Solana increasing fastStill very early with many limitations — but it's clear that the future we all envisioned is now happening in real timeMuch more progress to come https://t.co/gfB4F6niQ7 ...

Fintech features:💵Cheaper = pool user deposits to negotiate lower fees and/or higher yields🏃Faster = optimize deposit & withdraw actions across partner banks & other sources of yield📲Easier = bundle complex deposit & withdraw flows into one click actions⚠️ Each feature comes w/ an extra layer of risk that the fintech company managesBridget (@bridge__harris):Most equity tokenization today is analogous to the Variable Interest Entity (VIE) structure used by Alibaba ($BABA) on US exchanges, where American inve ...

🔥 TODAY: Gate xStocks launched perpetual contract trading for tokenized stocks like GOOGLX, TSLAX, and MSTRX.All settled in USDT with up to 10x leverage. https://t.co/h8pTJ4P9wr ...

RT VIKTOR (@thedefivillain)I don't hold any USELESS token, but the last time I've seen this PA was when FARTCOIN went from $20M to $400M with zero drawdownThere can only be one FARTCOIN though https://t.co/ImScrKHvAT ...

🇭🇰 LATEST: Hong Kong will regularize tokenized bonds, with a third batch in the works.It also plans to expand tokenization to metals, renewables, and more. https://t.co/e3qjDLLRn3 ...

🇷🇺 NEW: $38B Russian state corporation Rostec to launch RUBx, a ruble-backed token on Tron. https://t.co/AjTlptoI2J ...

在AI竞赛如火如荼的今天，真正制约各家公司发展的核心瓶颈究竟是什么？ semianalysis写了一篇文章，通过深入分析DeepSeek和Anthropic两家公司的策略选择，揭示了一个行业共同面临的根本挑 战：计算资源的稀缺。 DeepSeek R1发布128天后的数据呈现出一个看似矛盾的现象：官方平台用户流失，但第三方托管的模型使用量却暴增20 倍。为什么用户会抛弃价格极低的官方服务，转而选择第三方平台？ 本文通过Token经济学这一分析框架找到了答案。 文章指出，AI服务的定价本质上是三个性能指标的权衡游戏。 第一是延迟，即用户发送请求到收到第一个字符的等待时间； 第二是吞吐量，即模型每秒能生成多少个token，直接影响对话的流畅度； 第三是上下文窗口，决定了模型能"记住"多少对话历史，对于分析长文档或大型代码库至关重要。 关键洞察在于：通过调整这三个参数，服务商可以实现任何价格水平。 DeepSeek选择了极端配置——数秒的延迟、每秒仅25个token的输出速度、业界最小的64K上下文窗口，换取了极低的价 格和最大化的研发资源。 有趣的是，作为西方AI领军企业的Anthropic也面临着相似的困境。C ...