Would you like to inspect the original subtitles? These are the user uploaded subtitles that are being translated:
1
00:00:00,005 --> 00:00:03,001
- [Narrator] Remember the definition of agent AI?
2
00:00:03,001 --> 00:00:05,000
It is automation of AI
3
00:00:05,000 --> 00:00:08,006
using the reasoning power of large language models.
4
00:00:08,006 --> 00:00:14,001
What does reasoning in LLMs mean? Let's find out.
5
00:00:14,001 --> 00:00:16,003
LLMs are summarizing texts
6
00:00:16,003 --> 00:00:18,008
and writing marketing copy for us.
7
00:00:18,008 --> 00:00:22,003
They chat with us for customer support or legion.
8
00:00:22,003 --> 00:00:24,005
Are these LLMs reasoning?
9
00:00:24,005 --> 00:00:27,009
No, they're not reasoning in these examples.
10
00:00:27,009 --> 00:00:29,009
What is reasoning?
11
00:00:29,009 --> 00:00:34,005
Write your thoughts on what is reasoning on a notepad.
12
00:00:34,005 --> 00:00:37,006
Reasoning is the ability of the LLM
13
00:00:37,006 --> 00:00:39,009
to make decisions like a human
14
00:00:39,009 --> 00:00:43,001
in the absence of the human.
15
00:00:43,001 --> 00:00:45,007
When we chat with an intelligent chatbot,
16
00:00:45,007 --> 00:00:49,006
is the chatbot learning live from our conversations?
17
00:00:49,006 --> 00:00:52,008
When you ask a chatbot about a product or country,
18
00:00:52,008 --> 00:00:56,007
it might give incorrect answers sometimes.
19
00:00:56,007 --> 00:00:59,006
When you chat back saying that is not right,
20
00:00:59,006 --> 00:01:04,001
it will apologize and try to get you another answer,
21
00:01:04,001 --> 00:01:08,007
but the chatbot does not learn from the past mistakes.
22
00:01:08,007 --> 00:01:10,005
Do you know why an intelligent chatbot
23
00:01:10,005 --> 00:01:12,003
cannot learn from its mistakes
24
00:01:12,003 --> 00:01:14,009
even if it apologizes in the chat?
25
00:01:14,009 --> 00:01:17,006
Write your thoughts in your notepad.
26
00:01:17,006 --> 00:01:20,005
The chatbot is not learning from its mistakes
27
00:01:20,005 --> 00:01:24,003
because it is built upon an LLM model that is not changing.
28
00:01:24,003 --> 00:01:27,008
When you chat with the chatbot, the model is fixed
29
00:01:27,008 --> 00:01:31,007
and the chatbot is an application built on top of it.
30
00:01:31,007 --> 00:01:35,004
One of the abilities of making a LLM do reasoning
31
00:01:35,004 --> 00:01:38,005
is by building agent AI
32
00:01:38,005 --> 00:01:41,002
and giving it the ability to reason
33
00:01:41,002 --> 00:01:43,003
and correct its mistakes.
34
00:01:43,003 --> 00:01:46,009
To do this, the agent AI has to access
35
00:01:46,009 --> 00:01:49,003
new data sources and services
36
00:01:49,003 --> 00:01:52,006
and debate or ponder its decisions
37
00:01:52,006 --> 00:01:55,000
to choose the correct course of action.
38
00:01:55,000 --> 00:01:58,003
That is the ability of reasoning.
39
00:01:58,003 --> 00:02:03,000
Next, we'll learn how LLM reasoning can be achieved.
3063
Can't find what you're looking for?
Get subtitles in any language from opensubtitles.com, and translate them here.
Afrikaans
Akan
Albanian
Amharic
Arabic
Armenian
Azerbaijani
Basque
Belarusian
Bemba
Bengali
Bihari
Bosnian
Breton
Bulgarian
Cambodian
Catalan
Cebuano
Cherokee
Chichewa
Chinese (Simplified)
Chinese (Traditional)
Corsican
Croatian
Czech
Danish
Dutch
English
Esperanto
Estonian
Faroese
Finnish
Galician
Georgian
German
Greek
Guarani
Haitian Creole
Hausa
Hawaiian
Hebrew
Hmong
Hungarian
Icelandic
Indonesian
Irish
Italian
Japanese
Kazakh
Kinyarwanda
Kirundi
Kongo
Korean
Krio (Sierra Leone)
Kurdish
Kyrgyz
Laothian
Latin
Latvian
Lithuanian
Luganda
Luo
Luxembourgish
Macedonian
Malagasy
Malay
Maltese
Maori
Mauritian Creole
Moldavian
Mongolian
Myanmar (Burmese)
Montenegrin
Nepali
Norwegian
Pashto
Persian
Polish
Portuguese (Portugal)
Quechua
Romanian
Romansh
Russian
Samoan
Scots Gaelic
Serbian
Sesotho
Setswana
Seychellois Creole
Shona
Sindhi
Sinhalese
Slovak
Slovenian
Somali
Spanish
Swedish
Tajik
Thai
Tigrinya
Tonga
Turkish
Turkmen
Ukrainian
Uzbek
Vietnamese
Welsh
Wolof