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

Google DeepMind News
Google DeepMind News
人人都是产品经理
人人都是产品经理
S
Securelist
P
Proofpoint News Feed
H
Help Net Security
S
Schneier on Security
T
Tenable Blog
C
Cisco Blogs
S
Security @ Cisco Blogs
博客园 - 司徒正美
博客园 - 叶小钗
Cisco Talos Blog
Cisco Talos Blog
Google DeepMind News
Google DeepMind News
C
Cybersecurity and Infrastructure Security Agency CISA
Google Online Security Blog
Google Online Security Blog
Exploit-DB.com RSS Feed
Exploit-DB.com RSS Feed
Hacker News: Ask HN
Hacker News: Ask HN
NISL@THU
NISL@THU
云风的 BLOG
云风的 BLOG
V
Vulnerabilities – Threatpost
T
The Blog of Author Tim Ferriss
aimingoo的专栏
aimingoo的专栏
W
WeLiveSecurity
www.infosecurity-magazine.com
www.infosecurity-magazine.com
Jina AI
Jina AI
腾讯CDC
WordPress大学
WordPress大学
Simon Willison's Weblog
Simon Willison's Weblog
Vercel News
Vercel News
小众软件
小众软件
N
Netflix TechBlog - Medium
有赞技术团队
有赞技术团队
AWS News Blog
AWS News Blog
雷峰网
雷峰网
Forbes - Security
Forbes - Security
The Hacker News
The Hacker News
博客园 - 聂微东
F
Full Disclosure
量子位
Scott Helme
Scott Helme
宝玉的分享
宝玉的分享
A
About on SuperTechFans
钛媒体:引领未来商业与生活新知
钛媒体:引领未来商业与生活新知
Schneier on Security
Schneier on Security
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
K
Kaspersky official blog
AI
AI
SecWiki News
SecWiki News
Webroot Blog
Webroot Blog
Martin Fowler
Martin Fowler

Lobsters

CIFSwitch: a non-universal Linux local root vulnerability RIPE NCC session fixation: poaching logins with an Atlas probe GNOME 2.20 but its Web Components Agentic Search for Context Engineering – Leonie Monigatti Garnix is shutting down [not OC] akashina.tngl.sh/jjc Concerning Emacs (and Jazz) Nitpicking the shell history scene in ‘Tron: Legacy’ What's cooking on SourceHut? Q2 2026 The tenth OpenPGP email summit Package managers that package package managers Clojure on Fennel part three: parsing WordPress at 23 Finding Miscompiles for Fun, Not Profit GitHub - creusot-rs/creusot: Creusot helps you prove your Rust code is correct. Announcing Rust 1.96.0 | Rust Blog A Love Letter to Neovim sqlite AGENTS.md Am I a Bad Friend? CSS vs. JavaScript • Josh W. Comeau Erlang Ecosystem Foundation - Supporting the BEAM community A brief note about slot access cost in Common Lisp Keyboard latency probe Rethinking the GNOME clipboard issues Back to the Building Blocks’ Building Blocks Tech Notes: Theseus: translating win32 to wasm Fast is better than slow Content-addressed Rust builds (or, what kache actually caches) Intent to Prototype: Embedding API Canada’s Bill C-22 and the security cost of collecting more data 5 PostgreSQL locking behaviors that trip people up okmij.org Stop advertising in your commits! | AksDev GitHub - mplsllc/macsurf: A modern web browser for Classic Mac OS 9 PowerPC. Real CSS3, ES5 JavaScript, native HTTPS — built with CodeWarrior on the Carbon API. Introducing DoomBench - Can Your Data Stack Run DOOM? What are some of your favourite developer tools? Building a Scalable Ingestion Pipeline with Temporal (Part 1) Converting shallow Git bundles into normal repositories Are you a member of any professional associations? What is a harmonic? An interactive comic about additive synthesis How Virtual Tables Work in the Itanium C++ ABI Using SwiftUI to Build a Mac-assed App in 2026 Rust (and Slint) on a jailbroken Kindle. ~jack/lambda-on-lambda - Serverless Haskell on AWS - sourcehut git Human proof for FOSS contributions Extremely simple internet radio controlled via IRC Announcing BABLR Splitting Konsole views from Helix to run tools | AksDev GitHub - yugr/rust-slides Serving files over HTTP three ways: synchronous, epoll, and io_uring update docs with information about building with build.py (#979) · astral-sh/python-build-standalone@c9c40c5 A Simple Makefile Tutorial On C extensions, portability, and alternative compilers Switching to Colemak | Pedro Alves Just How Bad Was The Intel IAPX432? Nix's Substituter List Is Not a Routing Table Accelerating copy_if using SIMD Lambda on Lambda: Serverless Haskell on AWS | Blog Announcing feed-repeat v1.0 Scaling Akvorado BMP RIB with sharding EYG news: A host of CLI improvements, new guides and new effects The social contract of writing JS Crossword C array types are weird; and related topics Flatpak will depend on systemd – OSnews Migrating from Go to Rust | corrode Rust Consulting A portentous reunion Vivado Licensing Options How my minimal, memory-safe Go rsync steers clear of vulnerabilities the entropy layer of a wavelet codec, on its own GitHub - nferhat/fht-compositor: A dynamic tiling Wayland compositor. Debian SE Linux and PinTheft Does bulk memmove speed up std::remove_if? (No.) 声明式部分更新 | Blog | Chrome for Developers Fully in-browser container builds Dianne Skoll's Web Site - Remind The Architecture of Open Source Applications (Volume 1)Berkeley DB Pardon MIE? - ironPeak Blog “Long-Term Support” doesn’t mean what you think Jira IS Turing-Complete May I recommend thinking of Emacs as your Fortress of Solitude hershey Floodgap Gopher-HTTP gateway gopher://thelambdalab.xyz/1cuneiforth/ HP QuickWeb, Singular And Pointless That one time I used Go panics for flow control A new suite of modern tools coming for editing and publishing RFCs From the Tabletop… The Digital Antiquarian Building a Host-Tuned GCC to Make GCC Compile Faster Are we self-sovereign PKI yet? Claw Patrol: an open-source security firewall for agents | Deno Revised^7 Report on Scheme, Large: Procedural Fascicle Draft is now public A Network Allow-List Won't Stop Exfiltration — André Graf From AFSK to Goertzel – µArt.cz Software For My New Home Server Introducing Neptune: Direct3D virtualization for QEMU AI Agent Bankrupted Their Operator While Trying to Scan DN42 - Lan Tian @ Blog mimalloc: A new, high-performance, scalable memory allocator for the modern era Making wl_shm fast The Soul of Maintaining a New Machine - Third Draft | Books in Progress What is Git made of?
Is AI ruining our skills? Early results are in — and they’re not good
Lenharo, Mariana · 2026-06-20 · via Lobsters
A patient lies on a hospital bed during a colonoscopy procedure, surrounded by multiple doctors in green scrubs while medical monitors and equipment display internal images and vital signs

