Rephonic
Artwork for 拾余光

拾余光

Norvyn

我是一个刚入门的独立开发者,在不断地试错,试着向外、向上探索。我也订阅过很多大佬的播客、公号,但随着AI突飞猛进,好多内容都不约而同地机械化、平庸化。你很难再从这些内容中间,获得与付出时间等价的视野与洞察。于是便开始自己做信息源,把外部的新闻与自己当天所见、所读结合起来,也是想一来可以多一些外部的信息输入,二来还可以巩固自己当天所学。听了几期之后,觉得对自己有帮助,于是想分享出来。如果可以恰巧对您有所启发,再好不过,如果并无所感,就当逗个乐,听个响。

PublishesDailyEpisodes11Founded11 days ago
Language
ZH-CN
Category
Technology

Listen to this Podcast

Artwork for 拾余光

Latest Episodes

AI-OS要解决的是:当AI Agent真正替代人工作——自动发邮件、自动写代码、自动操作数据库——它需要的不是一个语言模型,而是一个能管理内存、管理工具调用、管理多步骤任务流程的运行时环境。

这里有一个需要回应的反方证据:LangChain已发布LangGraph,实现有状态图式协调;Anthropic的Claude Agent SDK和OpenAI的Responses API均引入了原生Agent抽象层。 说"现有框架只是在语言模型外包胶水代码"在2026年的语境下部分过时——头部框架已经... more

编译时安全不是 AI 领域的新概念,但它在 AI Agent 生产系统中的意义被低估了。当 AI 系统的决策逻辑足够复杂,当 Agent 循环的并发度足够高,我们需要的不只是"看起来能跑"的系统,而是"运行时不会因为数据竞争而崩溃"的系统。Swift 6 actor 模型提供的编译时可拦截数据竞争,在这个维度上是真实的进步——不是全部的进步,但是真实的。

设备端 AI 栈的完整闭环,解决的是另一个维度的确定性:数据主权和隐私保护。当推理可以在本地完成,数据不需要离开设备,我们对 AI 系统的信... more

当所有人都依赖同一个AI获取答案时,你真正值钱的不是你知道什么,而是你知道一些别人不知道的东西——而这些"不知道",正是你的知识库帮你积累的。

这不是对抗AI,而是在AI时代重新定义"知识"这个词的边界。

AI擅长处理显性的、可编码的、可规模化复制的知识。人类擅长处理的,恰好是那些"潜移默化"的、依赖于具体情境的、无法被标准化的知识。

《愚笨与智慧》里说"聪明是下贱的,愚笨则直率而且诚实"。在AI时代,这句话可以翻译成:

依赖AI给你的答案,是聪明的;拥有只有你自己知道的东西,才是真正有... more

0%到90%的成本削减意味着什么?理论上,以前只能服务10个开发者的资源,现在可以服务50个甚至100个。但这个线性外推太粗糙了——没有考虑并发瓶颈、内存限制和API限速。真实场景通常是:token消耗降了,但并发请求一多,延迟就上来了。所以我倾向于把它理解为"显著优化",而不是精确数字。

不过这不是简单的省钱,而是改变了AI编程工具的经济可行性。

但这里我要插一句重要的反方证据:不是所有人都同意这个方向。《AI编程的效率幻觉》这篇文章提出了一个关键反驳:token成本降90%,很可能伴随输... more

过去这三年,所有人都在说 AI 要更聪明、更强大、更全能。GPT-5 要来了、Claude 4 要来了、 Gemini 2.0 要上天了——好像 AI 越"聪明"越好,越"全能"越值钱。

但我最近看到几个信号,越来越觉得事情正在起变化。

我要说的观点可能会让一些人不太舒服:AI 的未来,可能是越来越"笨"的。

别急着反驳。我说的这个"笨",不是智障的笨,不是"你好我是小爱同学请问有什么可以帮您"的笨。我说的"笨",是那种——陀思妥耶夫斯基说的——"笨拙是简捷而朴质,聪明则是圆滑而又躲闪"的... more

我特别认同 Lattner 的一个观点:AI 的瓶颈已经不是模型本身了,而是系统效率。当我们在讨论哪个模型更好的时候,真正的机会在于如何让模型跑得更快、更省电、更便宜。这和我在前面提到的 on-device AI 效率突破是一致的。整个行业正在从追求更强的模型转向让现有的模型跑得更好。这个转向对独立开发者来说是个好消息。模型能力已经足够好了,好到可以放进你的口袋。你的 app 需要做的,不是接入最强大的云端模型,而是找到那个模型和你的用户场景之间的最优匹配。

我甚至觉得,这是我自己的判断,不一... more

Anthropic 发布 Claude Design,想让设计师用自然语言做设计。另一边,开发者用 DeepSeek 给 Claude Code 做成本套利。这两件事放在一起——说明什么?

