1
00:00:07,173 --> 00:00:10,453
Speaker 1: You're listening to the Saturday Morning with Jack team podcast

2
00:00:10,613 --> 00:00:11,773
Speaker 1: from News Talks.

3
00:00:11,573 --> 00:00:15,053
Speaker 2: D twenty two to eleven on News Talks, said be

4
00:00:15,373 --> 00:00:19,133
Speaker 2: Paul Stenhouse our text. Bert has undertaken a bit of

5
00:00:19,173 --> 00:00:21,733
Speaker 2: an experiment this week. He wanted to see if AI

6
00:00:22,013 --> 00:00:24,533
Speaker 2: could build an app. He's put it to the test

7
00:00:24,573 --> 00:00:24,773
Speaker 2: for us.

8
00:00:24,853 --> 00:00:27,893
Speaker 3: Goal to Paul, Good morning Jack. I don't know if

9
00:00:27,893 --> 00:00:30,773
Speaker 3: it was a good idea, and it's a bit a

10
00:00:30,813 --> 00:00:33,813
Speaker 3: frustrating experience, but I can say there is a working

11
00:00:33,893 --> 00:00:37,853
Speaker 3: app that I wrote zero lines of code for. That's

12
00:00:38,093 --> 00:00:38,853
Speaker 3: kind of crazy.

13
00:00:39,493 --> 00:00:41,693
Speaker 2: Yeah, that is amazing. Okay, so tell us about the

14
00:00:41,733 --> 00:00:44,853
Speaker 2: idea for the app or for your website, and then

15
00:00:45,133 --> 00:00:47,333
Speaker 2: explain to us how you managed to build it without

16
00:00:47,333 --> 00:00:48,573
Speaker 2: actually writing any code.

17
00:00:49,173 --> 00:00:49,413
Speaker 1: Yeah.

18
00:00:49,533 --> 00:00:53,453
Speaker 3: So I hate this thought. Every time there's an election call,

19
00:00:53,493 --> 00:00:56,173
Speaker 3: when each of his networks makes a call, I thought,

20
00:00:56,173 --> 00:00:58,293
Speaker 3: wouldn't it be good if we could track that so

21
00:00:58,333 --> 00:01:00,973
Speaker 3: that we could see which network thinks we're going to

22
00:01:01,013 --> 00:01:03,893
Speaker 3: have who as the president of the United States. The election,

23
00:01:03,973 --> 00:01:08,173
Speaker 3: of course, is on Tuesday. And I took to an

24
00:01:08,213 --> 00:01:12,133
Speaker 3: application that is called Cursor, and Cursor is like a

25
00:01:12,173 --> 00:01:16,853
Speaker 3: development tool crossed with AI, so you can actually ask

26
00:01:16,893 --> 00:01:19,333
Speaker 3: it things, and it will then create the code for it,

27
00:01:19,373 --> 00:01:22,493
Speaker 3: and you can commit that to your you know, dev environment,

28
00:01:23,493 --> 00:01:26,453
Speaker 3: your code just with the click of a button. It's incredible.

29
00:01:26,493 --> 00:01:30,093
Speaker 3: So you can say, give me a pigraph that shows

30
00:01:30,373 --> 00:01:33,293
Speaker 3: X and y it uses this is a variable. That

31
00:01:33,453 --> 00:01:37,213
Speaker 3: is a variable. Push submit. It will give you back

32
00:01:37,493 --> 00:01:40,613
Speaker 3: fully working code. And then what he basically did is

33
00:01:40,773 --> 00:01:43,693
Speaker 3: layered these on and on and on and on. So

34
00:01:44,253 --> 00:01:45,693
Speaker 3: you ask it to do one thing, then you ask

35
00:01:45,733 --> 00:01:47,173
Speaker 3: it to do another thing, then you ask it to

36
00:01:47,173 --> 00:01:49,013
Speaker 3: connect those two things, and then you ask it to

37
00:01:49,013 --> 00:01:53,573
Speaker 3: do something else. Yeah, it does get rather complicated. It

38
00:01:53,653 --> 00:01:56,733
Speaker 3: is not without headaches, it is not without frustrations. But

39
00:01:56,813 --> 00:02:00,333
Speaker 3: there is a working app app jecktame at all the

40
00:02:00,493 --> 00:02:03,973
Speaker 3: calls dot app, all the calls dot apps.

