惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

SecWiki News
SecWiki News
M
MIT News - Artificial intelligence
博客园 - 司徒正美
I
InfoQ
V
V2EX
L
LangChain Blog
人人都是产品经理
人人都是产品经理
T
Tailwind CSS Blog
让小产品的独立变现更简单 - ezindie.com
让小产品的独立变现更简单 - ezindie.com
The GitHub Blog
The GitHub Blog
freeCodeCamp Programming Tutorials: Python, JavaScript, Git & More
WordPress大学
WordPress大学
H
Help Net Security
美团技术团队
Y
Y Combinator Blog
G
Google Developers Blog
小众软件
小众软件
The Cloudflare Blog
博客园 - 三生石上(FineUI控件)
Jina AI
Jina AI
量子位
奇客Solidot–传递最新科技情报
奇客Solidot–传递最新科技情报
D
Darknet – Hacking Tools, Hacker News & Cyber Security
Spread Privacy
Spread Privacy
博客园 - 聂微东
The Register - Security
The Register - Security
F
Full Disclosure
S
Securelist
G
GRAHAM CLULEY
Cyberwarzone
Cyberwarzone
F
Fox-IT International blog
H
Hacker News: Front Page
C
Cisco Blogs
D
Docker
L
LINUX DO - 热门话题
Google Online Security Blog
Google Online Security Blog
T
Troy Hunt's Blog
Hacker News - Newest:
Hacker News - Newest: "LLM"
T
ThreatConnect
aimingoo的专栏
aimingoo的专栏
Last Week in AI
Last Week in AI
J
Java Code Geeks
宝玉的分享
宝玉的分享
Project Zero
Project Zero
L
LINUX DO - 最新话题
博客园_首页
MongoDB | Blog
MongoDB | Blog
Stack Overflow Blog
Stack Overflow Blog
P
Proofpoint News Feed
博客园 - 叶小钗

Hacker News - Newest: "AI"

greenvilleAI.coffee · Gloomy Doomy Ask HN: Where AI Researchers Congregate? Show HN: Directionally bad – a newsletter about risks of AI centralization The AI Great Leap Forward (a warning) AMD's Lemonade SDK For AI Promotes macOS To GA Status, ROCm 7.13 Integrated Embedded acoustic AI with <16ms latency running on 8MB RAM Does anyone in your organisation own "correctness" in your AI products? Return on Intelligence, Part 3: Moats | rebecca powell GitHub - monkidy/ai-ops-sop-pack: Documentation-only SOP pack for bounded AI-agent engineering operations: PR audit, crash recovery, handoff discipline, templates, and stop conditions. AI Was Used to Recreate the Voices of Dead Pilots. The NTSB Responded by Locking Down Its Database. - Fire ... AI Visibility Engineering Glossary — AIMENSION™ Terminology AI-Declaration.md 10 AI Prompt Examples and Techniques for Better AI Outputs in 2026 Is U.S. AI Adoption Plateauing? A Comprehensive Analysis Is AI Becoming Too Smart for Its Own Good? [audio] ThinkLLM — Think through your LLM choices Show HN: Waiting for AI Grand Prix racing SIM? Me too So I made one How the Library of Congress is using both AI and volunteers to unlock public broadcasting history Verification Tree Architecture: A Probabilistic Attention Orchestration Framework for Bug Report Management in the AI Era Let Me AI That For You The dominant paradigm in AI development is scale. Bigger models, more parameters, more compute. PHI // DRIFT is a different bet. It's a cognitive middleware architecture built on a single thesis: that distinct, continuous, contextually coherent behavior in an AI companion emerges not from model weights alone — but from what is assembled into the prompt, what is retrieved from memory, and what structured state is updated between turns. Five architectural contributions: DMU — Decision Memory Unit. Replaces cosine similarity retrieval with exp(-t/τ) × reinforcement × contextual × extra. Memories are scored by what mattered to the system's ongoing state — not just what was semantically adjacent. Ablation confirmed 14.8% more context injected per prompt than cosine-only RAG. On CPU-only hardware that's a 45.4% latency difference. PEDI / DII — Persistence-Embodiment-Drift Index. A five-component falsifiable proxy metric for behavioral continuity across context resets. Not a claim about consciousness. A measurement Is AI Profitable Yet? Chemical & Engineering News What it takes to run an AI coworker on iMessage 94% will keep spending on AI even when it fails Purr - Apps on Google Play Ask HN: What to learn and do, that makes me least affected by AI in STEM? Second Brain — Your AI Tools Finally Remember You atom.plumocracy.com Cheap AI could derail OpenAI and Anthropic's IPOs The Future of AI-Facilitated Medicine I used $30,983 of AI tokens last month in Claude code on $200/mo plan Don't just 'quote' the AI Did Google’s AI agents really build an operating system for $916? AI and Doctrinal Collapse Two Loops: How China's Open AI Strategy Reinforces Its Industrial Dominance [pdf] Battle over WiseTech AI job cuts intensifies amid China staff accusations Provenance Exclusive: Departing Meta staffer posts biting anti-AI video internally amid mass layoffs GitHub - ppserapiao/mneme: the open, user-sovereign memory layer for AI. local-first · client-side encrypted · open protocol. your memory. your keys. every model. GitHub - Hades-HY-LI/ai-native-founder-playbook-skills: Provider-neutral AI agent skills for AI-native startup founders across Idea, MVP, Launch, and Scale. AI is minting new billionaires, and workers want their share GitHub - anomalyco/models.dev: An open-source database of AI models. Pivoting to reach a wider audience and hitting a 5-figure MRR Datapoint AI China: What I Learned from the AI Labs, Robotics Startups and Academia Home Why Tech Companies Are Quietly Cancelling AI Data Centers [video] On AI Maybe AI Bots Are (Mostly) Harmless
AI 并没有区分开发者——它只是揭示了他们
fred1268 · 2026-05-23 · via Hacker News - Newest: "AI"

