1
00:00:00,000 --> 00:00:12,160
I see my veins. I've been driving this train. He is in this lane. There's no stop in this

2
00:00:12,160 --> 00:00:17,400
frame because I came to the game and I changed it to play. I was.

3
00:00:17,400 --> 00:00:25,640
All right. How's everybody doing? This is Dr. Infant. For those of you watching bearded

4
00:00:25,640 --> 00:00:36,360
doctor and find out. I've got my winter beard going right now during break and I'm excited

5
00:00:36,360 --> 00:00:45,920
to share this video and to really go over what we're talking about. And this is because

6
00:00:45,920 --> 00:00:52,080
you know some of this technology, you know, I've been using this technology that chat

7
00:00:52,080 --> 00:01:04,040
GPT, Claude, pie, you name it. I'll just say you name it. So I use and particularly I use

8
00:01:04,040 --> 00:01:14,000
chat GPT every day. And many times I may forget that you're always learning. So there's things

9
00:01:14,000 --> 00:01:21,280
I know. There's things that other people know about the platforms because they're always

10
00:01:21,280 --> 00:01:28,200
changing. And that's why I'm always asking and connecting with people and meeting with

11
00:01:28,200 --> 00:01:39,880
them and talking to them about what's going on. I have one person I talked about co-pilot

12
00:01:39,880 --> 00:01:47,280
and it's very interesting. So what happened yesterday was that I was in a meeting and

13
00:01:47,280 --> 00:01:57,360
I was sharing how I use chat GPT's vision. So if you didn't know chat GPT can see. So

14
00:01:57,360 --> 00:02:07,760
what do I mean? Can it see? Yes, it has eyes and it from my assessment, it's better in

15
00:02:07,760 --> 00:02:15,360
reading text than images. And I'm going to show you that in a minute. However, this is

16
00:02:15,360 --> 00:02:26,220
very remarkable. And I think it's a very interesting feature. And it's only going to get better.

17
00:02:26,220 --> 00:02:32,760
And I think that's important because many times we focus on the downsides of technology

18
00:02:32,760 --> 00:02:42,760
and what it can't do, which is true. I like what I heard the CEO and one of the founders

19
00:02:42,760 --> 00:02:56,480
of Anthropic AI, he said, well, hey, look, if the AI can do maybe 60%, so it fails 40%,

20
00:02:56,480 --> 00:03:06,320
but it can do 60% right now in a year, it'll probably be at 80 or 90%. So go with that

21
00:03:06,320 --> 00:03:14,480
assumption. So looking at chat GPT and this vision function, I've used it with text and

22
00:03:14,480 --> 00:03:22,520
it was amazing. And what I mean by text, the example I gave was I took a picture of some

23
00:03:22,520 --> 00:03:30,680
writing that I did and then asked it to summarize it and to make comments for that writing and

24
00:03:30,680 --> 00:03:43,040
it did it within seconds. Now, what I'm doing now is having it actually look at images.

25
00:03:43,040 --> 00:03:50,760
So it goes, I took pictures of my refrigerator and asked it based on what it sees, what can

26
00:03:50,760 --> 00:03:57,780
I make to eat for lunch and dinner? And I put the constraint that it had to be 300 calories

27
00:03:57,780 --> 00:04:04,360
and high protein. And it gave me some answers, which was very interesting. But before we

28
00:04:04,360 --> 00:04:12,560
do that, please make sure you click the notification button, you click like and subscribe so you

29
00:04:12,560 --> 00:04:19,520
get notified about any new content that we put out, as well as for those of you who are

30
00:04:19,520 --> 00:04:26,760
listening, make sure to leave a comment, as well as come over to YouTube. Come on over

31
00:04:26,760 --> 00:04:37,280
to YouTube. And for the YouTube family, we do have a podcast. So the podcast of this,

32
00:04:37,280 --> 00:04:45,440
so the audio version of this, it posted on YouTube, but also you can find it on Apple,

33
00:04:45,440 --> 00:04:54,840
Google, Spotify. So any platforms, about I think eight or nine platforms as well. So

34
00:04:54,840 --> 00:05:00,120
you can listen to the audio. So you can just look up my name and you'll find the podcast.

