1
00:00:00,020 --> 00:00:03,500
Let's talk about the ethics of using Chatgpt.

2
00:00:03,530 --> 00:00:06,980
This goes for any use of Chatgpt.

3
00:00:07,340 --> 00:00:12,730
I'm not here to be the morality police of what you should or shouldn't do.

4
00:00:12,740 --> 00:00:17,690
I think everyone is going to have their own moral compass and they'll ultimately decide how they can

5
00:00:17,690 --> 00:00:18,740
use Chatgpt.

6
00:00:19,070 --> 00:00:25,160
They can see determine if what they're doing with chatgpt is ethical or not.

7
00:00:25,190 --> 00:00:29,240
I think for me, the line is crossed between what's ethical and not.

8
00:00:29,270 --> 00:00:38,510
When you are using Chatgpt to do something that you are supposed to be doing yourself and you're claiming

9
00:00:38,510 --> 00:00:40,220
that it is your own writing.

10
00:00:40,220 --> 00:00:46,010
So if you're a student in school and you're just using an essay written by Chatgpt, then I feel like

11
00:00:46,010 --> 00:00:54,470
that is unethical versus if I'm writing a blog article and I'm just trying to help and provide information,

12
00:00:54,470 --> 00:01:02,550
it's not for a class where the purpose is to learn how to write or to learn about a subject and to prove

13
00:01:02,550 --> 00:01:09,870
that you have learned how to write about a subject, then providing great information that AI technology

14
00:01:09,870 --> 00:01:14,460
can do better or just as good or with assistance of chatgpt.

15
00:01:15,060 --> 00:01:17,250
I am comfortable doing that.

16
00:01:17,280 --> 00:01:20,190
Am I going to be doing that for all of my web content?

17
00:01:20,220 --> 00:01:21,180
No, not at all.

18
00:01:21,180 --> 00:01:27,090
I don't think that that's not the game that I play as a business myself, but I think that's just an

19
00:01:27,090 --> 00:01:33,300
example of what how the line of what's ethical or not can can be drawn.

20
00:01:33,300 --> 00:01:34,230
Just one example.

21
00:01:34,230 --> 00:01:39,900
So at the end of the day, you're going to have to decide what you think is ethical or not.

22
00:01:40,620 --> 00:01:50,430
Something that I tied with ethics is the this idea of like, how good is AI ultimately going to get

23
00:01:50,430 --> 00:01:57,450
in the future and what are people going to be doing to prevent AI based text.

24
00:01:57,480 --> 00:02:02,040
Is there going to be a push to ban AI based text or not?

25
00:02:02,040 --> 00:02:06,480
And the truth is that this AI technology is going to continue to grow.

26
00:02:06,510 --> 00:02:08,400
It's going to continue to improve.

27
00:02:08,400 --> 00:02:16,950
And there are chatgpt checkers now that potentially someday will be not be able to to work.

28
00:02:16,950 --> 00:02:18,480
I want to show you an example of that.

29
00:02:18,480 --> 00:02:23,700
So in the student essay writing lesson, I brought up this list of writing prompts.

30
00:02:23,700 --> 00:02:25,470
I'm going to show you another example.

31
00:02:25,470 --> 00:02:26,670
So here's an example.

32
00:02:26,700 --> 00:02:29,550
Do machines represent a threat to humans?

33
00:02:30,060 --> 00:02:39,270
I'm going to ask Chatgpt Write a 500 word essay or just say article on this prompt.

34
00:02:40,870 --> 00:02:43,990
And so Chatgpt is going to work this out.

35
00:02:44,020 --> 00:02:48,550
Now, The New York Times also links to an article about this topic.

36
00:02:48,550 --> 00:02:53,410
So we have some text that's already written and we can copy and paste this.

37
00:02:53,440 --> 00:03:01,990
What I'm going to do is copy and paste this into GPT 0.me, which is a AI text checker, and if I paste

38
00:03:01,990 --> 00:03:08,020
this text and I choose get results, it gives us back the results that the text is likely to be written

39
00:03:08,020 --> 00:03:09,640
by entirely by a human.

