Shadowing Practice: Episode 72 : Neuralk - AI for Structured Data - Learn English Speaking with Video

בחר מצב תרגול

כל מצב פותח חדר תרגול במסך מלא.

161 משפטים
אם המשפטים קצרים או ארוכים מדי, לחץ על ערוך כדי להתאים אותם.
1
Hey, everyone, and welcome back.
2
Today, we're going deep on a company that's tackling, well, a challenge you might not have even realized existed.
3
Yeah.
4
AI for the stuff that actually runs most businesses, structured data.
5
It's like, you know, everyone's obsessed with AI writing poems or making art, but this company is focused on the unsung hero of the data world.
6
Yeah.
7
The spreadsheets and databases we all rely on.
8
OK, so tell me more.
9
Yeah.
10
What exactly does this company do?
11
They're called Neuralc AI, and they're based in Paris, France.
12
Ooh la la.
13
Founded by Alexandre Pescu and Antoine Musnot.
14
Okay.
15
Their whole mission is to build AI specifically designed to understand and analyze structured data.
16
Okay.
17
Particularly in the commerce sector.
18
Paris, France.
19
That's kind of cool.
20
Yeah.
21
It adds a little mystique to the AI world, right?
22
It does.
23
But what's the problem they're solving?
24
Right.
25
Is analyzing spreadsheets really that hard?
26
It's not just about analyzing.
27
Okay.
28
It's about understanding.
29
Got it.
30
Most businesses are drowning in data, but they're only scratching the surface of what they can learn from it.
31
Yeah.
32
Traditional analytics can tell you what happened, but Neuroc .ai is building AI that can predict what will happen and help businesses make smarter decisions.
33
So it's like having a crystal ball for your business data.
34
Kind of, yeah.
35
Give me an example.
36
Okay.
37
What kind of problems can they solve with this AI?
38
Think about a massive online retailer with millions of products.
39
Just keeping their product catalog clean and consistent is a huge task.
40
I can imagine.
41
Imagine trying to make sure every product description is accurate, categorized correctly, and all the images are high quality.
42
It's a logistical nightmare.
43
Oh, I bet.
44
And humans probably make mistakes or things slip through the cracks.
45
Exactly.
46
That's where Neuralg AI comes in.
47
They've built models that can automatically clean up product catalogs,
48
standardize information, and even enrich existing data with additional attributes.
49
Wow.
50
They actually did this for a major European retailer called Akan Retail, and the results were impressive.
51
So instead of having a team of people manually combing through millions of product entries, they have AI doing the heavy lifting.
52
That's got to save them a ton of time and money.
53
Absolutely.
54
And it reduces errors, too.
55
And it reduces errors, too.
56
But it goes beyond just cleaning up data, right?
57
Right.
58
They're also using AI to personalize customer experiences.
59
Exactly.
60
Ooh, that's where it gets really interesting for the shoppers like me.
61
Yeah.
62
Tell me more about that.
63
By analyzing things like purchase history and browsing behavior, all structured data, by the way,
64
their AI can predict what products a customer is most likely to want and recommend them accordingly.
65
So it's like those, you might also like the gestures you see online, but powered by really sophisticated AI that understands your individual preferences.
66
Exactly.
67
And it's not just based on simple keyword matching.
68
Okay.
69
They have a model called a tabular foundation model that actually understands the underlying meaning of product attributes.
70
It's kind of like how some AI can understand images or text.
71
Right.
72
But this one is built for tabular data.
73
Okay.
74
The kind you find in spreadsheets and databases.
75
Wait, hold on.
76
Tabular foundation model.
77
That sounds pretty technical.
78
It is a little bit, yeah.
79
Do you break that down for me and the listeners?
80
Sure.
81
What makes it so special?
82
Think of it this way.
83
Okay.
84
Imagine a chef who can not only read a recipe that's like basic analytics.
85
Okay.
86
But also understand flavor profiles and predict how different ingredients will work together.
87
Yeah.
88
That's what this model can do with data.
89
I see.
90
It can see patterns and relationships that traditional methods would miss.
