1
00:00:07,133 --> 00:00:10,453
Speaker 1: You're listening to the Saturday Morning with Jack team podcast

2
00:00:10,573 --> 00:00:13,413
Speaker 1: from News Talks at be You know how.

3
00:00:13,373 --> 00:00:15,573
Speaker 2: There's a smart everything these days. You can get like

4
00:00:15,613 --> 00:00:19,213
Speaker 2: a smart toothbrush. You can get a smart bed or

5
00:00:19,213 --> 00:00:22,533
Speaker 2: a smart mattress. So it's a mattress that tracks your

6
00:00:22,573 --> 00:00:24,973
Speaker 2: sleep habits and it can heat up and cool down

7
00:00:25,013 --> 00:00:27,733
Speaker 2: and do various bits and pieces. But users of these

8
00:00:27,973 --> 00:00:30,653
Speaker 2: eight that's the name of the brand, eight Sleep smart

9
00:00:30,693 --> 00:00:33,813
Speaker 2: mattresses had a bit of a rough week. Some of

10
00:00:33,853 --> 00:00:35,933
Speaker 2: them were woken up when their mattresses got really hot,

11
00:00:36,013 --> 00:00:38,733
Speaker 2: and other ones the mattresses started sitting up and changing

12
00:00:38,813 --> 00:00:40,693
Speaker 2: their positioning even though they didn't want them to. And

13
00:00:40,773 --> 00:00:45,173
Speaker 2: it all came down to a fault with Amazon Web Services.

14
00:00:45,373 --> 00:00:50,013
Speaker 2: So these web services that provide kind of infrastructure cloud

15
00:00:50,053 --> 00:00:55,053
Speaker 2: infrastructure for lots of different products and websites around the

16
00:00:55,053 --> 00:00:57,853
Speaker 2: world went down during the week, which had all sorts

17
00:00:57,893 --> 00:01:01,453
Speaker 2: of unintended consequences. Our texpert poor Steinhouse has all the

18
00:01:01,493 --> 00:01:03,133
Speaker 2: details for us. Paul, you don't have one of those

19
00:01:03,173 --> 00:01:05,333
Speaker 2: fancy mattresses, do you that turned out not to be

20
00:01:05,413 --> 00:01:07,013
Speaker 2: quite as smart as you might have hoped.

21
00:01:07,693 --> 00:01:10,853
Speaker 3: I don't, but I have to say, Jack the enthusiasm

22
00:01:10,893 --> 00:01:13,573
Speaker 3: in your voice talking about that was a little telling.

23
00:01:13,933 --> 00:01:16,573
Speaker 2: Well, I don't know. I mean some things like do

24
00:01:16,613 --> 00:01:18,693
Speaker 2: you really need to does everything need to be smart?

25
00:01:18,733 --> 00:01:22,333
Speaker 2: Does everything need to be connected? It's just it's funny, right, Like, look,

26
00:01:22,413 --> 00:01:23,093
Speaker 2: it's funny.

27
00:01:23,213 --> 00:01:26,093
Speaker 3: But there was a problem where people had smart homes

28
00:01:26,613 --> 00:01:30,413
Speaker 3: and with smart locks, smart keys, and they couldn't get

29
00:01:30,413 --> 00:01:34,493
Speaker 3: into their homes. So that was a problem. Yes, but no, Look,

30
00:01:34,653 --> 00:01:37,413
Speaker 3: here's the one thing that it says, right, I mean,

31
00:01:37,533 --> 00:01:39,213
Speaker 3: the thing you have to remember about the cloud is

32
00:01:39,253 --> 00:01:42,333
Speaker 3: it's not mythical. It's not magical. It's someone else's server

33
00:01:42,453 --> 00:01:46,973
Speaker 3: that they maintain. You're renting a server effectively, and it

34
00:01:47,013 --> 00:01:49,493
Speaker 3: really does show that, you know, a lot of people