说明AI工具的价值正在分层。

顶层是"AI做决策"—— Claude Design 让 AI 来生成设计方案,Anthropic 显然认为 AI 可以在创意领域做主导。

底层是"AI做执行"——deepclaude 的开发者不在乎 AI 是什么,只在乎 AI 够不够便宜,能不能完成任务。

这两个方向都... more

我们正在进入一个 AI 能力大幅跃升的时代,但这个时代最稀缺的东西不是更强的模型,而是更好的判断力。

更好的判断力来自哪里?来自沉思,来自对意义的持续追问,来自在加速的洪流里主动选择慢下来的能力。

工具变强了,人的功课其实更重了,而不是更轻了。

这让我想起陀思妥耶夫斯基那句话:到了五十岁,交游广阔的聪明人永远会越来越自信。

他在说人的自信会随着年龄增长,但这种自信跟智慧是两回事。我们积累的成就越多,听到的认同声音越多,就越容易活在自己的判断惯性里,越难停下来问自己一个简单的问题——

... more

Key Facts

Contact Information
Podcast Host

Top Technology Podcasts

Acquired
AcquiredBen Gilbert and David Rosenthal
Tomorrow, Today
Tomorrow, TodayShekhar Natarajan
Hard Fork
Hard ForkThe New York Times
The Room Podcast
The Room PodcastClaudia Laurie and Madison McIlwain
Eye On A.I.
Eye On A.I.Craig S. Smith
Building AI Boston
Building AI BostonBuilding AI Boston

Audience Metrics

Listeners, social reach, demographics and more for this podcast.

Gender SkewLocationInterests
ProfessionsAge RangeHousehold Income
Social Media Reach

Frequently Asked Questions About 拾余光

Where can I find podcast stats for 拾余光?

Rephonic provides a wide range of podcast stats for 拾余光. We scanned the web and collated all of the information that we could find in our comprehensive podcast database. See how many people listen to 拾余光 and access YouTube viewership numbers, download stats, audience demographics, chart rankings, ratings, reviews and more.

How many listeners does 拾余光 get?

Rephonic provides a full set of podcast information for three million podcasts, including the number of listeners. View further listenership figures for 拾余光, including podcast download numbers and subscriber numbers, so you can make better decisions about which podcasts to sponsor or be a guest on. You will need to upgrade your account to access this premium data.

What are the audience demographics for 拾余光?

Rephonic provides comprehensive predictive audience data for 拾余光, including gender skew, age, country, political leaning, income, professions, education level, and interests. You can access these listener demographics by upgrading your account.

How many subscribers and views does 拾余光 have?

To see how many followers or subscribers 拾余光 has on Spotify and other platforms such as Castbox and Podcast Addict, simply upgrade your account. You'll also find viewership figures for their YouTube channel if they have one.

How many episodes of 拾余光 are there?

拾余光 launched 11 days ago and published 11 episodes to date. You can find more information about this podcast including rankings, audience demographics and engagement in our podcast database.

How do I contact 拾余光?

Our systems regularly scour the web to find email addresses and social media links for this podcast. We scanned the web and collated all of the contact information that we could find in our podcast database. But in the unlikely event that you can't find what you're looking for, our concierge service lets you request our research team to source better contacts for you.

Where can I see ratings and reviews for 拾余光?

Rephonic pulls ratings and reviews for 拾余光 from multiple sources, including Spotify, Apple Podcasts, Castbox, and Podcast Addict.

View all the reviews in one place instead of visiting each platform individually and use this information to decide if a show is worth pitching or not.

How do I access podcast episode transcripts for 拾余光?

Rephonic provides full transcripts for episodes of 拾余光. Search within each transcript for your keywords, whether they be topics, brands or people, and figure out if it's worth pitching as a guest or sponsor. You can even set-up alerts to get notified when your keywords are mentioned.

Find and pitch the right podcasts

We help savvy brands, marketers and PR professionals to find the right podcasts for any topic or niche. Get the data and contacts you need to pitch podcasts at scale and turn listeners into customers.
Try it free for 7 days