1
00:00:00,040 --> 00:00:02,840
Speaker 1: South African universities are our stepping up measures to curb

2
00:00:02,880 --> 00:00:07,120
Speaker 1: the unethical use of artificial intelligence by students in their

3
00:00:07,600 --> 00:00:12,200
Speaker 1: assignments and assessments. While AI tools like Chat, GPT and

4
00:00:12,280 --> 00:00:15,360
Speaker 1: Gemini can enhance learning, there is growing concern that some

5
00:00:15,440 --> 00:00:18,640
Speaker 1: students are using them in ways that misrepresent their own

6
00:00:18,680 --> 00:00:23,560
Speaker 1: work or bypass critical thinking. Universities are now balancing the

7
00:00:23,600 --> 00:00:27,440
Speaker 1: promise of AI with the need to maintain academic integrity.

8
00:00:27,720 --> 00:00:31,640
Speaker 1: Joining us to unpack these changes is Professor Jonathan Jansen,

9
00:00:31,680 --> 00:00:36,280
Speaker 1: a leading education expert. Professor Jansen will explain the strategies

10
00:00:36,320 --> 00:00:40,360
Speaker 1: being implemented and why they are necessary. Good afternoon, Thanks

11
00:00:40,400 --> 00:00:41,760
Speaker 1: very much for joining us.

12
00:00:42,200 --> 00:00:42,760
Speaker 2: You welcome.

13
00:00:43,960 --> 00:00:47,880
Speaker 1: I will start this conversation by asking what your opinion

14
00:00:48,040 --> 00:00:51,760
Speaker 1: is of the use of AI in tertiary education.

15
00:00:53,159 --> 00:00:55,640
Speaker 2: While I teach under the idea to third year class

16
00:00:55,640 --> 00:00:58,560
Speaker 2: and I welcome, I tell my students to bring it

17
00:00:58,600 --> 00:01:01,560
Speaker 2: along and to use it wisely. I'll show them not

18
00:01:01,600 --> 00:01:05,959
Speaker 2: to use advisorly, and I tell them that one of

19
00:01:05,959 --> 00:01:11,920
Speaker 2: the biggest threats to our development designation is universities are

20
00:01:11,959 --> 00:01:14,240
Speaker 2: trying to clam down an AI. We tried that or

21
00:01:14,280 --> 00:01:20,399
Speaker 2: the typewriter because we thought years ago, this would take

22
00:01:20,440 --> 00:01:22,640
Speaker 2: away curse of writing ability. Then we try to go

23
00:01:22,720 --> 00:01:26,000
Speaker 2: the calculator because we thought kids won't use their heads

24
00:01:26,000 --> 00:01:29,880
Speaker 2: to calculate, and that problem now it's inevitable. AI too,

25
00:01:30,360 --> 00:01:34,559
Speaker 2: over time will become inevitable. One of the reasons why

26
00:01:34,600 --> 00:01:39,759
Speaker 2: we are correcting with AI is because it exposed assessment fraud.

27
00:01:40,319 --> 00:01:44,280
Speaker 2: It exposed the really poor quality of assessments by lecturers

28
00:01:44,280 --> 00:01:48,880
Speaker 2: in our universities. And if you ask a poor question,

29
00:01:49,720 --> 00:01:51,600
Speaker 2: then of course gis it any is AI to get

30
00:01:51,640 --> 00:01:54,920
Speaker 2: the answer quickly. But if you ask questions with context

31
00:01:55,160 --> 00:01:59,520
Speaker 2: and nuanced and complexity bolt in AI can't get around

32
00:01:59,560 --> 00:02:05,120
Speaker 2: that because AI is normed according to Western models of

33
00:02:05,640 --> 00:02:08,760
Speaker 2: assessment and so on. So no, I think that's so

34
00:02:09,120 --> 00:02:13,880
Speaker 2: fuss about how to ban you know, AI in the

35
00:02:13,960 --> 00:02:17,320
Speaker 2: assessment of my stats. Bring AI and I will give

36
00:02:17,360 --> 00:02:19,400
Speaker 2: you a task in which you can get the AI

37
00:02:19,520 --> 00:02:22,160
Speaker 2: answer and get the answers that GRIEVE generated in class.

38
00:02:22,480 --> 00:02:25,000
Speaker 2: So there's a whole lot of ways around this sort

39
00:02:25,160 --> 00:02:28,480
Speaker 2: of wanting to clamp down on AI, because I think

40
00:02:28,480 --> 00:02:32,040
Speaker 2: it's an amazing resource if you used correctly.

41
00:02:33,000 --> 00:02:37,160
Speaker 1: I appreciate what you're saying about passing on to students

42
00:02:37,639 --> 00:02:42,480
Speaker 1: the message that they should use AI ethically. But the

43
00:02:42,520 --> 00:02:44,480
Speaker 1: fact of the matter is there are always people that

44
00:02:44,520 --> 00:02:48,320
Speaker 1: will try to cheat the system. There are hopefully fewer

45
00:02:48,320 --> 00:02:51,640
Speaker 1: than more, but there are people that will always find

46
00:02:51,720 --> 00:02:54,320
Speaker 1: a way to cheat. Now, how do you crack down

47
00:02:54,440 --> 00:02:57,120
Speaker 1: on those people using AI path.