35
00:01:49,573 --> 00:01:52,333
Speaker 3: are renting these servers. Fun fact for you, did you

36
00:01:52,373 --> 00:01:55,933
Speaker 3: know Amazon's stock while they were having fourteen hours of downtime,

37
00:01:56,213 --> 00:01:58,973
Speaker 3: the stock popped on that day. It was up like

38
00:01:59,053 --> 00:02:02,173
Speaker 3: three percent, and I yes, it didn't go down. And

39
00:02:02,213 --> 00:02:04,573
Speaker 3: I think it's because it really shows that there is

40
00:02:04,933 --> 00:02:06,253
Speaker 3: such a reliance on these server.

41
00:02:06,733 --> 00:02:06,853
Speaker 1: Right.

42
00:02:06,893 --> 00:02:08,933
Speaker 3: We've gone all in on the cloud. It's been the

43
00:02:09,013 --> 00:02:12,573
Speaker 3: talking point in corporate America certainly for the past ten years.

44
00:02:12,613 --> 00:02:14,613
Speaker 3: Get your cloud. It's cheaper don't have to maintain it

45
00:02:14,653 --> 00:02:16,053
Speaker 3: all this stuff. Well, we've moved to the cloud and

46
00:02:16,053 --> 00:02:20,293
Speaker 3: now we've got some failover issues. But what was really

47
00:02:20,373 --> 00:02:23,693
Speaker 3: interesting about this one was it's kind of like Amazon

48
00:02:24,013 --> 00:02:27,693
Speaker 3: uses Amazon's own services to run other Amazon services, and

49
00:02:27,733 --> 00:02:30,693
Speaker 3: that was what the big issue was this week. So

50
00:02:31,493 --> 00:02:34,693
Speaker 3: there's a database server, a service called Dynamo dB. I

51
00:02:34,733 --> 00:02:36,813
Speaker 3: won't bore you with the details of it, but what

52
00:02:36,893 --> 00:02:40,573
Speaker 3: effectively happened was the IP addresses of all of the

53
00:02:40,693 --> 00:02:44,773
Speaker 3: Dynamo dbs and the Virginia data center known as US

54
00:02:44,853 --> 00:02:49,693
Speaker 3: East one were lost. They were removed. So well, that

55
00:02:49,853 --> 00:02:52,893
Speaker 3: basically means is you've got a phone book, but all

56
00:02:52,933 --> 00:02:55,213
Speaker 3: of the phone numbers have been removed from the phone books.

57
00:02:55,293 --> 00:02:59,173
Speaker 3: You don't know anyone's phone numbers anymore. But the problem

58
00:02:59,293 --> 00:03:02,533
Speaker 3: is is that all the other Amazon services, they actually

59
00:03:02,573 --> 00:03:04,693
Speaker 3: need the numbers in the phone book to operate all

60
00:03:04,733 --> 00:03:09,813
Speaker 3: of their own services. So taking down this one database cluster,

61
00:03:10,013 --> 00:03:12,733
Speaker 3: which is a problem in and upon itself, right, like

62
00:03:13,053 --> 00:03:16,693
Speaker 3: losing entire access to a service that you sell as

63
00:03:16,773 --> 00:03:20,253
Speaker 3: bad one, but two all of these other services. I

64
00:03:20,253 --> 00:03:22,333
Speaker 3: think at one point there was something like twenty five

65
00:03:22,413 --> 00:03:26,133
Speaker 3: different Amazon services that were down because they all relied

66
00:03:26,293 --> 00:03:31,213
Speaker 3: on this one surface, yus, which is just incredible.