35
00:05:00,120 --> 00:05:12,680
So let's take a look at this. So I'm in chat GPT, chat GPT-4. Now don't judge my refrigerator.

36
00:05:12,680 --> 00:05:22,440
Do not judge the refrigerator. So I'm in chat GPT-4 plus the premium version. So I've got

37
00:05:22,440 --> 00:05:31,920
the paid version of chat GPT. So you can go in here. You see there's a carton of eggs.

38
00:05:31,920 --> 00:05:41,360
There's some other already cooked eggs, some vegetables, a little eggnog, a cup of ice.

39
00:05:41,360 --> 00:05:49,400
So I took the picture and this is all through the Android app on my phone. So you got to

40
00:05:49,400 --> 00:05:58,240
use your phone for this. If you want the fastest way to do it, otherwise you'll take a picture

41
00:05:58,240 --> 00:06:05,280
on your laptop or upload it to your computer. So I took a picture using my phone and I asked

42
00:06:05,280 --> 00:06:12,680
chat GPT, what can I make from lunch and dinner from my fridge? Make each meal 300 calories

43
00:06:12,680 --> 00:06:19,600
and high in protein. That's a big bar and those are some big constraints. Now the first

44
00:06:19,600 --> 00:06:25,960
thing in this first iteration instantly it said, okay, you can create an egg white omelet

45
00:06:25,960 --> 00:06:33,640
with spinach. Use about three egg whites, handful of spinach, heavy cream. So this I

46
00:06:33,640 --> 00:06:40,760
don't know. I mean, where's the spinach? They didn't have the spinach. And then it talked

47
00:06:40,760 --> 00:06:50,560
about dinner, a grilled chicken strips, a chicken salad. Don't have the chicken strips

48
00:06:50,560 --> 00:07:00,360
or the chicken salad in there. So it tells us to adjust the portions to meet the 300

49
00:07:00,360 --> 00:07:07,000
calorie goal high protein. So this may be because the constraints are too hard and it

50
00:07:07,000 --> 00:07:12,080
wanted to give me an answer. So I asked it, where's the spinach and chicken strips? My

51
00:07:12,080 --> 00:07:17,080
apologies for the oversight. I cannot visually identify specific items such as spinach or

52
00:07:17,080 --> 00:07:23,020
chicken strips in your fridge. However, based on what I can see, you could make the following.

53
00:07:23,020 --> 00:07:28,680
So now it let me know it didn't see it, but it can make the following based on what it

54
00:07:28,680 --> 00:07:34,640
can see. Egg salad using the eggs. You can make them hard boil. Okay, we've got hard

55
00:07:34,640 --> 00:07:43,800
boil eggs. Mustard, any herbs or spices available in the refrigerator. Okay, so this is an assumption.

56
00:07:43,800 --> 00:07:51,040
Now tuna salad, if you have canned tuna in your pantry, so it's going outside of the

57
00:07:51,040 --> 00:07:59,040
refrigerator and then it says mayo on the fridge door. The fridge door, unless it's

58
00:07:59,040 --> 00:08:08,560
got some superpowers, it's not in the picture. However, okay, we'll give it to it. Or maybe

59
00:08:08,560 --> 00:08:15,440
it's making a hunch, right? A good guess. And then any greens and sandwich. So this

60
00:08:15,440 --> 00:08:22,760
is bread if you have it on hand. So now it took another picture of the refrigerator and

61
00:08:22,760 --> 00:08:29,000
you can get some of the stuff that's down there. Now it says based on the contents of

62
00:08:29,000 --> 00:08:35,840
your fridge, it says use the boiled eggs. So now I know it can see the boiled eggs,

63
00:08:35,840 --> 00:08:44,340
which is amazing. This is great. Then it says serve this with a few of the sugar snap peas

64
00:08:44,340 --> 00:08:51,840
on the side for crunch and fiber. There are sugar snap peas, those right in that bag.

