Pratica di Shadowing: Tragic mistake... Anthropic leaks Claude’s source code - Impara a parlare inglese con i video
Creazione lezione...
1
Yesterday, the most ironic thing ever happened.
2
Anthropic, a $380 billion startup built on the idea of safety first, that advocates for closed-source software for the supposed benefit of humanity,
3
a company Elon calls Miss Anthropic, whose logo is definitely not a sphincter, whose CEO has been warning us for years that human programmers will be replaced by AI in six months,
4
just accidentally leaked Claude Code's entire source code to the internet at 4am, officially making Anthropic more open than OpenAI.
5
Within minutes of the leak, Chaofan Xiao, a security researcher,
6
discovered that version 2.1.88 of the Cloud Code NPM package was shipped with a 57MB source map file.
7
You know, that file that's only used in development with the full readable source code of your project.
8
This holy grail of leaks containing over 500,000 lines of TypeScript code quickly spread across the internet like wildfire.
9
Anthropic's legal team courageously issued DMCA takedowns, but by the time they woke up in San Francisco, it was already too late.
10
The code was mirrored countless times and cloned by slop tubers like Fireship, which the Supreme Court says I can do legally, by the way, as a world-renowned journalist.
11
I know a lot about the law and various other lawyerings, um...
12
In today's video, we'll look at all of the incredible discoveries in the code that Anthropic doesn't want you to know about.
13
Like its anti-distillation poison pills, its mysterious unreleased features, its undercover mode, its regular expression-based frustration detector,
14
and many other super-secret techniques at the foreskin of AI research.
15
No, this is not an April Fool's Day joke, it is April 1st, 2026, and you're watching the code report.
16
Unfortunately, my lawyer just informed me that showing you Anthropik's TypeScript code would be a violation of my parole, and I refuse to go back to jail.
17
But luckily, the open source community has already created a loophole.
18
Ironically, Claude's most prolific user used OpenAI Codex to rewrite Claude Code's TypeScript code to Python code, resulting in a new, barely legal project called ClawCode,
19
and it's already become the fastest repo in history to surpass 50,000 GitHub stars.
20
Not only that, but somebody else slop-forked the leaked code and made it work with any model.
21
They're calling this new project OpenClaude, and it makes projects like OpenCode completely obsolete.
22
But maybe, now that the code's out in the open, Anthropic will just make it open source.
23
Somebody tried to make a pull request with the leaked code, but not surprisingly, it looks like Anthropic already deleted it.
24
What a mother-effin' crazy 24 hours, but how did this code end up leaked in the first place?
25
Well, as I mentioned, the source map was accidentally packaged in an NPM release, but that's weird because build tools normally strip out source maps automatically.
26
Well, Cloud Code is built on BunJS, which as you might recall was recently acquired by Anthropic.
27
And it just so happens that about three weeks ago, somebody opened up an issue on GitHub about BunJS serving source maps in production.
28
Wouldn't it be ironic
29
if the fastest JavaScript runtime in the world also turned out
30
to be the fastest way to ship your entire codebase to the internet?
31
It's unclear if that was the root cause, and it's also possible that some unfortunate developer did this by accident, or perhaps some rogue developer did it on purpose.
32
We may never know the truth, but now it's time for the fun part.
33
What did we actually learn from the leak?
34
Well first, we learned that Claude uses Axios.
35
If you subscribe to my channel, then you know that Axios was compromised by North Korean hackers yesterday.
36
The exploit can install a remote access Trojan on anyone using this package, and in theory, if that happened on Anthropic servers, it could be a massive disaster.
37
But the next thing we learned is that Claude code is basically just a dynamic prompt sandwich glued together with TypeScript, and not some magical piece of futuristic technology.
38
In a basic AI chatbot, you typically have a hidden system prompt that gets combined with your prompt,
39
then the base model uses statistics to regurgitate a bunch of data it stole from the internet.
40
But in Claude code, things are far more complex, with a total of 11 steps from input to output.