67
00:03:31,253 --> 00:03:34,853
Speaker 2: It's amazing when something like one domino basically is what

68
00:03:34,893 --> 00:03:37,573
Speaker 2: you're telling us, one dominant complex terms, you know, but

69
00:03:38,093 --> 00:03:41,693
Speaker 2: one domino falls and then nick minute, you've woken up

70
00:03:41,693 --> 00:03:43,093
Speaker 2: in the middle of the night because your smart bed's

71
00:03:43,093 --> 00:03:47,493
Speaker 2: trying to follow itself in half. Hey, chet gpt has

72
00:03:47,573 --> 00:03:50,333
Speaker 2: launched a browser. So open Ai, the company behind chet

73
00:03:50,373 --> 00:03:53,213
Speaker 2: gptv been pursuing a few different products and tools and

74
00:03:53,253 --> 00:03:56,213
Speaker 2: things lately, but now they've launched an internet browser. So

75
00:03:56,253 --> 00:03:59,093
Speaker 2: if you have Google Chrome or Safari or something like

76
00:03:59,093 --> 00:04:01,013
Speaker 2: that on your computer, this is designed to compete.

77
00:04:02,213 --> 00:04:04,933
Speaker 3: It is and I'm full of fun facts today. But

78
00:04:04,973 --> 00:04:09,853
Speaker 3: another fun fact, the open ai browser is effectively using

79
00:04:09,853 --> 00:04:14,173
Speaker 3: the same technology as Google Chrome. Google actually make it

80
00:04:14,213 --> 00:04:16,333
Speaker 3: open source and you can build browsers on top of it.

81
00:04:16,373 --> 00:04:18,973
Speaker 3: So it's a little bit kind of amusing the people

82
00:04:19,013 --> 00:04:21,733
Speaker 3: who want to go and take down Google and take

83
00:04:21,773 --> 00:04:24,613
Speaker 3: all of your attention away from Google searches is using

84
00:04:25,213 --> 00:04:29,133
Speaker 3: Google funded technology from Google Search and Google Ads to

85
00:04:29,373 --> 00:04:33,853
Speaker 3: then create their own things. Right that aside, Yes, So

86
00:04:33,973 --> 00:04:37,533
Speaker 3: what they're doing is they're baking chat gpt in into

87
00:04:37,613 --> 00:04:40,253
Speaker 3: the browser so when you first open it up, instead

88
00:04:40,293 --> 00:04:42,573
Speaker 3: of typically seeing what we've all seen for the past

89
00:04:42,933 --> 00:04:46,653
Speaker 3: twenty odd years, the beautiful Google search bar, you now

90
00:04:46,733 --> 00:04:50,613
Speaker 3: see a chat GPT box and you can start typing

91
00:04:50,733 --> 00:04:54,453
Speaker 3: into that instead of your Google Search, and you can

92
00:04:54,493 --> 00:04:56,893
Speaker 3: ask it questions. You can ask it to find information,

93
00:04:56,933 --> 00:04:58,533
Speaker 3: you can ask it to go to websites, all those

94
00:04:58,573 --> 00:05:01,053
Speaker 3: types of things. But it's also got a sort of

95
00:05:01,133 --> 00:05:03,813
Speaker 3: side panel, so when you're browsing to a website, you

96
00:05:03,813 --> 00:05:07,093
Speaker 3: can then say, like, chat about this week site with

97
00:05:07,333 --> 00:05:09,853
Speaker 3: chat GPT, so it takes all the context of that

98
00:05:10,013 --> 00:05:13,093
Speaker 3: page and it can even start to do actions. Now.

99
00:05:13,333 --> 00:05:15,333
Speaker 3: I tried it. I was hoping it was going to

100
00:05:15,373 --> 00:05:18,213
Speaker 3: be quite exciting, but it let me down twice. The

101
00:05:18,253 --> 00:05:20,013
Speaker 3: first time I was I was trying to get it

102
00:05:20,013 --> 00:05:22,093
Speaker 3: to give me I was like, okay, give me a

103
00:05:22,173 --> 00:05:24,533
Speaker 3: summary of the day's news from CNN or the New

104
00:05:24,613 --> 00:05:26,973
Speaker 3: York Times, right, and both of those websites were restricted

