New York Stock Exchange signals move to build an NFT marketplace

Bitcoinist发布于2022-02-16更新于2022-05-26

文章摘要

The New York Stock Exchange (NYSE) is catching many off guard this week with an interesting and unique trademark filing that is signaling the exchange’s intent to build it’s own NFT marketplace.

The New York Stock Exchange (NYSE) is catching many off guard this week with an interesting and unique trademark filing that is signaling the exchange’s intent to build it’s own NFT marketplace.

A patent from the NYSE has been filed within the past week with the United States Patent and Trademark Office (USPTO) has suggested that the exchange is making it’s first ‘metaverse’ move. Let’s look at what we know from early reports, and what it could mean moving forward.

New York, New York

According to the USPTO filing, the New York Stock Exchange is registering “NYSE” for utilization of several different blockchain and crypto-adjacent goods and services, including areas like “virtual reality and augmented reality software, non-fungible tokens, and/or NFTs, and online marketplaces.”

The exchange could be looking to provide “downloadable virtual goods,” according to reports, and could be looking to build “authentication of data in the field on NFTs using blockchain technology.”

It’s the latest twist in the unexpected road of stock exchange and NFT crossovers, after the NYSE minted six commemorative NFTs last year – celebrating major public listings as part of a series titled “First Trade.”

The New York Stock Exchange's flagship composite index, NYA, has had a strong performance over the past year. | Source: NYSE: NYA on TradingView.com

The “State” Of Affairs

The New York Stock Exchange is the world’s largest stock exchange by market cap, with average trading volumes that are often estimated in the hundreds of billions of dollars each day.

Unlike last April’s NFT release, this latest endeavor from the NYSE signals a much broader and bigger investment into the blockchain territory; while details are still spotty, it seems that the exchange will be creating a full-fledged NFT marketplace. It comes at a time when competition from market leader OpenSea has becoming increasingly fierce. Coinbase is now seemingly weeks away from their NFT marketplace, and growing marketplace LooksRare has been increasingly mentioned in recent weeks (despite a rough week in the headlines this week) – and these are just two names among many.

The trademark application also includes APIs and “cryptocurrency trading services” as well. Could this move be the first signal that the NYSE could be moving towards supporting crypto trading more broadly? With minimal U.S. regulation that has been implemented to date, the territory can easily be described as a ‘grey area’ at best for the exchange – but innovation isn’t known to wait. While ‘metaverse’ is the “flavor of the month” lately, the New York Stock Exchange utilizing blockchain APIs in stock tracking, or adopting NFTs at a much more substantial level is nothing to scoff at.

你可能也喜欢

Deepseek 能否为中国节省 1 万亿美元?

《DeepSeek 能否为中国节省 1 万亿美元?》一文探讨了DeepSeek如何通过技术创新,可能大幅降低中国AI基础设施的建设成本。 核心在于,AI硬件的成本正从计算芯片向昂贵的内存(如HBM)倾斜。DeepSeek通过三方面技术突破,系统性降低了对这类稀缺硬件的依赖: 1. **压缩“记性”**:采用多头潜在注意力(MLA)等技术,将长上下文对话所需的缓存(KV Cache)压缩至原来的极小部分,极大节省显存。 2. **按需唤醒“身体”**:利用混合专家(MoE)模型架构,每次只激活部分参数,并结合智能的存储调度,将多数参数移出昂贵显存。 3. **复用计算结果**:将算过的上下文作为缓存复用,用低成本的内存读取替代高成本的重复计算。 这些技术协同,使DeepSeek V4等模型在长上下文场景下,用同等硬件可能产出数倍的有效Token,等效于大幅提升硬件效率、摊薄单位成本。 文章据此进行推演:到2030年,全球AI硬件投资预计达数万亿美元。若DeepSeek的技术路线能使中国未来AI基建的硬件需求等效减少75%,在日均Token消耗达到数千万亿级的规模时,节省的投资额可能接近1万亿美元(约7万亿元人民币),相当于少建数万座智算中心。 更重要的是,此举将产业价值从受制于人的尖端算力芯片,部分转移至中国已有所突破的存储芯片及系统工程领域,提升了供应链安全性。DeepSeek的意义不在于“消灭硬件”,而在于通过极致优化,降低行业对最昂贵、最稀缺硬件的边际依赖,让AI能力更普惠,重塑未来AI基建的账本。 (注:文中关于万亿节省的具体数字属于基于行业趋势的推演观点。)

marsbit5分钟前

Deepseek 能否为中国节省 1 万亿美元?

marsbit5分钟前

推翻主流治幻觉思路:元认知,才是大模型破幻觉的全新解法

谷歌研究院与特拉维夫大学联合发表论文,提出对抗大模型“幻觉”问题的新思路:与其追求让AI全知全能或过度拒答,不如培养其“元认知”能力,即让模型能够感知并诚实表达自身对每个答案的确信程度。 论文指出,当前主流方法存在局限:一味增加知识覆盖无法穷尽所有事实;而通过大幅拒答来降低错误率则会征收沉重的“实用性税”,牺牲大量本可正确回答的问题。核心原因在于模型缺乏“判别力”,难以精准区分具体问题的对错,导致校准良好但实用性低下。 论文重新定义了“幻觉”:问题不在于AI输出错误信息,而在于其“没有资格确定却以确定的语气给出错误信息”。因此,解决路径应是实现“忠实不确定性”——让AI语言表达的确信度与其内部状态的真实确信度对齐。这比消灭所有错误更可行,是一个依赖内部信号的闭环问题。 在AI代理(Agent)时代,元认知更为关键。没有它,Agent在调用外部工具(如搜索)时将陷入“盲飞”,无法智能决策何时需要搜索、如何评估信息可信度。 实现元认知面临几大挑战:“自举悖论”涉及用静态数据训练动态能力的困难;“对齐破坏信号”指RLHF等训练可能磨灭模型原有的内部不确定性信号;“因果性评估”则需区分真正的元认知与对其的表演。 论文建议,评估反幻觉方法应超越单一准确率指标,转而分析完整的“实用性-错误率权衡曲线”,并关注其在其他任务上的“附带损伤”。最终目标是让AI学会诚实地沟通其认知状态,从而在保留实用性的同时,将错误信息的危害降至最低,建立可靠信任。

marsbit10分钟前

推翻主流治幻觉思路:元认知,才是大模型破幻觉的全新解法

marsbit10分钟前

交易

现货
合约
活动图片