48
00:02:57,720 --> 00:03:01,000
Speaker 2: In high education? Quite simply, you teach well, you teach deep,

49
00:03:01,160 --> 00:03:06,399
Speaker 2: you teach wide, and you ask questions that are intellectually challenging.

50
00:03:06,520 --> 00:03:10,080
Speaker 2: AI can't help you with, as I said, contexts and nuance.

51
00:03:10,480 --> 00:03:13,080
Speaker 2: But if you ask kids name the three parts of belief,

52
00:03:13,360 --> 00:03:17,320
Speaker 2: then of course the use AI get quick answers. So

53
00:03:17,520 --> 00:03:20,120
Speaker 2: I don't see A as a threat at all. As

54
00:03:20,120 --> 00:03:22,720
Speaker 2: I said, I use. I welcome it into my classroom

55
00:03:23,160 --> 00:03:26,360
Speaker 2: at Selmash University, and the students still get er if

56
00:03:26,400 --> 00:03:30,120
Speaker 2: they try to use AI inappropriately, because AI can't help

57
00:03:30,160 --> 00:03:34,520
Speaker 2: you if you miss my class, because the assessment is

58
00:03:34,560 --> 00:03:37,760
Speaker 2: based on the interaction and not just on notes. So

59
00:03:37,880 --> 00:03:41,040
Speaker 2: if you follow my drift, AIE is not the threat.

60
00:03:41,680 --> 00:03:46,000
Speaker 2: The threat here is full quality teaching and low quality assessments.

61
00:03:46,480 --> 00:03:48,080
Speaker 2: And so AI is my friend.

62
00:03:48,520 --> 00:03:51,160
Speaker 1: So you are placing the responsibility on the lecturers and

63
00:03:51,200 --> 00:03:52,960
Speaker 1: the teachers and.

64
00:03:52,920 --> 00:03:56,240
Speaker 2: The whole university system, you know, but it is, as

65
00:03:56,280 --> 00:03:59,040
Speaker 2: I said, in the history of new technologies, our first

66
00:03:59,080 --> 00:04:02,480
Speaker 2: reaction is always to ban, to clamp down, to destroy.

67
00:04:02,560 --> 00:04:04,520
Speaker 2: AA is going to be here for the next one

68
00:04:04,560 --> 00:04:07,200
Speaker 2: hundred and two hundred years what and it will get

69
00:04:07,280 --> 00:04:12,720
Speaker 2: smarter and smarter, you know, at circum venting whatever rules,

70
00:04:12,680 --> 00:04:16,479
Speaker 2: salut rules, senates of university set up. And so my

71
00:04:16,720 --> 00:04:19,279
Speaker 2: argument is, let's try and get better at teaching and

72
00:04:19,360 --> 00:04:21,840
Speaker 2: even better at assessment and the problem goes away.

73
00:04:22,279 --> 00:04:24,880
Speaker 1: I want to share this message with you from a listener, prof.

74
00:04:25,040 --> 00:04:28,280
Speaker 1: Because it speaks to what you're saying. They say, great

75
00:04:28,279 --> 00:04:31,359
Speaker 1: topic about AI in tertiary education. AI has changed me

76
00:04:31,400 --> 00:04:34,440
Speaker 1: from a seventy percent to a ninety percent student, not

77
00:04:34,560 --> 00:04:37,279
Speaker 1: because it's stopped my critical thinking, but because it's highlighted

78
00:04:37,279 --> 00:04:41,000
Speaker 1: that sometimes poer teaching abilities of lecturers. I've often sat

79
00:04:41,040 --> 00:04:44,440
Speaker 1: with thick textbooks that are convoluted and technical, and lecturers

80
00:04:44,440 --> 00:04:47,279
Speaker 1: don't help much. AI has helped simplify subjects so that

81
00:04:47,360 --> 00:04:48,160
Speaker 1: I understand.

82
00:04:48,680 --> 00:04:51,080
Speaker 2: And that's exactly what I do as a researcher as well.

83
00:04:51,120 --> 00:04:53,440
Speaker 2: You know, I would get into answer. Just give me

84
00:04:53,480 --> 00:04:55,600
Speaker 2: a rough sense of what let's teacher is out there

85
00:04:56,040 --> 00:04:58,840
Speaker 2: it's my job to mind deeper. It is my job

86
00:04:58,960 --> 00:05:01,680
Speaker 2: to look more broad. But it's a wonderful way to

87
00:05:01,800 --> 00:05:05,520
Speaker 2: quickly get at least a sample of thinking of writing

88
00:05:06,240 --> 00:05:09,400
Speaker 2: that is readily available. If that's where it stops, of course,

89
00:05:09,920 --> 00:05:12,720
Speaker 2: you know, then that will be a problem. But I

90
00:05:13,480 --> 00:05:17,240
Speaker 2: do workshops around the country with lecturers who are interested

91
00:05:17,320 --> 00:05:19,800
Speaker 2: in how they can improve their teaching so that it

92
00:05:20,000 --> 00:05:26,160
Speaker 2: is deep and engaging and complex and compelling, as opposed

93
00:05:26,200 --> 00:05:31,320
Speaker 2: to simply handing out notes and that I am now.