105
00:05:26,973 --> 00:05:28,253
Speaker 3: and it said I'm not allowed to do it. So

106
00:05:28,293 --> 00:05:31,973
Speaker 3: I was like, okay, well that was frustrating. Good good

107
00:05:31,973 --> 00:05:33,373
Speaker 3: for seeing in good for the New York Times that

108
00:05:33,373 --> 00:05:35,893
Speaker 3: I started to block chat GPT, But I was like,

109
00:05:35,933 --> 00:05:37,773
Speaker 3: that was annoying for me as a person trying to

110
00:05:37,853 --> 00:05:39,933
Speaker 3: use it anyway. Second of all, I thought, okay, let's

111
00:05:39,973 --> 00:05:41,693
Speaker 3: see if you can really test it. I'm going to

112
00:05:41,773 --> 00:05:44,093
Speaker 3: ask it to draft me an email and then get

113
00:05:44,133 --> 00:05:47,213
Speaker 3: it ready in my Gmail because it can do these actions, right.

114
00:05:47,253 --> 00:05:49,173
Speaker 3: It says it can take over and it can, you know,

115
00:05:49,173 --> 00:05:52,853
Speaker 3: effectively work the screen like a user could, and it

116
00:05:52,893 --> 00:05:56,133
Speaker 3: wrote the email just like it wouldn't chat GPT. But

117
00:05:56,253 --> 00:05:59,013
Speaker 3: then it asked me so many times for permissions to

118
00:05:59,053 --> 00:06:00,653
Speaker 3: do this and open it up here and do this.

119
00:06:00,973 --> 00:06:02,533
Speaker 3: It would have just been easier for me to copy

120
00:06:02,573 --> 00:06:06,053
Speaker 3: and paste it enter Gmail directly. So I will say

121
00:06:06,453 --> 00:06:10,253
Speaker 3: I was left a little bit disappointed under in the

122
00:06:10,293 --> 00:06:13,133
Speaker 3: browser underwell, and look, I have to say this too.

123
00:06:13,173 --> 00:06:15,813
Speaker 3: I mean, the US is a really big Apple market, right,

124
00:06:15,893 --> 00:06:18,813
Speaker 3: huge with the iPhone, sure, huge with the MacBook. And

125
00:06:18,853 --> 00:06:21,293
Speaker 3: so they've actually launched this only for Mac at the moment,

126
00:06:21,693 --> 00:06:23,613
Speaker 3: So if you have a Windows computer at the moment,

127
00:06:23,893 --> 00:06:25,933
Speaker 3: you're out of luck. You can't even try the check.

128
00:06:26,253 --> 00:06:28,893
Speaker 2: This sounds like it's not that not that impressive. See,

129
00:06:28,893 --> 00:06:31,973
Speaker 2: I've found it's interesting how it's only once you really

130
00:06:31,973 --> 00:06:33,693
Speaker 2: play around with these things and kind of get into

131
00:06:33,733 --> 00:06:37,413
Speaker 2: a habit that it goes from being you know, like

132
00:06:37,453 --> 00:06:39,773
Speaker 2: a bit of a quirk quirky sort of party trick

133
00:06:39,893 --> 00:06:41,293
Speaker 2: or something, a bit of a novelty.

134
00:06:41,853 --> 00:06:41,973
Speaker 3: You know.

135
00:06:42,013 --> 00:06:43,613
Speaker 2: It takes a bit of time to work out exactly

136
00:06:43,653 --> 00:06:45,613
Speaker 2: if you can use these and kind of integrate them

137
00:06:45,653 --> 00:06:49,653
Speaker 2: as part of your daily life. I use chat GPT

138
00:06:49,893 --> 00:06:53,653
Speaker 2: heaps as a kind of search tool, Like I'm I'm

139
00:06:53,693 --> 00:06:55,893
Speaker 2: asking it questions that once I would have used Google,

140
00:06:55,933 --> 00:06:58,333
Speaker 2: and it's you know, I can only imagine that Google's

141
00:06:58,533 --> 00:07:03,333
Speaker 2: traffic has been significantly impacted by chat GPT, but the

142
00:07:03,373 --> 00:07:06,733
Speaker 2: fact that the browser hasn't quite nailed it yet kind

143
00:07:06,733 --> 00:07:07,333
Speaker 2: of intriguing.