91
Okay, that's a great analogy.
92
Thanks.
93
So it's not just about crunching numbers.
94
Right.
95
It's about understanding the deeper meaning behind those numbers and using that knowledge to make predictions.
96
Exactly.
97
Okay, I'm starting to see how this could be a game changer for companies.
98
Yeah.
99
What other cool stuff have they done with this tabular foundation model?
100
Another example is in similarity searches.
101
Okay.
102
Imagine you're shopping online for a new pair of shoes.
103
Instead of just relying on basic keywords, this model can understand the underlying meaning of product attributes,
104
like style, color, material, and even brand reputation.
105
So it could suggest shoes that are similar in style, even if they don't share the exact same keywords in the description.
106
Yes.
107
It's about understanding the essence of what you're looking for, not just matching words.
108
Right.
109
This makes product discovery much more intuitive and helps customers find what they really want, even if they don't know the exact terms to use.
110
This is all very cool stuff.
111
Yeah.
112
And I can see how this kind of AI could be applied to so many different industries beyond just retail.
113
You're absolutely right.
114
Yeah.
115
Structured data is.. everywhere.
116
And the potential applications for this technology are vast.
117
Well, listeners, it sounds like Neurok AI is definitely a company to watch for them.
118
Where are they headed next?
119
Well, they recently secured $4 million in funding to further develop their technology, specifically their tabular foundation models.
120
That's a pretty significant investment.
121
What are they planning to do with that funding?
122
It seems they're staying laser focused on the commerce sector makes sense
123
which makes sense given their early success there yeah
124
they're also gearing up to launch the first official version of
125
their model oh cool very soon with a full benchmark release expected later this year oh a benchmark release
126
that sounds interesting what does
127
that mean for people like me who might not be you
128
know super technical think of it like a public test they'll be releasing their model
129
and data to researchers and developers so they can see how it performs compared to other AI models out there.
130
It's a way to prove the effectiveness of their technology and show the world what it can do.
131
So it's kind of like a competition for AI models.
132
In a way, yeah.
133
It's a chance for Neuralc AI to demonstrate how their approach to structured data AI stacks up against the competition.
134
This is all very exciting.
135
It sounds like Neuralc AI is really pushing the boundaries of what's possible with AI and data.
136
Absolutely.
137
They're at the forefront of a very important trend.
138
Yeah.
139
We're moving beyond the era of just collecting data and analyzing it in a basic way.
140
Now it's about using AI to truly understand the meaning behind that data and unlock its full potential.
141
It makes you wonder what the world will look like when AI can truly understand all types of data.
142
Right.
143
Not just structured data.
144
It's a fascinating question to consider.
145
Yeah.
146
Imagine a world where AI can seamlessly integrate information from spreadsheets, databases, text documents, images, even audio and video.
147
Wow.
148
The possibilities are almost endless.
149
That's pretty mind -blowing.
150
But for now, it sounds like Neuralc AI is making incredible strides in helping businesses make sense of the data they already have.
151
Exactly.
152
And their upcoming model launch and benchmark release will be key milestones to watch.
153
Right.
154
I'm very curious to see what kind of impact they'll have on the industry.
155
Me too.
156
I'm definitely adding Neuralc AI to my list of companies to keep an eye on.
157
Well, listeners, that brings us to the end of this deep dive into Neuralc AI.
158
But don't worry, the learning doesn't stop here.
159
We'll be back tomorrow evening with another exciting deep dive into a new topic.
160
Make sure you subscribe and hit that notification bell so you don't miss it.
161
Until then, keep exploring, keep learning, and keep asking those big questions.

