1
00:00:02,160 --> 00:00:07,119
Speaker 1: Zone Media. Hello and welcome to this week's Better Offline Monologue.

2
00:00:07,200 --> 00:00:17,159
Speaker 1: I'm at Zychra and your host. As a reminder, you

3
00:00:17,200 --> 00:00:19,800
Speaker 1: can buy Better Offline merchandise now. You'll find a link

4
00:00:19,840 --> 00:00:21,639
Speaker 1: to it in the episode notes. You all seem to

5
00:00:21,640 --> 00:00:24,240
Speaker 1: really like it. It's cool stuff. But I know some

6
00:00:24,280 --> 00:00:26,680
Speaker 1: of you have also said recently that I've done too

7
00:00:26,840 --> 00:00:30,080
Speaker 1: much AI stuff recently, and I wanted to take a

8
00:00:30,120 --> 00:00:32,080
Speaker 1: little time to explain why I've been doing so, in

9
00:00:32,120 --> 00:00:34,960
Speaker 1: part because I'm not going to change the generative AI

10
00:00:35,040 --> 00:00:38,199
Speaker 1: boom is about far more than art, official intelligence, or

11
00:00:38,280 --> 00:00:40,760
Speaker 1: cloud storage or data centers or what part of Hawaii

12
00:00:40,800 --> 00:00:43,760
Speaker 1: Mark Zuckerberg will buy next. I believe this movement is

13
00:00:43,760 --> 00:00:46,199
Speaker 1: symbolic of a greater rot in the tech industry and

14
00:00:46,240 --> 00:00:49,280
Speaker 1: indeed in media criticism, both inside and outside of the

15
00:00:49,320 --> 00:00:51,960
Speaker 1: tech media, and that the nature of criticism must indeed

16
00:00:52,040 --> 00:00:54,440
Speaker 1: change to meet this moment. Now, a few of the

17
00:00:54,480 --> 00:00:57,080
Speaker 1: most consistent critiques in my work are mostly around my tone.

18
00:00:57,080 --> 00:00:59,200
Speaker 1: I'm a hater, I'm a cynic, I'm a skeptic, I'm

19
00:00:59,360 --> 00:01:02,840
Speaker 1: too emotional, I've gone overboard. The local townspeople should throw

20
00:01:02,880 --> 00:01:05,280
Speaker 1: tomatoes at me. I should be treated in the way

21
00:01:05,319 --> 00:01:08,240
Speaker 1: of Shrek and exiled from society, things like that. But

22
00:01:08,360 --> 00:01:10,560
Speaker 1: in all seriousness, it seems the only way my work

23
00:01:10,600 --> 00:01:14,240
Speaker 1: is reliably critiqued is to suggest that my emotions invalidate

24
00:01:14,240 --> 00:01:18,880
Speaker 1: my criticisms somehow, and that saying fuck somehow weakens my arguments. Basically,

25
00:01:18,920 --> 00:01:22,679
Speaker 1: that giving a shit is invalidating when it comes to

26
00:01:22,720 --> 00:01:25,440
Speaker 1: criticizing something. And I find that putrid, by the way,

27
00:01:25,480 --> 00:01:27,760
Speaker 1: but it's how things are, and I kind of get it.

28
00:01:28,120 --> 00:01:31,200
Speaker 1: Most financial tech criticism is expected to be dry and clinical,

29
00:01:31,240 --> 00:01:33,600
Speaker 1: and any emotional content should be positive or at the

30
00:01:33,720 --> 00:01:36,600
Speaker 1: very least optimistic, And I do not believe this tone

31
00:01:36,600 --> 00:01:38,800
Speaker 1: is sufficient for the seriousness of the matters at hand.

32
00:01:39,440 --> 00:01:42,080
Speaker 1: We are, as I have said repeatedly several years, into

33
00:01:42,120 --> 00:01:44,880
Speaker 1: a hysterical, illogical bubble, one launched off the back of

34
00:01:44,920 --> 00:01:48,160
Speaker 1: two smaller yet no less hysterical bubbles, at times by

35
00:01:48,200 --> 00:01:49,960
Speaker 1: the same people, some of them writing for the New

36
00:01:50,000 --> 00:01:52,720
Speaker 1: York Times of the verge, We as a society have

37
00:01:52,800 --> 00:01:55,600
Speaker 1: accepted the terms that the generative AI companies want us to.