40
00:03:09,670 --> 00:03:12,880
That's good because it should be.

41
00:03:12,880 --> 00:03:21,310
This is a New York Times article that is written by what we assume is a person.

42
00:03:21,400 --> 00:03:22,150
Okay.

43
00:03:22,210 --> 00:03:31,450
And I think that's also an ethical line when if you have a byline that you are written by Michael Gonchar,

44
00:03:31,930 --> 00:03:41,470
but you are using Chatgpt, that's another issue, I think on a technical website where a lot of times

45
00:03:41,470 --> 00:03:48,220
companies are outsourcing text, you're using ghostwriters, you're not necessarily having a byline,

46
00:03:48,940 --> 00:03:52,210
then that's maybe not as bad.

47
00:03:52,660 --> 00:03:59,290
Again, it's up to your ethical and I'm interested to know what you think is ethical or not.

48
00:03:59,290 --> 00:04:03,910
Some people might just think anything written with Chatgpt is unethical.

49
00:04:03,910 --> 00:04:07,570
Some might think that anything goes.

50
00:04:07,570 --> 00:04:08,170
But here.

51
00:04:08,170 --> 00:04:08,740
Let's go.

52
00:04:08,740 --> 00:04:12,520
Now we have this essay written by Chatgpt.

53
00:04:12,820 --> 00:04:14,170
I'm going to paste it.

54
00:04:14,760 --> 00:04:19,770
Just to show you that I had to sign up for an account to continue using this.

55
00:04:19,770 --> 00:04:24,240
But let's go ahead and check this text.

56
00:04:24,810 --> 00:04:29,940
And it comes back that it's likely entirely written by AI.

57
00:04:29,970 --> 00:04:36,300
And you can see that the perplexity score, the burstiness score, these are the measurements that it

58
00:04:36,300 --> 00:04:41,250
uses to determine whether text is written by a AI or not.

59
00:04:41,250 --> 00:04:48,060
And of course you can go in there and then you could add your own ideas, your own thoughts to try to

60
00:04:48,060 --> 00:04:53,640
improve it and make it less likely to be marked as written by AI.

61
00:04:53,790 --> 00:04:56,640
But again, this goes back to the ethics of it.

62
00:04:56,640 --> 00:04:58,410
Is that what you want to do?

63
00:04:58,440 --> 00:05:00,930
Are you trying to get away with something?

64
00:05:00,930 --> 00:05:05,040
And that's probably ultimately the compass that you need to listen to.

65
00:05:05,040 --> 00:05:10,200
Are you trying to get away with something using chatgpt or are you using it as a tool?

66
00:05:10,230 --> 00:05:16,170
I've found that in all of the things that we've taught you in this class, a lot of it is just using

67
00:05:16,170 --> 00:05:23,610
it as an assistant in using this tool as a way to not only speed up the process we're doing things,

68
00:05:23,610 --> 00:05:31,750
but also to do it in a better way, like responding to student questions, assisting in creating social

69
00:05:31,750 --> 00:05:39,670
media posts, rewriting an article into an email, those kinds of things that just speed up our work.

70
00:05:39,670 --> 00:05:45,430
And at the end of the day, though, it's something to have a conversation about.

71
00:05:45,430 --> 00:05:52,540
It's something to think about and it's something we'll continue to to be thinking about as the AI gets

72
00:05:52,540 --> 00:05:53,860
better and better.

73
00:05:54,100 --> 00:05:54,460
All right.

74
00:05:54,460 --> 00:05:56,650
If you have any thoughts about this, I'd love to hear it.

75
00:05:56,650 --> 00:06:00,010
Let us know in the comments of the course or in the Q&A.

76
00:06:00,040 --> 00:06:01,810
We'd love to hear your thoughts.

77
00:06:01,810 --> 00:06:03,940
About what?

78
00:06:04,440 --> 00:06:06,600
Is ethical and what is not.

79
00:06:06,720 --> 00:06:07,560
Cheers.