1
00:00:02,280 --> 00:00:02,840
Hi guys.

2
00:00:02,840 --> 00:00:03,760
Welcome to this session.

3
00:00:03,760 --> 00:00:10,240
So in this session now we will see how we can use these different versions of ChatGPT and use some practical

4
00:00:10,240 --> 00:00:15,560
prompts or same prompts on different versions to see the difference in the output which we get from

5
00:00:15,560 --> 00:00:15,920
them.

6
00:00:16,120 --> 00:00:17,360
So let's have a look at this.

7
00:00:17,640 --> 00:00:22,880
So the first one we are going to do a comparison is going to be between ChatGPT four or mini versus

8
00:00:22,880 --> 00:00:24,240
ChatGPT oh one.

9
00:00:24,400 --> 00:00:33,400
So let's start with that wherein we're asking ChatGPT four or mini to create a image of a dog running

10
00:00:33,400 --> 00:00:34,360
through the forest.

11
00:00:48,880 --> 00:00:50,280
It is able to do that absolutely.

12
00:00:50,280 --> 00:00:50,760
Fine.

13
00:00:50,960 --> 00:00:52,920
Now let's switch this over.

14
00:01:17,640 --> 00:01:20,280
And now we are going to give the same option again.

15
00:01:26,000 --> 00:01:30,280
Since looking at the reasoning part of it, as you can see, oh one will do a little bit of reasoning

16
00:01:30,280 --> 00:01:35,160
about the whole task and then it's going to give us information, as you can see.

17
00:01:39,200 --> 00:01:41,000
It's giving us all the information about it.

18
00:01:41,000 --> 00:01:47,400
It's also giving us some prompt which we can use over here, but specifically with oh one specifically,

19
00:01:47,400 --> 00:01:49,280
you can see this limitation that it has.

20
00:01:49,320 --> 00:01:54,720
It does not have that capability of converting text to image in this particular case.

21
00:01:54,920 --> 00:01:57,600
So this is one drawback which you will get with oh one.

22
00:01:58,880 --> 00:02:01,760
Same way you can do it with describing images as well.

23
00:02:02,200 --> 00:02:04,760
Now let's look at come back again to.

24
00:02:07,120 --> 00:02:09,560
Oh one mini Photo mini.

25
00:02:12,640 --> 00:02:15,480
And let's say we start with another prompt over here.

26
00:02:21,880 --> 00:02:27,880
So here we have talked about I work at an established tech consulting firm specializing in digital transformation.

27
00:02:27,880 --> 00:02:31,680
I lead our efforts in regional expansion and strategic partnerships.

28
00:02:31,680 --> 00:02:37,240
We are evaluating potential locations for a new office in Asia to strengthen our presence in emerging

29
00:02:37,240 --> 00:02:38,080
markets.

30
00:02:38,120 --> 00:02:42,480
Singapore is our choice, but we are also considering Tokyo and Hong Kong, where we already have some

31
00:02:42,480 --> 00:02:43,720
client relationships.

32
00:02:43,720 --> 00:02:46,360
Help me devise a comprehensive plan for this expansion.

33
00:02:46,520 --> 00:02:52,760
So now you can see it's giving us all the structure, market research, location analysis, operational

34
00:02:52,760 --> 00:02:59,080
setup, strategic partnerships and networking, cultural talent considerations, market and brand awareness,

35
00:02:59,280 --> 00:03:03,840
risk mitigation and contingency plans, timing and budget.

36
00:03:05,720 --> 00:03:11,280
So all that is given by ChatGPT for Omni, as you can see over here.

37
00:03:11,440 --> 00:03:15,270
But now if we switch to One.

38
00:03:22,190 --> 00:03:28,630
And give the same prompt is going to look at the reasoning of it and analyzing it.

39
00:03:28,670 --> 00:03:31,910
It also and then it is going to give us.

40
00:03:31,910 --> 00:03:36,430
So this is the expansion part where it is giving it few seconds to think through.

41
00:03:36,550 --> 00:03:38,710
And then it's going to give us all the information.

42
00:03:38,710 --> 00:03:45,030
Now if you see the the detailing is very precise with oh one you get a lot of pros and cons.

43
00:03:45,070 --> 00:03:48,510
This was not there in our previous prompt which we had looked at.

44
00:03:48,510 --> 00:03:49,070
Right?

45
00:03:49,110 --> 00:03:49,350
Okay.

46
00:03:49,390 --> 00:03:53,390
So if for every location it's giving you pros and cons okay.

47
00:03:53,590 --> 00:04:00,030
Then with every section it's telling us the legal and regularity, uh, regulatory things as well,

48
00:04:00,030 --> 00:04:01,750
financial and operational risks.

49
00:04:01,950 --> 00:04:02,190
Okay.

50
00:04:02,230 --> 00:04:04,110
So it's getting more detailed.

51
00:04:04,110 --> 00:04:11,710
So oh one is really good at doing these particular analysis bit where it will go into the specifics

52
00:04:11,710 --> 00:04:17,950
of it and give you very minute details, which possibly for many does not do that much.