如果说有一个话题能将一群开发者分成两派,那无疑是人工智能。一边是狂热者——坚信这是计算领域的下一个重大变革,急于向你介绍它。另一边是怀疑者——同样坚信这只是一场会像之前那样悄然破灭的炒作周期。

这种两极分化是真实存在的,而且非常尖锐。但我认为它的根源比人们意识到的要深——而且这与人工智能本身关系不大。

两种爱 #

要理解分歧,思考软件开发真正涉及的内容有助于理解。其核心同时存在两件事.

旅程:需要做的一切 这软件 — 架构决策、算法、调试过程、重构、将模糊想法变成实际可运行东西的技艺.

还有 目的地:最终成品本身 — 用户接触的东西、你试图解决的问题、最终落入某人手中的价值。

选择这个职业的每一位开发者都热爱这两者。只是热爱程度不同罢了。

目标导向型开发者往往更注重产品。他们能直觉地感知哪些功能缺失,哪些功能永远不会有人使用,哪些流程感觉不对劲。他们向往简单——不是因为简单容易,而是因为他们明白简单正是伟大产品的关键所在。代码只是手段;产品才是目的。

以过程为导向的开发者则相反。他们倾向于算法、架构、优雅。他们会花一个下午寻找合适的抽象,更干净的界面,让下一个开发者生活更轻松的方法。产品验证了工作;工作是乐趣。

人工智能实际做什么#

当你审视AI为软件开发带来的变革时,有一件事非常明显:它极大地缩短了过程,同时加速了抵达目标的速度。一个曾经需要数小时的设计思考、谨慎实施和迭代优化的功能,现在几分钟内就能实现。而且,如果你想要更多目标——更多功能、更多产品、更多创意可能——AI消除了曾经存在于想法和事物之间的技术摩擦。

透过这个视角,两极分化几乎不言而喻。

热爱这个目的地的开发者们兴奋不已。更多输出,更快速度,更多自由度。那些曾让他们慢下来的约束——样板代码、晦涩的API、机械性工作——基本上都消失了。AI就是他们的答案。

以旅程为导向的开发者感到更像是损失。AI从他们那里拿走的不是输出——而是路径。而且,它选择的路径往往不是他们想要走的。它有效,但感觉不优雅。它不属于他们。正如我在为什么我编程:从来不是关于输出中探索的那样,编程始终是一场智力游戏——输出只是证明你玩得好的证据。当游戏交给人工智能时,监督就成了剩下的东西。而监督从来不是目的。

一个光谱,而非二元#

这一切都不是非黑即白。大多数开发者既热爱过程也热爱结果——问题在于他们更热爱哪一个更多当差距足够大时,他们最终会坚定地站在一边。

这里有一个值得深思的问题:你更享受设计 是一个产品,或者是在 建造 中吗?是的,我故意用了两个不同的词。设计意味着思考某物应该是什么——它的形状、它的行为、它存在的理由。建造意味着让它实现——编写代码、组装零件、看着它诞生。两者都是软件开发的一部分,但它们带来的乐趣并不相同。

你的答案可能是一个不错的预测,说明你与AI的关系如何。

自我认知如同导航。 #

无论你更看重过程还是结果,都没有什么好处。这两种类型的开发者都是必要的。他们都能构建出优秀的软件。

但了解你是哪种人可以改变你如何应对AI对职业的影响。如果你热爱目的地,你可能已经感到解放了。如果你热爱旅程,有两个路径值得考虑。

一个是让变革成为一种邀请——也许AI让你从机械工作中解脱出来,会留出更多空间去欣赏产品、用户、成果。也许你内心比想象中更有目标之爱。另一种更自然的本能,是寻找AI难以触及的领域——问题足够复杂、足够新颖或足够奇特,以至于道路仍然属于你自己.

那些领域是存在的。它们是真实的。

目前。