144
00:07:07,853 --> 00:07:09,613
Speaker 3: I mean, it can do that, so I guess I

145
00:07:09,653 --> 00:07:11,533
Speaker 3: was trying to really take it through its paces.

146
00:07:11,573 --> 00:07:13,533
Speaker 2: Yeah, of course. I mean that's like you've got to

147
00:07:13,573 --> 00:07:15,333
Speaker 2: have a reason that you would move away from your

148
00:07:15,333 --> 00:07:18,453
Speaker 2: current system, right, correct. And it sounds like if it's

149
00:07:18,493 --> 00:07:21,533
Speaker 2: blocked from those major websites, then that then in itself

150
00:07:21,613 --> 00:07:23,653
Speaker 2: is kind of a reason not to be using it.

151
00:07:23,613 --> 00:07:25,893
Speaker 3: It's kind of annoying. Yeah, I will say. The thing

152
00:07:25,933 --> 00:07:27,853
Speaker 3: that I find with this AI stuff is if you

153
00:07:27,893 --> 00:07:31,093
Speaker 3: get it to generate something and you don't know that

154
00:07:31,173 --> 00:07:34,213
Speaker 3: what the outcome is supposed to be. It's incredible, right

155
00:07:34,573 --> 00:07:36,933
Speaker 3: because you can't verify it the moment you try. I

156
00:07:36,973 --> 00:07:39,173
Speaker 3: found personally if you try to use it for something

157
00:07:39,173 --> 00:07:42,533
Speaker 3: where you know what the outcome should be and you

158
00:07:42,773 --> 00:07:45,613
Speaker 3: say you've given it some documents and you're saying, hey,

159
00:07:45,693 --> 00:07:48,013
Speaker 3: summarize this, or take some values from a spreadsheet, or

160
00:07:48,013 --> 00:07:49,813
Speaker 3: do whatever it is trying to do, and then those

161
00:07:49,893 --> 00:07:52,933
Speaker 3: values and things aren't in the output. It's like, Okay,

162
00:07:53,533 --> 00:07:55,253
Speaker 3: can I trust it with the stuff that I thought

163
00:07:55,253 --> 00:07:55,773
Speaker 3: I could trust?

164
00:07:55,973 --> 00:07:56,253
Speaker 2: Yeah?

165
00:07:56,373 --> 00:07:58,213
Speaker 3: Like that to me is where I'm like, Okay, maybe

166
00:07:58,213 --> 00:08:00,333
Speaker 3: I shouldn't be using it as a research tool. Yeah,

167
00:08:00,373 --> 00:08:01,773
Speaker 3: but as long as it gives me the links I

168
00:08:01,773 --> 00:08:03,653
Speaker 3: can then click on and go and verify the information,

169
00:08:03,733 --> 00:08:04,573
Speaker 3: I'm usually okay with it.

170
00:08:04,573 --> 00:08:07,253
Speaker 2: Then you yeah, yeah, very interesting. Hey, thanks Paul. We

171
00:08:07,293 --> 00:08:10,453
Speaker 2: will let you get back to sleep on your smart snoozer,

172
00:08:11,733 --> 00:08:13,493
Speaker 2: Paul Steinhouse. Our textbook is.

173
00:08:13,933 --> 00:08:17,053
Speaker 1: For more from Saturday Morning with Jack Tame. Listen live

174
00:08:17,133 --> 00:08:19,933
Speaker 1: to News Talks at b from nine am Saturday, or

175
00:08:20,013 --> 00:08:21,933
Speaker 1: follow the podcast on iHeartRadio