53
00:04:18,470 --> 00:04:26,070
So this is the strength of one specifically, wherein it will give you far more accurate, precise and

54
00:04:26,470 --> 00:04:28,910
granular information versus the four.

55
00:04:30,430 --> 00:04:35,150
Now let's look at another example with four row many.

56
00:04:37,390 --> 00:04:40,430
For a coding example which we can look at.

57
00:04:43,870 --> 00:04:51,110
So we'll try with this is going to be oh one mini versus let's do oh one mini oh one specifically.

58
00:04:53,430 --> 00:05:01,030
And now with oh one specifically, we are looking at a asking ChatGPT to develop a complete code because

59
00:05:01,030 --> 00:05:02,430
we are looking to build a website.

60
00:05:02,430 --> 00:05:08,470
So I need to develop a new web application from scratch using Python with flask for the backend and

61
00:05:08,870 --> 00:05:10,710
Vue.js for the front end.

62
00:05:11,030 --> 00:05:16,550
The application will enable users to create accounts, post updates and search for other users by interest.

63
00:05:16,790 --> 00:05:22,070
Could you provide the complete project setup, including file structure, required package installations,

64
00:05:22,070 --> 00:05:27,550
and an initial implementation for user authentication and handling user generated content.

65
00:05:27,590 --> 00:05:33,350
So basically what I'm looking for, some kind of coding related, uh, pro results so that it can help

66
00:05:33,350 --> 00:05:34,670
me set up the website.

67
00:05:39,350 --> 00:05:40,790
So this is happening on oh one.

68
00:05:47,270 --> 00:05:49,870
You can see it's building and connecting okay.

69
00:05:49,910 --> 00:05:51,430
So crafting the blueprint.

70
00:06:01,230 --> 00:06:03,590
Is giving us the code over here now okay.

71
00:06:03,630 --> 00:06:08,430
It had to think through it a little bit 20s of thinking went into it.

72
00:06:09,550 --> 00:06:11,910
And then the project structure, the code came up.

73
00:06:13,070 --> 00:06:15,070
Explanation also been given for it.

74
00:06:15,990 --> 00:06:17,830
Back end setup is given.

75
00:06:19,990 --> 00:06:22,670
There will be a front end one as well which will come up.

76
00:06:23,590 --> 00:06:26,670
It's still writing the code for us for the back end.

77
00:06:29,670 --> 00:06:38,350
So it's a little slower in that respect in generating code related prompts responses, as you can see.

78
00:06:38,390 --> 00:06:39,990
Front end setup now.

79
00:06:42,070 --> 00:06:44,750
The code has been generated for front end setup as well.

80
00:06:51,950 --> 00:06:54,590
It is providing us the whole code for that also.

81
00:06:55,830 --> 00:07:00,630
This we are going to now compare with specifically O3 mini.

82
00:07:00,670 --> 00:07:00,990
Hi.

83
00:07:01,870 --> 00:07:02,310
Okay.

84
00:07:02,350 --> 00:07:08,750
So O3 is primarily built out for faster coding related responses which we can get over here.

85
00:07:09,110 --> 00:07:09,310
Okay.

86
00:07:09,350 --> 00:07:11,150
So so we'll see it in action.

87
00:07:11,310 --> 00:07:13,510
How fast or how frequent.

88
00:07:13,550 --> 00:07:15,670
How better it is than O1.

89
00:07:16,030 --> 00:07:18,710
So now you can see the O1 response has completed.

90
00:07:18,750 --> 00:07:21,470
Now we'll switch this to O3 mini high.

91
00:07:24,190 --> 00:07:26,150
And we are going to give the same prompt.

92
00:07:33,780 --> 00:07:35,620
And you can see how quick it is.

93
00:07:35,660 --> 00:07:39,020
It I think in less than five seconds it took for right down the whole thing.

94
00:07:54,740 --> 00:07:57,260
And now you can see the code has already started generating.

95
00:07:58,060 --> 00:08:00,260
You can see the the frequency.

96
00:08:00,260 --> 00:08:02,460
It is much faster over here.

97
00:08:02,500 --> 00:08:09,340
Now with respect to with respect to the previous one with 0103, mini high is far more faster in generating

98
00:08:09,340 --> 00:08:12,100
the codes which we are looking for out here.

99
00:08:12,420 --> 00:08:17,380
So you can literally now see the difference in the versions of ChatGPT which we have here.

100
00:08:17,500 --> 00:08:21,740
And that is what we were discussing in the previous videos as well, that each of the versions has their

101
00:08:21,740 --> 00:08:23,740
own strengths and limitations.

102
00:08:23,900 --> 00:08:29,820
So depending on what is your requirement, you need to select the right version for yourself and use

103
00:08:29,820 --> 00:08:30,180
that.

104
00:08:30,500 --> 00:08:31,500
I hope this makes sense.

105
00:08:31,500 --> 00:08:33,340
You understand the whole concept now.

106
00:08:33,380 --> 00:08:37,700
Thank you so much guys for listening to this and I will see you in the next video.