AI解决不了分配问题:
AI 再强大,解决的主要还是生产问题。生产力可以指数级提升,但财富如何分配、谁掌握资源、谁有消费能力,这些问题并不会自动解决。
甚至可能出现一种局面:技术越进步,分配矛盾反而越尖锐。
我越来越认同一种 Agent 架构:万物皆可插件。 平台只负责提供稳定的基础能力,其他功能都作为插件接入。 需要什么能力,就安装什么插件;不满意,就替换。 未来优秀的 Agent 平台,不一定是功能最多的平台,而可能是最容易组合外部能力的平台。
移民中介或者留学中介吹嘘的各移民国家的优缺点是偏离真实的,因为他的工作利益导向,这群人必然不中立。要真实准确的了解一个国家最好的方式是自己去考察,去生活定居一段时间。或者多去跟当地不同的人交流,获取多个维度的信息,才有可能做到客观,准确。
挪威的政府全球养老基金规模达到22.7万亿挪威克朗,约2.3万亿美元;累计投资收益达到15.2万亿克朗,约占基金规模的67%,自1998年以来年化名义回报约7%。按全国约560万人计算,相当于人均拥有约40万美元的公共基金资产。这笔财富通过专业管理、全球分散投资和长期复利,为公共财政与福利提供支持。挪威靠被动投资实现了所谓“几代人花不完”养老金规模,更准确地说,是依靠长期投资和以预期实际回报约3%为基准的财政使用规则,力求保留基金的实际购买力,让财富持续惠及当代人与子孙后代。
挪威和瑞典值得借鉴的,是通过专业管理、全球配置和长期投资,让公共财富持续服务于国民。挪威将石油财富积累为主权基金,由财政部制定投资框架、央行承担管理责任、NBIM执行投资,并全部配置于海外;瑞典则通过AP基金等机构管理养老体系的部分资产。两者都重视分散风险、长期回报和清晰的授权监督,让国民分享全球企业发展与资本积累的成果。这与VOO长期分散持有的理念有相通之处,但资产范围和治理结构不同。核心是:以制度保障专业投资,以长期积累支持公共财政、养老保障和代际公平,而非将基金规模等同于每个人可以直接领取的养老金。
中国的社保类似于定投,每个月给国家交钱,但这是亏本的定投。考虑到通货膨胀➕货币通胀,以及人的平均死亡年龄只比65岁大了不到15年,从概率上这个社保是100%亏本的。除非能活到100岁,那么就是赚的。因此年轻人一定要自己提前布局,坚持定投VOO,用定投的资产增值给自己养老。
2000年互联网泡沫的时候,标普500最大跌幅约50%,纳指80%。纳指主要为科技股,均衡性以及覆盖面都不如标普500。如果按照巴菲特的SPY 90%持仓,国债10%持仓,实际最底部总体回撤只有45%,回撤的心理感受更好。实际上,最底部停留的时间并不长,熬过去,反弹后,大概也就是20%-30%的回撤。只是在熊市大周期,要熬的时间比较长。但是这种时候,整个市场都不好。所以至少也是没有输给市场的。
MAX is awesome, Salute to MAX.
I finally wrote a book that covers vprogs, and in particular tic-tac-toe implemented as a vprog on testnet-10. It follows the game down to the machinery underneath, with diagrams and appendices for the details. Welcome to explore it and ask questions: biryukovmaxim.github.io/vpro…
Interestingly, of the 24 public chains ranked ahead of KAS by market cap, most are older chains. Among those launched after November 2021, only Hype, Sui, Canton, and M have a higher market cap than Kaspa, which means Kaspa ranks fifth by market cap among newer public chains.
Of these five projects, Kaspa is undoubtedly the strongest in technical strength and future growth potential. It is fair to say that Kaspa is the best-performing rising-star public chain. kaspa:native
Kaspa ranks 25th by market capitalization among all public blockchains.If ranked by market capitalization on CMC, Kaspa is in 61st place. However, if all stablecoins and exchange platform tokens are excluded and only public blockchains are considered, Kaspa ranks 25th. kaspa:native
Once KCC20, Argent, and Silverscript are fully mature, Kaspa will evolve from a high-speed PoW payment network into a truly programmable platform for native assets and decentralized finance. $kaspa
Silverscript provides a safer, higher-level way to build covenant-based smart contracts. Argent enables complex applications composed of multiple contracts and state transitions. KCC20 establishes a native token standard whose rules are enforced directly at the L1 consensus level.
Together, they could unlock native tokens, stablecoins, DEXs, AMMs, escrow, bonding curves, cross-chain assets, and other financial applications—all while preserving the parallelism and security of the UTXO model and the performance advantages of Kaspa’s high-speed BlockDAG.
The key shift: Kaspa moves from simply being “a fast blockchain” toward becoming fast, secure, permissionless, programmable financial infrastructure.
The crypto space is full of theft, scams, rug pulls, and Ponzi schemes. Then I discovered Kaspa. Its purity, mathematically grounded security, and rocket-like speed genuinely amazed me. That was when I knew I had finally found the real deal. $KASPA
Kaspa’s high speed, security, permissionless nature, and programmability provide a solid foundation for it to become a global financial settlement infrastructure in the future. $KASPA



