تدريب Shadowing: The End of Human Mathematics? AI's Recent Breakthrough -- And What's Next - تعلم التحدث بالإنجليزية عبر الفيديو
جارٍ إنشاء الدرس...
1
Maths is cooked!
2
I heard this recently on a podcast, and I agree.
3
I also agree that heard it on a podcast is not the greatest argument, but there's more to it,
4
including the question of why everyone's suddenly so interested in solving obscure maths problems that no one had heard of before,
5
or may I try to convince you it's the biggest thing ever.
6
Today I have a brief update on what AI has been doing to maths recently,
7
but mostly I want to look at how people have reacted to it and what's likely to come next.
8
Even just at the present pace, we're talking about essentially all Frontier Math Tier 4,
9
all professional research -grade math problems being solved in the next four or five years.
10
So math is cooked.
11
I'll say it, you know, second time, math is cooked.
12
That was physicist and computer scientist Alexander Wisner -Gross commenting on the AI takeover of maths.
13
He's not the only one who sees it coming, and things are moving quickly now.
14
In the past four years, large language models have progressed in maths extremely fast.
15
They've basically gone from not being able to add 2
16
and 2 to solving maths Olympia problems to solving problems that no one cares about.
17
Then, two weeks ago, for the first time, a large language model solved a reasonably well -known maths problem.
18
The model in question is not a publicly available one.
19
It's an internal OpenAI model, which is somewhat disappointing because we can't try it, and it also makes you wonder where OpenAI is going with this.
20
The problem the model solved was posed by Paul Erdős 80 years ago.
21
The solution made big headlines primarily, I think, because it's a case where the problem is actually understandable.
22
If you put points in a plane, how do you have to place them so that you have a maximal number of pairs with the same distance between them?
23
Erdős conjecture, that'd be a square lattice, but he didn't prove it.
24
The OpenAI model disproved Edo's conjecture with a counterexample.
25
Not just that, it allegedly one -shotted it on the first attempt.
26
OpenAI also checked how often the model arrived at the correct answer in multiple test runs.
27
With enough computing time, they found that the model got it right about 50 % of the time.
28
So it's clear that the model is still not any good at checking its own logic, but it's also clearly not just randomly outputting strings that look mathy,
29
which is more than can be said for some conference abstracts I've read.
30
Here are some recent reactions from mathematicians.
31
I find this existentially upsetting.
32
I see myself shriveling to nothing.
33
I'm saddened.
34
It may be the case that within the decade reasoning models will outperform the greatest living mathematicians.
35
We're cooked.
36
This is the best proof to come out in a year.
37
We're completely fucked.
38
And the computer scientist Scott Aronson relates,
39
I heard the news maybe an hour after it broke when some UT grad students came to my office to tell me.
40
For what it's worth, these students were morose, musing about how everything might soon be over for young scientists and mathematicians like themselves.
41
This is what makes academia special.
42
In industry, people worry that AI will take their jobs.
43
In academia, they worry that AI will take their jobs before they even get jobs.
44
But what I find most interesting about this recent proof is that, for all I can tell,
45
most mathematicians thought that a Deutsch conjecture was correct.
46
And the counterexample which the AI came up with is the kind of unintuitive, slightly ugly construction that humans are unlikely to think of.
47
It's not that the machine is smarter, but that it has worse taste.
48
That said, rumors of the demise of mathematics have been somewhat exaggerated because it took only a few hours until someone, and I mean a human being,
49
improved the Oak May Eye result.
50
There's also been a lot of pushback of the idea that maths is over.
51
A group of 150 mathematicians just signed a declaration warning
52
that these developments put the autonomy of mathematics under threat
53
and there's currently a strong commercial incentive on the part of the technology industry to overstate the capabilities of their products.
54
The latter quite plausibly answers the question of why the US media makes such a big deal of this
55
when usually they can't be bothered writing about maths at all.
56
It's a kind of sabre -retting for the Chinese, basically.
57
Personally, I think what's.. going to happen is not that mathematics will end.
58
What we'll see instead is that a lot of mathematicians will shift to AI -supported techniques.
59
This is because the existing mathematical literature contains sufficient information for a lot of conclusions,
60
proofs even, that are within reach for someone or something
61
that can rapidly digest and analyze a lot of data and go through many attempts at proofs quickly.
62
There must be many low -hanging fruits to pick, like this recent counterexample, and we're going to see more of this.
63
The open question is whether AI can go beyond extracting patterns in human methods
64
and filling in the holes in the literature to develop its own methods.
65
I continue to insist that large language models are not going to get us there.
66
So yes, maths may be cooked, but only partially, more like ardente.