41
00:02:04,693 --> 00:02:07,973
Speaker 2: You can go to the Yeah, okay, you've already fortunately

42
00:02:08,013 --> 00:02:10,173
Speaker 2: sent me a little previous so I gave you joyed

43
00:02:10,173 --> 00:02:11,853
Speaker 2: looking at it this week. No, it's pretty amazing what

44
00:02:11,893 --> 00:02:13,813
Speaker 2: you can do given you haven't had to do any code,

45
00:02:13,853 --> 00:02:16,613
Speaker 2: although obviously I haven't been privy to the frustrations of

46
00:02:17,053 --> 00:02:19,373
Speaker 2: all of the layering. But what do you think this

47
00:02:19,533 --> 00:02:22,413
Speaker 2: means for the future of coding? Because for years now

48
00:02:22,413 --> 00:02:25,413
Speaker 2: we've been saying, Oh, kids don't need to worry about handwriting.

49
00:02:25,493 --> 00:02:27,693
Speaker 2: Kids don't even need to worry about multiplication. All they

50
00:02:27,733 --> 00:02:29,693
Speaker 2: need to know is how to code, and they'll have

51
00:02:29,733 --> 00:02:30,493
Speaker 2: a job for life.

52
00:02:31,533 --> 00:02:33,853
Speaker 3: Yes, well, I think that's a bit that's changing, right,

53
00:02:33,933 --> 00:02:37,493
Speaker 3: So I think you need to know technical concepts. People

54
00:02:37,533 --> 00:02:39,573
Speaker 3: still need to be technical. You still need to know

55
00:02:39,653 --> 00:02:41,773
Speaker 3: kind of what the tech is trying to do, and

56
00:02:41,853 --> 00:02:44,333
Speaker 3: what the concepts are, and what the individual services are

57
00:02:44,573 --> 00:02:47,093
Speaker 3: that make an application in a website run. What you

58
00:02:47,173 --> 00:02:48,733
Speaker 3: don't need to do, though, is actually know how to

59
00:02:48,733 --> 00:02:52,173
Speaker 3: do the actual code. The language part. This runs on typescript.

60
00:02:52,173 --> 00:02:54,413
Speaker 3: I don't know any typescript A I wrote it for me,

61
00:02:56,493 --> 00:02:58,813
Speaker 3: So I think that's really important. And I think that

62
00:02:59,373 --> 00:03:02,573
Speaker 3: just understanding the concepts, understanding how to talk to it.

63
00:03:02,773 --> 00:03:04,613
Speaker 3: One of the things I found this week Jack was

64
00:03:05,373 --> 00:03:07,853
Speaker 3: I really had to work out out what context it needed.

65
00:03:08,813 --> 00:03:11,413
Speaker 3: Words are really important with AI, right, you need to

66
00:03:11,453 --> 00:03:14,173
Speaker 3: make sure you're really explicit and the takeaway price down

67
00:03:14,253 --> 00:03:17,173
Speaker 3: does really have to be You have to be spot on, Yeah,

68
00:03:17,213 --> 00:03:20,853
Speaker 3: absolutely spot on. You know, using words like overlaid versus

69
00:03:20,893 --> 00:03:24,373
Speaker 3: above mean very different things, and they can send you

70
00:03:24,413 --> 00:03:26,773
Speaker 3: down a very bad path. Now overlaid and above it

71
00:03:26,853 --> 00:03:29,693
Speaker 3: kind of sound reasonably similar and have a pretty You know,

72
00:03:29,733 --> 00:03:31,293
Speaker 3: when you have a bit of context, they mean something,

73
00:03:31,333 --> 00:03:33,413
Speaker 3: but in coding terms they mean something very different, and

74
00:03:33,453 --> 00:03:35,933
Speaker 3: so that can lead you down some bad paths. But

75
00:03:35,973 --> 00:03:38,493
Speaker 3: I will say the takeaway I have is that it

76
00:03:38,573 --> 00:03:42,173
Speaker 3: really is about the power of communication. Funnily enough, the

77
00:03:42,213 --> 00:03:47,933
Speaker 3: ability and the skill to communicate explicitly, be clear, be concise,

78
00:03:48,373 --> 00:03:52,013
Speaker 3: and also doing some active listing, checking and clarifying what

79
00:03:52,213 --> 00:03:55,213
Speaker 3: the AI is telling you actually helps you create something.

80
00:03:55,973 --> 00:03:58,933
Speaker 2: Yeah, okay, let's see it. That's very interesting. Hey, I've

81
00:03:58,973 --> 00:04:00,773
Speaker 2: got to ask you, since you are in the US,

82
00:04:00,853 --> 00:04:03,493
Speaker 2: what is your pick for Wednesday this week?

