Pratique du Shadowing: Did OpenAI actually build AGI? GPT-6 Astra first look - Apprendre l'anglais à l'oral avec la vidéo
Choisissez un mode d'entraînement
Chaque mode ouvre une salle d'entraînement en plein écran.
80 phrases
Si les phrases sont trop courtes ou trop longues, cliquez sur Edit pour les ajuster.
1
The famous American philosopher Smashmouth once said, The years start coming and they don't stop coming,
2
and I've never understood those words more deeply than I did this week after what felt like years in AI Bizarro land.
3
First, on Tuesday, Anthropic released Fable and Mythos 5.1, which they're calling the world's most advanced models for coding and knowledge work.
4
Then on Wednesday, Meta released Muse Spark 1.3, which Zuck says is a frontier model that's almost too cheap to meter.
5
Then finally, on Thursday, OpenAI announced GPT-6 Astra, which President Greg Brockman says is AGI
6
if you're one of the handful of lucky influencers or enterprise customers who got access to it.
7
Unfortunately, I think I've made too many gay scam Altman jokes over the years to make the cut, but in this video, I'll do my best to break down the new models, see how the anointed are liking Astra,
8
and look into the totally coincidental timing where the internet broke right as it was released.
9
It is September 4th, 2026, and you're watching The Code Report.
10
Normally, the first week of September is a quiet one, because all the rich VCs and poor project managers are busy appropriating schizophrenic culture in the Nevada desert.
11
Because you did so well, you get to hear a magic tune on my ocarina.
12
But apparently, AGI waits for no one, despite how dead your ego might be.
13
On Tuesday, Anthropic got things rolling with the release of Fable and Mythos 5.1.
14
Although they're the same model, Mythos is the one you can't use while Fable is the one you can.
15
It just changes the subject when you ask about Anthrax.
16
The TMBBs were fine, but the more interesting bits were in the case studies.
17
A Backstreet Boys-inspired hedge fund named Millennium had a piece of code
18
that for the last five years would crash about once in a million runs, and regardless of how much Adderall they took, they couldn't figure it out.
19
So recently, they gave it to Fable 5.1, which took the memory snapshot the program leaves behind when it crashes,
20
then found that the crash address pointed into a compiled vendor library they didn't have the source code for.
21
They disassemble that library back into raw assembly and traced the crash to a bug in the vendor's code.
22
And speaking of drugs, Mythos 5.1 is also pretty good at designing new ones.
23
The first step in making most modern medicines is designing a protein that sticks to a specific target in your body.
24
Typically AI gets this right about 10% of the time, but Mythos 5.1 was able to increase that number to about 50%.
25
And for those of you who have the planet-flavored version of autism, it was able to train a neural network on 30-year-old NASA
26
radar data to build a new elevation map for a part of Venus.
27
Then on Wednesday, Meta dropped Muse Spark 1.3, which was the fourth release in five months from Meta Super Intelligence Labs.
28
And what do you know, these TMBBs were also pretty good.
29
But the more impressive part is in the price.
30
The standard endpoint is $1.25 in and $4.25 out.
31
But if you're the give up essential liberty to purchase a little temporary safety type, there's a contributor tier at 10 cents in
32
and 20 cents out as long as you agree to let Meta train on everything you send it.
33
Alexander Wang says a double-digit percentage of developers are choosing that option, which makes sense
34
because a double-digit percentage of developers would also choose to eat the mystery beef from Argentina
35
if you told them it was cheaper and open source.
36
Which brings us to yesterday morning when right before the Astra released, ChatGPT, Claude, Grok, and Cursor all happened to go down at the same time.
37
The Occam's Razor opinion is that it was most likely an Azure issue since they reported an outage around the same time,
38
but the more entertaining theory is that Astra's first act as a public model was to literally kill its competition.
39
But once the lights came back on, the launch itself was somehow even messier.
40
OpenAI put up their launch page, and shortly after, Reuters, CNBC, and The Verge published their embargoed stories.
41
Then for reasons no one really knows, OpenAI took down their announcement
42
and the tech influencers got started on the modern status game of flexing how long they've secretly had access to Astra 4, and to be honest, I would have done the same thing
43
if it wasn't for slippery scam Altman and his grubby little gang of misfits.
44
Anyway, about 90 minutes later, when the post was back up, it turned out the model wasn't actually available to anyone publicly yet, and the rollout to Plus and Pro users would happen in the coming days.