41
Somebody already vibe-coded a website that breaks down every step, but the most interesting part about this codebase is
42
that it contains tons of hard-coded instructions and guardrails that basically beg Claude to please don't do anything weird.
43
Like there's just file after file of these massive hard-coded strings telling Claude to be a good boy.
44
And that's kind of surprising, because if this code were ever leaked, it would instantly turn from a black box into a blueprint for Claude's competition.
45
And ironically, that's exactly what happened.
46
What makes that even more funny, though, is that Claude was actively trying to stop their competition from copying Claude code by implementing anti-distillation poison pills.
47
It does that by pretending that certain tools exist, when in reality, they don't exist at all.
48
You're a big fat That means if you're some Chinese guy trying to train a new model on Claude's outputs, it's going to talk about tools that don't exist,
49
which will point your model in the wrong direction and just make it suck.
50
In reality, Claude Code only uses about 25 different tools or so, and now the Claude distillers know exactly what to look for, and they're likely going to have a field day with the Bash tool.
51
This file contains over a thousand lines of code that helps the large language model reliably parse and execute Bash commands,
52
which might be the single most important feature in an AI coding assistant.
53
The next thing we need to talk about, though, is undercover mode, which is a set of instructions that tell Claude to never mention itself in commit messages or outputs,
54
where the main idea is to make the outputs look as human as possible.
55
The stated purpose of this feature is to prevent things like model codename leaks, but many have speculated that the true purpose is more deceptive,
56
like they're trying to covertly use Claude in open-source projects so AI code doesn't get scrutinized when it breaks things catastrophically.
57
A very misanthropic idea, indeed.
58
But the irony continues.
59
Another funny thing found in the code is its regex frustration detector.
60
Your state-of-the-art AI model uses simple regular expression matching against your prompt to look for keywords like shit, fuck, cock, balls, and so on,
61
to determine if you're not having a good experience coding with Claude.
62
If it detects a match, it'll simply log an event.
63
The bottom line here is that we're not looking at some sort of alien superintelligence, but rather basic programming concepts that have been around for 50 years combined with a bunch of prompt spaghetti.
64
It's all just an illusion.
65
On top of that, this codebase has a ton of comments.
66
A lot more comments than you would typically find in a human-written codebase.
67
And what that tells us is that these comments aren't actually written for humans, but rather for the AI to write its own AI coding tool in an infinite loop.
68
But perhaps the biggest problem about this League 4 Anthropic is not the code itself, but rather the feature names and roadmap hidden within the code.
69
Like there's a hidden capability under a feature flag called Buddy,
70
which appears to be a new Tamagotchi-style companion that every developer can customize and raise like a little digital pet.
71
This might just be Anthropik's April Fool's Day joke, but there are also references to Opus 4.7 and a new model called Capybara,
72
which might be their new recently-teased Mythos model.
73
There's also things like Ultra Plan, Coordinator Mode, and Demon Mode, but perhaps the most interesting is Kairos,
74
which is a Greek word for an exact moment in time, or God's time.
75
I hate to beat off a dead horse here, but it's a bit ironic that Anthropic didn't get to reveal Kairos at the exact time it wanted to, and instead, God chose the right time.
76
The feature itself seems to be some kind of background agent that keeps a daily journal, uses dream mode to consolidate memories, and does work for you in the background on a specific schedule.
77
Pretty cool, but at the end of the day, this leak is a pretty huge setback for Anthropic, which hopes to IPO later this year and offload their bags to the retail public.
78
And it's yet another reminder that your top secret application is just one NPM publish away from becoming open source, whether you like it or not.
📺 Stesso Canale
B21
10 very promising Open Source Projects you haven’t heard of
American Accent, en-USUniversity Students+2
C11
Array Map in 100 Seconds
American Accent, en-USUniversity Students+2
B21
7 Programming myths that waste your time
American Accent, en-USUniversity Students+2
C15
The plan to break apart Google... RIP Chrome
American Accent, en-USUniversity Students+3
C11
DHH has gone completely off the rails...