67
One -shotting an Erdős problem is of course impressive, but wait until they can explain to normal people why they show care.
68
A few years ago, I received some ugly letters from a debt collection company.
69
They threatened to sue me for ordering something online and then not paying it.
70
Turned out that someone had stolen my private data.
71
I managed to get out of this, but since then I've become very protective of my personal information, which is why I've signed up to Incogni,
72
who've been sponsoring this video.
73
Each time you open a website, it'll try to collect data about who you are and where you are and what other websites you've visited.
74
Many firms then make money by selling your private information to data brokers.
75
Most countries have laws against that and you can ask for your data to be removed, but doing this takes up a lot of time.
76
Incogni automates the process of getting you out of those databases.
77
You sign up and they'll contact the big sinners, request that your personal details be removed.
78
They'll keep on doing that and, if you want, send you updates about the progress they're making.
79
They now also have a custom removal feature where you can submit specific websites.
80
I'll be using this a lot.
81
Incogni has really made my life better.
82
If that sounds like something you could need to, use my link Sabine or the custom link in the info
83
because the first 100 people to use it will get 60 % off.
84
Stay safe.
85
Thanks for watching.
86
See you tomorrow.
📺 نفس القناة
C16
Superfluid Dark Matter
IELTS PrepTOEFL PrepUniversity Students
C14
Make AI Believe in God to Keep Humans Safe, Scientist Argues
IELTS PrepTOEFL PrepUniversity Students
C1Has quantum mechanics proved that reality does not exist?
IELTS PrepTOEFL PrepUniversity Students
C13
FINALLY: A Proof that Particles Take All Paths At Once?
IELTS PrepTOEFL PrepUniversity Students
C1The Uncertainty Principle: What Does It Mean, How Does It Work?
IELTS PrepTOEFL PrepUniversity Students
C1Mathematician Collapses All Functions to One Weird Formula
Business EnglishUniversity StudentsWorking Adults
C13
AI Will End Philosophy, Physicist Claims
IELTS PrepTOEFL PrepUniversity Students
C1We Can Now Simulate a Human Brain, Scientists Show
IELTS PrepTOEFL PrepUniversity Students
C19
Does God exist? Science does not have an answer.
IELTS PrepTOEFL PrepUniversity Students
C15
What is time?
IELTS PrepTOEFL PrepUniversity Students
C1Our Solar System Lost Two Planets!
IELTS PrepTOEFL PrepUniversity Students
C1I was SO wrong about quantum computing.
IELTS PrepBusiness EnglishUniversity Students
✨ فيديو موصى به
C12
182 جمل
We Let $200 GPT-Pro Think for 7 Days on Unsolved Math
Business EnglishUniversity StudentsWorking Adults
C11
85 جمل
AI Can Write Code. And That’s the Problem.
Business EnglishUniversity StudentsWorking Adults
C11
74 جمل
AI & Inequality | Daron Acemoglu: Big Tech poses risks as AI reshapes society
IELTS PrepBusiness EnglishUniversity Students
B21
261 جمل
GMA First Look Full Broadcast - Sept. 9, 2026
Business EnglishDaily ConversationUniversity Students
C112
47 جمل
Essence of linear algebra preview
American AccentBusiness EnglishUniversity StudentsWorking Adults
C12
155 جمل
How to be a creative thinker | Carnegie Mellon University Po-Shen Loh
Business EnglishUniversity StudentsWorking Adults
C11
407 جمل
Why AI Tokens are so Expensive - Computerphile
Business EnglishUniversity StudentsWorking Adults
C13
54 جمل
Did Google just kickstart the intelligence explosion?
American AccentBusiness EnglishDaily ConversationUniversity Students
C12
86 جمل
The Map of Mathematics
British AccentBusiness EnglishUniversity StudentsWorking Adults
C18
102 جمل
World Models explained in 10min..
Business EnglishUniversity StudentsWorking Adults
C11
409 جمل
I think AI slop is about to trigger a renaissance, actually
Business EnglishUniversity StudentsWorking Adults
المفردات وملاحظات التحدث لهذا الدرس
يحتوي هذا الفيديو على 86 جملة و1168 كلمة للتدرب على الترديد. مدة الكلام 7:35. يتكلم المتحدث بسرعة طبيعية، نحو 154 كلمة في الدقيقة، قريبة من المحادثة اليومية. 80% فقط من الكلمات ضمن أكثر 3,000 كلمة شيوعًا في الإنجليزية، فالمفردات صعبة نسبيًا.