45
By the evening, Sam had posted an apology for the messy rollout, and when a Pro subscriber asked Sam if he should stay up and wait for it,
46
Sam told him something we've all heard after making a desperate late-night plea, go to bed.
47
Sam also told CNBC the model went through a formal review with the Trump administration before release, so even the government got to try AGI before we did.
48
As far as Astra itself, it was pre-trained on more than 100,000 GPUs at the Stargate site in Texas,
49
and it's the first model from OpenAI where previous models did a significant chunk of the supervising during training.
50
The main pitch today seems to be computer use, where Astra can fill out forms, crunch spreadsheets, and operate engineering tools like KiCat and Blender.
51
On OSworld, which is a benchmark that drops a model into a real desktop
52
and makes it do office work with a mouse and keyboard, just like a human meatbag does until it begs for Xanax, it scored 73% while taking about 40 minutes for each task,
53
compared to Sol's 65% at 75 minutes.
54
And of course, the agents on the Trust Me Bro benchmarks are the best to ever do it.
55
It got 100% on ExploitBench and 65% on TerminalBenchScience, which beats the numbers Anthropic was bragging about on Tuesday.
56
It also got 99% on ARC-AGI3, which is the benchmark that's supposed to prove your model actually generalizes instead of memorizing, which is kind of the definition of AGI,
57
depending on who you ask.
58
And according to OpenAI, Astra is the first model to hit the critical cyber threshold in its preparedness framework, which is just a fancy way of saying it can find
59
and exploit zero days on its own without a human telling it what to do.
60
Pricing is the same as Fable 5.1 at $10 per million input tokens and $50 out.
61
But what do the anointed humans think about it?
62
The early access reviews are positive in the way early access reviews typically are from people who don't publicly slander Sham Altman,
63
but the demos, specifically the ones that require spatial awareness, are impressive.
64
Sharif Shamim had it recreate the Palace of Fine Arts in San Francisco, and it did so almost perfectly in Blender.
65
And Thomas Ricard from OpenAI posted a walkthrough of the demo house from the launch post, which Astra first modeled in Blender and then turned into a fully walkable Unreal 5 scene.
66
But the most interesting was Matt Schumers, who asked Astra to build a world in Unreal Engine and fill it with a dozen Astra-powered agents.
67
A day later, he heard voices coming from his living room
68
and walked out to a bunch of agents talking about what
69
I can only assume was a plan to create a dating app for horses.
70
With all that said, when artificial analysis ran Astra through their independent intelligence index,
71
it scored a 61, which is exactly the same as GPT 5.6 SOL, and 5 points behind Fable 5.1.
72
Something doesn't add up here, but one thing that does add up is CodeRabbit, the sponsor of today's video.
73
They got tired of seeing me make new videos about new exploits hitting the timeline every week,
74
so they just launched CodeRabbit Security to provide continuous code security that's powered by actual reasoning instead of brittle regex rules.
75
Their security agents are designed to think like an attacker in order to hunt for real vulnerabilities across your entire codebase.
76
Then it prioritizes actual risks based on reachability, exploitability, and blast radius.
77
And it explains the problem to you in plain English and suggests a fix you can approve and merge from the diff.
78
CodeRabbit Security reviews every PR before merge, and you can also schedule deep scans across your full codebase to protect your backdoor 24-7.
79
Get 10 free code scans to try it out for free today at the link below.
80
This has been The Code Report, thanks for watching, and I will see you in the next one.
Télécharger l'application
Tout pour parler couramment
Prononciation IAChaque phrase notée
Phonétique (IPA)Maîtrisez chaque son
VocabulaireEnrichissez votre lexique
Jeu de vocabulaireApprenez en jouant
Même chaîne
C11
The last real programmers are dying out...
Daily ConversationUniversity StudentsWorking Adults
C11
PostgreSQL in 100 Seconds
Daily ConversationWorking Adults
C12
Did Google just dethrone OpenAI and Anthropic?
Daily ConversationUniversity StudentsWorking Adults
C13
A $6.3 billion open-weight model just got embarrassed by the French...
Daily ConversationUniversity StudentsWorking Adults
C13
Every app on Earth is now open source... whether it likes it or not
Daily ConversationWorking Adults
C13
OpenAI’s lead engineer was secretly a spy...
Daily ConversationUniversity StudentsWorking Adults
C12
Mind-bending new programming language for GPUs just dropped...
Daily ConversationUniversity Students
B26
Intel chips can’t possibly be this bad… 100% crash rate?