83
00:04:04,933 --> 00:04:08,413
Speaker 3: Oh? You mean who I think it's going to be

84
00:04:08,453 --> 00:04:10,053
Speaker 3: and who I want it to be?

85
00:04:10,053 --> 00:04:15,293
Speaker 2: Because well, who do you think? Yeah? So you you

86
00:04:15,693 --> 00:04:19,493
Speaker 2: I'm guessing here that you are not a Donald Trump fan.

87
00:04:19,853 --> 00:04:22,973
Speaker 3: That would be true. Yes, So are we voting? I

88
00:04:23,013 --> 00:04:23,773
Speaker 3: think vote over here?

89
00:04:23,813 --> 00:04:23,973
Speaker 1: Now?

90
00:04:24,613 --> 00:04:26,573
Speaker 2: You can vote, yeah, I can vote.

91
00:04:26,653 --> 00:04:26,813
Speaker 1: Yeah.

92
00:04:26,893 --> 00:04:29,773
Speaker 3: Yeah, so I'll be casting my vote for Kamala Harris.

93
00:04:30,573 --> 00:04:34,373
Speaker 3: But watching with anticipation. The problem is is I think

94
00:04:34,373 --> 00:04:37,293
Speaker 3: I'm just going back to that last election where you know,

95
00:04:37,373 --> 00:04:41,333
Speaker 3: where it was him versus Clinton and just the polls

96
00:04:41,333 --> 00:04:42,813
Speaker 3: were just so wrong. And I guess I have no

97
00:04:43,013 --> 00:04:46,453
Speaker 3: faith in the polls. I faith that the election, but

98
00:04:46,773 --> 00:04:49,413
Speaker 3: the polling, I guess is what I'm nervous about. So

99
00:04:49,453 --> 00:04:51,253
Speaker 3: we'll see the middle of the country's doing.

100
00:04:51,933 --> 00:04:53,853
Speaker 2: I think. I think that's very wise. I think not

101
00:04:54,533 --> 00:04:56,493
Speaker 2: Donald Trump himself has always said he never puts any

102
00:04:56,493 --> 00:04:59,493
Speaker 2: faith in the polls. I just think, you know, because

103
00:04:59,533 --> 00:05:01,973
Speaker 2: all of the polsters correct on new information, right so

104
00:05:02,013 --> 00:05:04,413
Speaker 2: over the last couple of elections that have undercounted Donald

105
00:05:04,413 --> 00:05:07,093
Speaker 2: Trump support heading into the election, and they might do

106
00:05:07,133 --> 00:05:09,413
Speaker 2: the same thing this time around. They might also overcorrect.

107
00:05:09,533 --> 00:05:11,413
Speaker 2: We just don't know, and we won't know until the

108
00:05:11,413 --> 00:05:13,973
Speaker 2: results come through. Thank you so much, Paul, what a

109
00:05:14,013 --> 00:05:16,693
Speaker 2: fascinating experiment. I'm going to give you Paul's website again

110
00:05:16,733 --> 00:05:17,733
Speaker 2: so you can go and have a look or his

111
00:05:17,813 --> 00:05:20,213
Speaker 2: app once again. I hope that he's managed to build

112
00:05:20,253 --> 00:05:23,133
Speaker 2: in plenty of capacity for the millions of hits. Maybe

113
00:05:23,133 --> 00:05:25,173
Speaker 2: he's got a good advertising framework in there as well,

114
00:05:25,253 --> 00:05:27,893
Speaker 2: so we can make sure that Paul is sitting pretty.

115
00:05:27,893 --> 00:05:33,493
Speaker 2: It's all the calls dot app, so allthecalls dot app. Well,

116
00:05:33,493 --> 00:05:35,173
Speaker 2: make sure that we put a link on our website

117
00:05:35,173 --> 00:05:37,293
Speaker 2: as well, so you can see what Paul was able

118
00:05:37,333 --> 00:05:39,173
Speaker 2: to build using only AI.

119
00:05:40,253 --> 00:05:43,373
Speaker 1: For more from Saturday Morning with Jack Tame, listen live

120
00:05:43,453 --> 00:05:46,773
Speaker 1: to Newstalk ZB from nine am Saturday, or follow the

121
00:05:46,773 --> 00:05:48,253
Speaker 1: podcast on iHeartRadio