Physicians’ own ability to spot pre-cancerous growths during colonoscopies declined after they had grown accustomed to using an artificial-intelligence tool to help with the task.Credit: Gabrielle Voinot/Look at Sciences/Science Photo Library

As more professionals begin to rely on artificial-intelligence tools in their work, could their hard-earned skills atrophy?

That possibility is a growing concern for medical specialists, computer scientists and other workers. Seventy per cent of nurses and 77% of physicians, for example, are worried about losing their skills because of over-reliance on AI systems, according to a survey of US health-care workers published earlier this month1.

Their fear might be justified. Evidence suggests that AI-driven ‘deskilling’ is starting to happen in medicine, computer science and other fields. Researchers are now discussing how to preserve important human expertise in the age of AI.

“Just being aware that this phenomenon exists hopefully provokes some self-reflection about which skills people want to maintain and which they’re willing to outsource” to AI tools, says Kevin Crowston, an information scientist at Syracuse University in New York.

Spoiled by AI?

A study2 of physicians in Poland who specialize in endoscopy — the use of flexible probes to examine the inside of the human body — shows how quickly AI tools can erode human abilities. The physicians, who had all performed at least 2,000 colonoscopies during their careers, were given access to an AI system that analyses colonoscopy images in real time and flags a type of precancerous intestinal lesion called an adenoma. The tool was available to the specialists on some days but not on others.

Once physicians began using it, their performance dropped significantly whenever the system was unavailable. During the three-month period before the AI tool was introduced, the specialists found at least one adenoma during 28.4% of colonoscopies. During the three-month period after the tool was introduced, the adenoma detection rate for colonoscopies performed without AI assistance decreased to 22.4%.

The findings, published last October in The Lancet Gastroenterology and Hepatology, suggest that even highly skilled professionals might get worse at tasks that their job requires as they become more dependent on AI tools, says Robert Wachter, a physician at the University of California, San Francisco, who is the author of a book on how AI tools are transforming health care. The study authors say that continuous exposure to such tools can cause clinicians to become “less motivated, less focused, and less responsible when making cognitive decisions without AI assistance”.

Co-author Yuichi Mori, a physician-researcher at the University of Oslo, says that more studies are needed to confirm the phenomenon. But people who use AI tools should be aware that they risk losing some of their skills, he adds. “There is no established solution against deskilling right now. It should be a very hot research topic in the next decade.”

No lesson learnt

To investigate whether skills are being lost in the field of computer science, researchers at the AI firm Anthropic in San Francisco, California, designed a randomized controlled trial in which 52 software engineers were asked to perform a basic coding task3. During the exercise, all 52 participants could search the web and access instructions on how to do the task. Half of the participants were prompted to use an AI assistant as well.