American Accent, en-USUniversity Students+2
C12
DNS Explained in 100 Seconds
American Accent, en-USUniversity Students+2
B23
Anthropic begged the world to stop AI… then shipped this
American Accent, en-USUniversity Students+2
B22
One man just liberated Fable... and now it’s illegal
American Accent, en-USUniversity Students+2
B22
7 Things No Programmer Ever Wants to Hear
American Accent, en-USBusiness English+1
B21
80% of programmers are NOT happy… why?
American Accent, en-USUniversity Students+2
C14
The most interesting hack in history just got weirder...
American Accent, en-USUniversity Students+2
B24
Raspberry Pi Explained in 100 Seconds
American Accent, en-USUniversity Students+2
C12
Did Google just kickstart the intelligence explosion?
American Accent, en-USUniversity Students+3
C1100+ Docker Concepts you Need to Know
American Accent, en-USUniversity Students+2
C17
When open-sourcing your code goes wrong...
American Accent, en-USUniversity Students+2
C129
Docker in 100 Seconds
American Accent, en-USUniversity Students+2
C13
The mystery is solved... and the answer is 40x cheaper than Claude
American Accent, en-USUniversity Students+2
B28
An ex-OpenAI researcher just deleted language from the LLM...
American Accent, en-USUniversity Students+2
B29
25 crazy software bugs explained
American Accent, en-USUniversity Students+2
B29
Git Explained in 100 Seconds
American Accent, en-USUniversity Students+2
✨ Video consigliato
B21
629 frasi
Last Cheater Standing Wins $10,000!
American Accent, en-USUniversity Students+2
C11
309 frasi
Claude Code: The Advanced Guide (99% of Devs Skip These Features)
University StudentsBusiness English+1
![You’ll stop using ChatGPT after listening to this | Jonathan Pageau [ARC 2026]](https://img.youtube.com/vi/yZUuKzDQSsI/mqdefault.jpg)
C13
173 frasi
You’ll stop using ChatGPT after listening to this | Jonathan Pageau [ARC 2026]
University StudentsBusiness English+1
B22
190 frasi
He’s always wrong
University StudentsBusiness English+2
C18
55 frasi
When AIs act emotional
University StudentsBusiness English+1
C12
542 frasi
The Future Of Work (& The New High-Income Skill Stack)
American Accent, en-USUniversity Students+2
C12
68 frasi
Oh, wait, actually the best Wordle opener is not “crane”…
American Accent, en-USUniversity Students+2
C11
565 frasi
Garry Tan: Own Your Intelligence
University StudentsBusiness English+1
C11
853 frasi
Why Coding Agents Keep Making Your Codebase Worse
University StudentsBusiness English+1
C11
207 frasi
How I Made $30,000 Hacking Broken Access Control
University StudentsBusiness English+1
Informazioni su questa lezione
Cos'è la tecnica dello Shadowing?
Shadowing è una tecnica di apprendimento delle lingue supportata da studi scientifici, originariamente sviluppata per la formazione dei traduttori professionisti e resa popolare dal poliglotta Dr. Alexander Arguelles. Il metodo è semplice ma potente: ascolti un audio in inglese di madrelingua e lo ripeti immediatamente ad alta voce — come un'ombra che segue il parlante con un ritardo di solo 1–2 secondi. A differenza dell'ascolto passivo o degli esercizi di grammatica, lo shadowing costringe il tuo cervello e i muscoli della bocca a elaborare e riprodurre simultaneamente i modelli di discorso reale. La ricerca dimostra che migliora significativamente la precisione della pronuncia, l'intonazione, il ritmo, il discorso connesso, la comprensione dell'ascolto e la fluidità del parlato — rendendolo uno dei metodi più efficaci per la preparazione alla prova di speaking dell'IELTS e per la comunicazione reale in inglese.