94
00:05:33,600 --> 00:05:36,680
Speaker 1: I suspect that perhaps not all universities and not all

95
00:05:36,800 --> 00:05:40,560
Speaker 1: lecturers are as progressive in terms of their thinking around

96
00:05:40,600 --> 00:05:43,440
Speaker 1: AI as you are, and that can lead to problems.

97
00:05:43,480 --> 00:05:45,640
Speaker 1: And I want to ask you this question because the

98
00:05:45,720 --> 00:05:48,120
Speaker 1: listener was so quick to message at the very very

99
00:05:48,120 --> 00:05:49,960
Speaker 1: beginning of the show. As soon as I'd announced that

100
00:05:50,000 --> 00:05:52,720
Speaker 1: we were talking to you, they said, I understand profiance

101
00:05:52,760 --> 00:05:55,440
Speaker 1: and will be on your program speaking on AI. I

102
00:05:55,600 --> 00:05:59,240
Speaker 1: was accused, amongst other students, of cheating last year by

103
00:05:59,320 --> 00:06:03,839
Speaker 1: using AI and not using their plagiarism detection tool, neither

104
00:06:03,920 --> 00:06:06,640
Speaker 1: of which are true. I've contacted both the legal Department

105
00:06:06,720 --> 00:06:09,680
Speaker 1: and the ombudsman or with no success or even a

106
00:06:09,680 --> 00:06:12,719
Speaker 1: reply to my inquiries. Then the person goes on to

107
00:06:12,800 --> 00:06:16,320
Speaker 1: name the university which I won't mention, and says, I

108
00:06:16,320 --> 00:06:18,159
Speaker 1: have all the documents in front of me and would

109
00:06:18,160 --> 00:06:19,880
Speaker 1: like to hear Profiance's opinion on this.

110
00:06:21,240 --> 00:06:23,920
Speaker 2: And yeah, I feel sorry for that student because one

111
00:06:23,960 --> 00:06:29,200
Speaker 2: of there's clear evidence of these different ways of catching

112
00:06:29,200 --> 00:06:33,599
Speaker 2: out students with counter technology. Is one of the big

113
00:06:33,640 --> 00:06:37,839
Speaker 2: problems is false positives. In other words, you're added to

114
00:06:37,839 --> 00:06:41,120
Speaker 2: do the diservice. I owe the student cheated, but actually

115
00:06:41,440 --> 00:06:48,320
Speaker 2: the detective technology is actually not good enough to distinguish

116
00:06:48,320 --> 00:06:53,239
Speaker 2: a genuine case of assessment fraud versus students who actually

117
00:06:53,279 --> 00:06:55,880
Speaker 2: did the hard work. But why are we swimming in

118
00:06:55,920 --> 00:06:59,680
Speaker 2: that shallow pool? Is my response? Right, So, trying to

119
00:06:59,720 --> 00:07:02,640
Speaker 2: sort of put out the fire, you know, like the

120
00:07:02,720 --> 00:07:09,360
Speaker 2: fire intellectually, you know, for students such that you don't

121
00:07:09,400 --> 00:07:12,080
Speaker 2: spend huge amounts of money as a university. And by

122
00:07:12,120 --> 00:07:18,800
Speaker 2: the way, every university has something called the Deputy avised

123
00:07:18,840 --> 00:07:22,760
Speaker 2: Chancellor for teaching and learning. The job of those people

124
00:07:23,080 --> 00:07:26,280
Speaker 2: they overpaid. The job of those people is to do

125
00:07:26,320 --> 00:07:29,320
Speaker 2: the training workshops, is to bring in I can give

126
00:07:29,360 --> 00:07:32,160
Speaker 2: you names of people who would do this for free,

127
00:07:32,480 --> 00:07:37,120
Speaker 2: to bring in experts in alternative assessments, experts in the

128
00:07:37,240 --> 00:07:40,440
Speaker 2: uses of AI in such a way that you boost learning.

129
00:07:40,800 --> 00:07:43,640
Speaker 2: This can be done right. So a lecturally doesn't have

130
00:07:43,680 --> 00:07:46,280
Speaker 2: to be like me, but a deputy a vised chancellor

131
00:07:46,320 --> 00:07:49,360
Speaker 2: has to bring in those kinds of resources, that kind

132
00:07:49,400 --> 00:07:54,480
Speaker 2: of expertise, instead of the first impulsive act being how

133
00:07:54,480 --> 00:07:55,280
Speaker 2: do we stop this?

134
00:07:55,520 --> 00:07:57,680
Speaker 1: Thank you very much for your time this afternoon. Professor

135
00:07:57,760 --> 00:08:01,440
Speaker 1: Jonathan Johansen, Professor and Education at Stellenbosch University,