أهم المفردات في هذا الفيديو
15 كلمة من الفيديو تستحق التعلّم، مع النطق والمعنى:
| الكلمة | النطق | المعنى |
|---|---|---|
| mathematics اسم | /mæθ(.ə)ˈmæt.ɪks/ | رِيَاضَة, رِيَاضِيَّات |
| solve فعل | /sɒlv/ | ذَابَ, حَلَّ |
| scientist اسم | /ˈsaɪəntəst/ | عَالِم, عَالِمَة |
| custom اسم | /ˈkʌstəm/ | عَادَة, عُرْف |
| ugly صفة | /ˈʌɡli/ | قَبِيح, بَشِع |
| podcast اسم | /ˈpɑdˌkæst/ | بُودْكَاسْت m پُودْكَاسْت, بَثّ صَوْتِيّ |
| technique اسم | /tɛkˈniːk/ | أُسْلُوب, تِقْنِيَة |
| output اسم | /ˈaʊtpʊt/ | مُخرَج |
| steal فعل | /stiːl/ | سَرَقَ |
| string اسم | /stɹɪŋ/ | وَتَر |
| bother فعل | /ˈbɑðɚ/ | أزْعَجَ |
| arrive فعل | /əˈɹaɪv/ | وَصَلَ |
| collect فعل | /ˈkɑlɪkt/ | جَمَعَ |
| database اسم | /ˈdeɪtəˌbeɪs/ | قَاعِدَة بَيَانَات |
| unlikely صفة | /ʌnˈlaɪkli/ | مِنْ غَيْرِ المُرَجَّحِ |
الأفعال المركبة التي ستسمعها
| الكلمة | المعنى |
|---|---|
| come up with فعل | جَاءَ بِـ, أَتَى بِـ |
نطق يحتاج إلى انتباه
يستخدم المتحدث 22 من الصيغ المختصرة، مثل we're, I've, can't. انطقها بصيغتها القصيرة كما تسمعها.
- أصوات “th”: mathematics /mæθ(.ə)ˈmæt.ɪks/, bother /ˈbɑðɚ/, mathematical /ˌmæθ(.ə)ˈmæt.ɪ.kəl/, threaten /ˈθɹɛt.n̩/, mathematician /ˌmæθ.(ə.)məˈtɪʃ.ən/
- أصوات “sh” و“zh”: essentially /ɪˈsɛnʃəli/, conclusion /kənˈkluːʒən/, sufficient /səˈfɪʃənt/, partially /ˈpɑɹʃəli/, declaration /ˌdɛk.ləˈɹeɪ.ʃən/
- الكلمات الطويلة — انتبه لموضع النبر: mathematics /mæθ(.ə)ˈmæt.ɪks/, academia /ˌæk.əˈdi.mi.ə/, primarily /ˈpɹaɪ.mə.ɹə.li/, essentially /ɪˈsɛnʃəli/, declaration /ˌdɛk.ləˈɹeɪ.ʃən/
كيف تتدرب بهذا الفيديو
- استمع إلى الفيديو كاملًا مرة واحدة دون أن تتكلم، ودوّن الكلمات التي لا تعرفها.
- ابدأ بسرعة 0.75×، وردّد جملة جملة، ثم عُد إلى السرعة العادية عندما يصبح الأمر سهلًا.
- سجّل صوتك وقارنه بالأصل، مع الانتباه إلى كلمات مثل mathematics, solve, scientist.
القواعد في هذا الفيديو
أكثر التراكيب التي يستخدمها المتحدث، مع العبارات نفسها من الفيديو:
| التركيب | في الفيديو |
|---|---|
| المضارع التام المستمر have/has been + -ing — حدث بدأ سابقًا وما زال مستمرًا | has been doing · who've been sponsoring |
| المبني للمجهول be + اسم المفعول — التركيز على ما يحدث لا على من يفعله | is cooked · being solved · was posed |
| المضارع التام have/has + اسم المفعول — حدث ماضٍ ما زال له أثر الآن | have reacted · have progressed · I've read |
| جمل الوصل who / which + جملة — معلومة إضافية عن شخص أو شيء | one who sees · model, which is · information, which is |
ما هي تقنية التظليل الصوتي؟
التظليل الصوتي (Shadowing) تقنية تعلم لغة مدعومة علمياً، طُورت أصلاً لتدريب المترجمين الفوريين المحترفين. الطريقة بسيطة لكنها قوية: تستمع لصوت إنجليزي أصلي وتكرره فوراً بصوت عالٍ — كظل يتبع المتحدث بتأخير 1-2 ثانية. تُظهر الأبحاث تحسناً كبيراً في دقة النطق والتنغيم والإيقاع وربط الأصوات والاستماع والطلاقة.