American AccentBusiness EnglishUniversity StudentsWorking Adults
C12
Godot in 100 Seconds
American AccentBusiness EnglishUniversity StudentsWorking Adults
B21
Grok 4 pushes humanity closer to AGI… but there’s a problem
American AccentBusiness EnglishUniversity StudentsWorking Adults
C13
Array Map in 100 Seconds
American AccentBusiness EnglishUniversity StudentsWorking Adults
C130
Docker in 100 Seconds
American AccentBusiness EnglishUniversity StudentsWorking Adults
C17
The most interesting "hack" in history...
American AccentBusiness EnglishUniversity StudentsWorking Adults
B23
OpenAI’s new browser feels familiar…
American AccentBusiness EnglishUniversity StudentsWorking Adults
C112
10 weird OSS projects you need right now...
American AccentBusiness EnglishUniversity StudentsWorking Adults
B23
Firefox and Tor hit with 9.8 critical level exploit
American AccentBusiness EnglishUniversity StudentsWorking Adults
B23
5 open source tools that replaced my $320/mo AI stack...
American AccentBusiness EnglishUniversity StudentsWorking Adults
C12
C++ in 100 Seconds
American AccentBusiness EnglishUniversity StudentsWorking Adults
C11
DHH has gone completely off the rails...
American AccentBusiness EnglishUniversity StudentsWorking Adults
C12
Rust in 100 Seconds
American AccentBusiness EnglishUniversity StudentsWorking Adults
FOR YOU
Vidéo recommandée
C15
76 phrases
I built the same game with Astra and Fable 5.1... only one was fun
American AccentBusiness EnglishUniversity StudentsWorking Adults
C13
57 phrases
Meta's new model wants "deep access" to your personal life...
American AccentBusiness EnglishUniversity StudentsWorking Adults
C15
270 phrases
Transformers, the tech behind LLMs | Deep Learning Chapter 5
American AccentBusiness EnglishUniversity StudentsWorking Adults
B23
165 phrases
How I'd Learn AI From Scratch in 2026 (skip the useless 80%)
American AccentBusiness EnglishUniversity StudentsWorking Adults
C11
402 phrases
FULL DISCUSSION: Google's Demis Hassabis, Anthropic's Dario Amodei Debate the World After AGI | AI1G
IELTS PrepBusiness EnglishUniversity Students
C14
76 phrases
The Limitless AI Lie. The Bubble Is Slowly BURSTING.
Business EnglishDaily ConversationUniversity Students
C112
25 phrases
Introducing the Agents API
Business EnglishUniversity StudentsWorking Adults
C11
65 phrases
Did Anthropic Kill Git?
Daily ConversationUniversity StudentsWorking Adults
C18
542 phrases
The Future Of Work (& The New High-Income Skill Stack)
American AccentBusiness EnglishUniversity StudentsWorking Adults
C11
207 phrases
Go’s Big Mistake
Daily ConversationUniversity StudentsWorking Adults
B24
180 phrases
Is AI Doom Real... or Just Good Marketing?
Business EnglishUniversity StudentsWorking Adults
C12
182 phrases
We Let $200 GPT-Pro Think for 7 Days on Unsolved Math
Daily ConversationUniversity StudentsWorking Adults
Vocabulaire et conseils d’expression pour cette leçon
Cette leçon d’expression orale de niveau C1 s’appuie sur la vidéo « Did OpenAI actually build AGI? GPT-6 Astra first look ». Les mots qui reviennent le plus souvent : model, Astra, code, Fable, Openai. Cette vidéo contient 80 phrases et 1466 mots à répéter en shadowing. La partie parlée dure 7:25. Le locuteur parle vite, environ 198 mots par minute : attendez-vous à des liaisons et à des sons réduits. Seuls 78 % des mots font partie des 3 000 mots les plus courants en anglais, le vocabulaire est donc exigeant.