65
00:08:51,840 --> 00:09:04,320
So that is either there or let's see. There are sugar snap peas. No, the sugar snap peas

66
00:09:04,320 --> 00:09:13,280
I think are right here in the middle. Those are pickles in those bags. So it did find

67
00:09:13,280 --> 00:09:20,600
the something in there. Now let's look at this. Here's this grilled chicken. Where on

68
00:09:20,600 --> 00:09:28,040
the middle shelf the grilled chicken. I'll see the grilled chicken in here. So there's

69
00:09:28,040 --> 00:09:36,160
no grilled chicken in here. So and it's saying to shred it, mix it with a little barbecue

70
00:09:36,160 --> 00:09:45,280
sauce with the snap peas. And then it says please pour some corn on the stay within your

71
00:09:45,280 --> 00:09:52,800
calorie limit. The egg whites are about 17 calories each. The chicken calories, okay,

72
00:09:52,800 --> 00:09:59,480
depend on the weight. There is no chicken. So I say given the items visible here. So

73
00:09:59,480 --> 00:10:05,280
I saying given the items visible. So it's kind of revealing that it knows there's no

74
00:10:05,280 --> 00:10:13,400
chicken. Now it says canned tuna. Where does it see the canned tuna? Maybe it thinks that's

75
00:10:13,400 --> 00:10:22,720
canned tuna. So it's saying to create high protein tuna salad with using the canned

76
00:10:22,720 --> 00:10:28,160
tuna. That appears to be on the middle. So it might be there. And then it's assuming

77
00:10:28,160 --> 00:10:34,680
we have mayonnaise and then it's throwing in leafy greens and sugar snap peas. And then

78
00:10:34,680 --> 00:10:47,280
it talks about tofu on the middle shelf in the sugar snap peas. Now there's no tofu.

79
00:10:47,280 --> 00:10:56,120
There are no tofu. So there's a lot of eggs, but there's no tofu there. So now I go back

80
00:10:56,120 --> 00:11:02,480
in with another picture and I just zero shot it and just give it a picture. And then it

81
00:11:02,480 --> 00:11:10,600
says the boiled eggs with the sugar snap peas, deli turkey meat. I think there's turkey meat

82
00:11:10,600 --> 00:11:21,760
under these egg bites somewhere in there. Pasta in the blue. That is not pasta. Those

83
00:11:21,760 --> 00:11:32,880
are vegetables. So it's assuming a lot. It's saying that in those blue containers, that's

84
00:11:32,880 --> 00:11:41,540
pasta. It is not pasta. It's Brussels sprouts in there. And then it talks about sugar snap

85
00:11:41,540 --> 00:12:11,440
peas. So let's ask it. Let's see what it does. So now I'm throwing this in here to

86
00:12:11,440 --> 00:12:27,200
see what it can make without the calorie constraints from what it can see. Okay. We cook pasta.

87
00:12:27,200 --> 00:12:34,360
So it's still not able to see the... So this is good. So I took away the calorie constraints.

88
00:12:34,360 --> 00:12:43,720
It sees the eggs. So it says make an egg sandwich. And this is based on what it can see. So there

89
00:12:43,720 --> 00:12:50,960
are whole eggs in there. If there are bread and wraps available. We don't have that slice

90
00:12:50,960 --> 00:12:59,080
of cheese. So it's adding some things. Then pasta dish. It's still thinking that there's

91
00:12:59,080 --> 00:13:09,760
pasta in these blue containers, which they're not. And so overall, I think what's good about

92
00:13:09,760 --> 00:13:23,640
this is that it gives us some ideas. And with the images from the refrigerator, it wasn't

93
00:13:23,640 --> 00:13:31,640
so good. Now with the text, it was amazing. I mean, it was amazing. And I think that may

94
00:13:31,640 --> 00:13:38,880
be the next video that I'll do with the text. However, if you have the chat GPT app and

95
00:13:38,880 --> 00:13:45,560
you're able to upload pictures, make sure try it. I'd love to hear what you think, your

96
00:13:45,560 --> 00:13:52,780
thoughts on this. Think of some ideas on how you can use this. And I mean, it can only

97
00:13:52,780 --> 00:13:58,960
get better as we move forward. You take a picture, you can ask questions on that picture.

98
00:13:58,960 --> 00:14:06,200
It has visions, which I think is amazing. And I could focus on the negative, what it

99
00:14:06,200 --> 00:14:11,440
couldn't do. But I'm going to lean towards the positive side and think about it what

100
00:14:11,440 --> 00:14:19,400
it can do. So I'd love to hear what you think. As always, have a Merry, Merry Christmas.

101
00:14:19,400 --> 00:14:31,200
Thank you so much.