הורד אפליקציה

כל מה שאתה צריך כדי לדבר בשטף

הגיית AIציון לכל משפט
תרגול IPAשלוט בכל צליל
אוצר מיליםבנה את מאגר המילים שלך
משחק אוצר מיליםלמד תוך כדי משחק
FOR YOU

המלץ על סרטון

אוצר מילים והערות הגייה לשיעור הזה

שיעור הדיבור הזה מבוסס על הסרטון “Episode 72 : Neuralk - AI for Structured Data”. המילים שחוזרות הכי הרבה: model, product, Neuralc, structured, exactly. בסרטון הזה יש 161 משפטים ו-1305 מילים לתרגול Shadowing. הדיבור נמשך 7:33. הדובר מדבר בקצב טבעי של כ-173 מילים בדקה, קרוב לשיחה יומיומית. 82% מהמילים נמצאות בין 3,000 המילים הנפוצות ביותר באנגלית; כדאי ללמוד את השאר לפני שמתחילים.

אוצר מילים מרכזי בסרטון הזה

15 מילים מהסרטון שכדאי ללמוד, עם ההגייה והמשמעות שלהן:

מילההגייהמשמעות
database שם עצם/ˈdeɪtəˌbeɪs/בְּסִיס נְתוּנִים
attribute שם עצם/ˈæt.ɹɪˌbjut/לוואי, לְוַאי
listener שם עצם/ˈlɪs(ə)nɚ/מַאֲזִין
solve פועל/sɒlv/פָּתַר
commerce שם עצם/ˈkɑ.mɚs/סַחַר, מִסְחָר
shoe שם עצם/ˈʃu/נַעַל
dive פועל/daɪv/צָלַל
topic שם עצם/ˈtɑpɪk/נוֹשֵׂא
consistent שם תואר/kənˈsɪs.tənt/עקבי ikví
audio שם תואר/ˈɔ.di.oʊ/שמע
impressive שם תואר/ɪmˈpɹɛsɪv/מרשים
reputation שם עצם/ˌɹɛpjʊˈteɪʃən/מוֹנִיטִין
compare פועל/kəmˈpɛɚ/השווה \ הִשְׁוָה
curious שם תואר/ˈkjʊɹ.i.əs/סַקְרָן, סקרנית
discovery שם עצם/dɪˈskʌv.ɹi/תַּגְלִית

הגייה ששווה לשים לב אליה

הדובר משתמש ב-25 קיצורים וצורות מקוצרות, כמו they're, don't, I'm. אמרו אותם בצורה הקצרה, כפי שאתם שומעים אותם.

  • הצלילים של “sh” ו-“zh”: benchmark /ˈbɛn(t)ʃmɑːk/, shoe /ˈʃu/, reputation /ˌɹɛpjʊˈteɪʃən/, chef /ʃɛf/, prediction /pɹɪˈdɪkʃən/
  • מילים ארוכות — שימו לב להטעמה: analytics /ˌæn.əˈlɪt.ɪks/, reputation /ˌɹɛpjʊˈteɪʃən/, developer /dɪˈvɛləpɚ/, effectiveness /ɪˈfɛk.tɪv.nɪs/, accordingly /əˈkɔɹ.dɪŋ.li/

צלילים שדוברי עברית מתקשים בהם:

  • /æ/ — פתוח יותר מה-“אֶ” העברית, לא כמו “a” ב-“cup”: analyze /ˈæn.əˌlaɪz/, attribute /ˈæt.ɹɪˌbjut/, analytics /ˌæn.əˈlɪt.ɪks/, vast /væst/, crack /kɹæk/
  • /r/ — הלשון לא נוגעת בשום מקום, לא ה-“ר” הגרונית: predict /pɹɪˈdɪkt/, attribute /ˈæt.ɹɪˌbjut/, rely /ɹɪˈlaɪ/, retailer /ˈɹiteɪlɚ/, impressive /ɪmˈpɹɛsɪv/
  • עיצור קולי בסוף מילה — /b/, /d/, /g/, /z/, /v/ בלי להפוך אותם לאטומים: analyze /ˈæn.əˌlaɪz/, solve /sɒlv/, dive /daɪv/, impressive /ɪmˈpɹɛsɪv/, subscribe /səbˈskɹaɪb/

איך לתרגל עם הסרטון הזה

  1. הקשיבו לכל הסרטון פעם אחת בלי לדבר, ורשמו את המילים שאתם לא מכירים.
  2. התחילו במהירות 0.75×, תרגלו Shadowing משפט אחר משפט, וחזרו למהירות רגילה כשזה מרגיש קל.
  3. הקליטו את עצמכם והשוו למקור, תוך תשומת לב למילים כמו database, attribute, listener.

מהי טכניקת השאדואינג?

שאדואינג היא טכניקת למידת שפה מבוססת מדע שפותחה במקור להכשרת מתורגמנים מקצועיים ופופולרית על ידי הפוליגלוט ד"ר אלכסנדר ארגואלס. השיטה פשוטה אך עוצמתית: אתה מקשיב לאודיו באנגלית ילידית וחוזר עליו בקול רם מיד — כמו צל העוקב אחרי הדובר עם עיכוב של 1–2 שניות בלבד. בניגוד להאזנה פסיבית או תרגילי דקדוק, שאדואינג מכריח את המוח והשרירים בפה לעבד ולשחזר דפוסי דיבור אמיתיים בו זמנית. מחקרים מראים שזה משפר משמעותית את דיוק ההגייה, האינטונציה, הקצב, הדיבור המחובר, הבנת השמיעה ושטף הדיבור — מה שהופך אותו לאחת השיטות היעילות ביותר להכנה לIELTS Speaking ולתקשורת באנגלית בעולם האמיתי.

טכניקת Shadowing: קרא את המדריך המלא שלב אחר שלב →