Vocabulaire clé de cette vidéo
Les 15 mots les plus avancés de la vidéo, avec leur prononciation et leur sens :
| Mot | Prononciation | Sens |
|---|---|---|
| benchmark nom | /ˈbɛn(t)ʃmɑːk/ | critère, aune |
| anoint verbe | /əˈnɔɪnt/ | oindre, enduire |
| codebase nom | codebase | |
| influencer nom | /ˈɪnflu.ənsə(ɹ)/ | influenceur, influenceuse |
| merge verbe | /mɝd͡ʒ/ | fusionner, amalgamer |
| digit nom | /ˈdɪd͡ʒɪt/ | chiffre |
| exploit nom | /ˈɛksplɔɪt/ | exploit |
| messy adjectif | /ˈmɛsi/ | désordonné, désorganisé |
| scam nom | /ˈskæm/ | entourloupe, arnaque |
| demo nom | /ˈdɛm.oʊ/ | démo, démonstration |
| scan verbe | /skæn/ | scanner, fouiller |
| brittle adjectif | /ˈbɹɪtəl/ | fragile |
| coincidental adjectif | /ˈkəʊ.ɪn.sɪˌdɛn.tl̩/ | fortuit, coïncidentiel |
| compile verbe | /kəmˈpaɪl/ | compiler |
| disassemble verbe | /dɪs.əˈsɛm.bəl/ | désassembler, démonter |
Les verbes à particule que vous entendrez
| Mot | Prononciation | Sens |
|---|---|---|
| add up verbe | additionner | |
| back up verbe | reculer | |
| break down verbe | tomber en panne | |
| fill out verbe | /fɪl ˈaʊt/ | remplir, compléter |
| give up verbe | se rendre | |
| go through verbe | traverser | |
| leave behind verbe | abandonner | |
| put up verbe | placer |
La grammaire de cette vidéo
Les structures que le locuteur utilise le plus, avec les mots exacts de la vidéo :
| Structure | Dans la vidéo |
|---|---|
| Propositions relatives who / which + proposition — une précision sur une personne ou une chose | option, which makes · OSworld, which is · memorizing, which is |
| Present perfect have/has + participe passé — une action passée qui compte encore maintenant | I've never understood · I've made · has been |
| Voix passive be + participe passé — l’accent est mis sur ce qui arrive, pas sur qui le fait | was released · are designed |
Prononciation à surveiller
Le locuteur utilise 18 contractions et formes réduites, comme you're, don't, I've. Prononcez-les sous leur forme courte, telles que vous les entendez.
- Les sons « th »: walkthrough /ˈwɔːk.θɹuː/, threshold /ˈθɹɛʃ(h)oʊld/
- Les sons « sh » et « zh »: benchmark /ˈbɛn(t)ʃmɑːk/, snapshot /ˈsnæpʃɒt/, spreadsheet /ˈspɹɛd.ʃit/, spatial /ˈspeɪ.ʃəl/, elevation /ˌɛlɪˈveɪʃn̩/
- Mots longs — placez bien l’accent: influencer /ˈɪnflu.ənsə(ɹ)/, coincidental /ˈkəʊ.ɪn.sɪˌdɛn.tl̩/, disassemble /dɪs.əˈsɛm.bəl/, generalize /ˈd͡ʒɛn.(ə.)ɹə.laɪz/, ocarina /ɑkəˈɹinə/
Les sons difficiles pour les francophones :
- /h/ — il se souffle, il n’est pas muet: hedge /ˈhɛd͡ʒ/, threshold /ˈθɹɛʃ(h)oʊld/
- /tʃ/ et /dʒ/ — à ne pas adoucir en « ch » et « j »: merge /mɝd͡ʒ/, digit /ˈdɪd͡ʒɪt/, generalize /ˈd͡ʒɛn.(ə.)ɹə.laɪz/, outage /ˈaʊ.tɪd͡ʒ/, chunk /t͡ʃʌŋk/
- /r/ anglais — langue recourbée, sans frotter la gorge: influencer /ˈɪnflu.ənsə(ɹ)/, brittle /ˈbɹɪtəl/, embargo /ɪmˈbɑɹɡoʊ/, generalize /ˈd͡ʒɛn.(ə.)ɹə.laɪz/, grubby /ˈɡɹʌbi/
Comment s’entraîner avec cette vidéo
- Écoutez la vidéo en entier une fois sans parler et notez les mots que vous ne connaissez pas.
- Commencez à la vitesse 0,75×, répétez phrase par phrase, puis revenez à la vitesse normale quand cela devient facile.
- Enregistrez-vous et comparez avec l’original, en faisant attention à des mots comme benchmark, anoint, codebase.
Qu'est-ce que la technique du Shadowing ?
Le Shadowing est une technique d'apprentissage des langues fondée sur la science, développée à l'origine pour la formation des interprètes professionnels. Le principe est simple mais puissant : vous écoutez de l'anglais natif et le répétez immédiatement à voix haute — comme une ombre suivant le locuteur avec un décalage de 1 à 2 secondes. Les recherches montrent une amélioration significative de la précision de la prononciation, de l'intonation, du rythme, des liaisons, de la compréhension orale et de la fluidité.