38
00:01:56,040 --> 00:01:58,559
Speaker 1: That we must destroy the environment, that we must steal

39
00:01:58,560 --> 00:02:01,000
Speaker 1: from millions of people, that we must burn billions of

40
00:02:01,040 --> 00:02:03,720
Speaker 1: dollars all in pursuit of a vague and specious outcome

41
00:02:03,760 --> 00:02:06,920
Speaker 1: that will never really arrive. This isn't to say that

42
00:02:06,960 --> 00:02:09,880
Speaker 1: we've had much choice in said acceptance. The media has

43
00:02:09,919 --> 00:02:13,400
Speaker 1: repeatedly accepted and promoted these narratives, helped justify these costs,

44
00:02:13,520 --> 00:02:16,960
Speaker 1: and presented ridiculous narratives as sensible ones. The outcome has

45
00:02:16,960 --> 00:02:19,960
Speaker 1: been a massive transfer of wealth upwards, both into the

46
00:02:19,960 --> 00:02:23,000
Speaker 1: market capitalization of hyperscalers and into the pockets of Sam

47
00:02:23,000 --> 00:02:26,000
Speaker 1: Mortman and other founders pretending generative AI will become some

48
00:02:26,040 --> 00:02:29,280
Speaker 1: sort of conscious intelligence or the next hypergrowth market, regardless

49
00:02:29,280 --> 00:02:32,239
Speaker 1: of whether of these things is possible. The other outcome

50
00:02:32,240 --> 00:02:34,880
Speaker 1: has been the rise of many new kinds of grifters,

51
00:02:35,320 --> 00:02:38,240
Speaker 1: the ethermolics of the world that creates scientific sounding yet

52
00:02:38,280 --> 00:02:41,560
Speaker 1: specious reviews of AI models, the slop filled AI newsletters

53
00:02:41,560 --> 00:02:44,480
Speaker 1: with fake subscribe accounts, and the many many AI influencers

54
00:02:44,639 --> 00:02:48,480
Speaker 1: that exist as extensions of AI companies pr departments. Yet

55
00:02:48,480 --> 00:02:50,360
Speaker 1: the most worrying part has been the members of the

56
00:02:50,400 --> 00:02:52,600
Speaker 1: media that have allowed this to happen. This isn't to

57
00:02:52,600 --> 00:02:55,280
Speaker 1: say anybody is even Kevin rus and Case and Newton

58
00:02:55,280 --> 00:02:57,800
Speaker 1: are included here. I don't think anyone's corrupt, but the

59
00:02:57,840 --> 00:03:00,640
Speaker 1: media is unprepared, unwilling, or able to push back on

60
00:03:00,680 --> 00:03:03,840
Speaker 1: the narratives. The structures that hold up tech and business

61
00:03:03,840 --> 00:03:06,480
Speaker 1: media are not built to truly explain or criticize what's

62
00:03:06,520 --> 00:03:09,239
Speaker 1: happening in the tech industry or even the economy at large.

63
00:03:10,560 --> 00:03:12,760
Speaker 1: These structures are built not to ask is this real?

64
00:03:12,880 --> 00:03:15,000
Speaker 1: Or will this work? But when this works, what will

65
00:03:15,040 --> 00:03:18,000
Speaker 1: it look like. They're built not to question whether an

66
00:03:18,040 --> 00:03:21,320
Speaker 1: industry is like real at all, but to assume that

67
00:03:21,639 --> 00:03:24,040
Speaker 1: there are always risks in building anything, and thus the

68
00:03:24,080 --> 00:03:26,360
Speaker 1: most important part is discussing what the person in question

69
00:03:26,480 --> 00:03:30,840
Speaker 1: wants to happen. Media's desperation for objectivity regularly deprives the

70
00:03:30,880 --> 00:03:33,800
Speaker 1: reader of the value of being objective because objectivity is

71
00:03:33,840 --> 00:03:39,000
Speaker 1: being conflated with being passive. True objective journalism, which is impossible,

72
00:03:39,040 --> 00:03:41,000
Speaker 1: by the way, would say that both the Open AI

73
00:03:41,120 --> 00:03:43,360
Speaker 1: raised a bunch of money and loses billions a year,

74
00:03:43,400 --> 00:03:45,440
Speaker 1: and that to continue doing business they'll have to raise

75
00:03:45,520 --> 00:03:49,240
Speaker 1: unbelievable sums of money every year. Instead, articles in open

76
00:03:49,240 --> 00:03:51,240
Speaker 1: AI just print that they raised money and why they

77
00:03:51,320 --> 00:03:54,360
Speaker 1: raised it, or that their products do something. No interest

78
00:03:54,400 --> 00:03:56,800
Speaker 1: in finding out what there's something is whether it's important

79
00:03:56,800 --> 00:03:59,160
Speaker 1: and whether it will lead to any outcome, probably because

80
00:03:59,160 --> 00:04:02,320
Speaker 1: it won't. The initial consequences of this passivity are that

81
00:04:02,440 --> 00:04:05,440
Speaker 1: venture capitalist anders select few startup founders have become very,

82
00:04:05,600 --> 00:04:08,040
Speaker 1: very rich, The big tech firms have had something new

83
00:04:08,080 --> 00:04:10,480
Speaker 1: to hock to their customers, and the access journalists have

84
00:04:10,520 --> 00:04:12,520
Speaker 1: had a new thing to pretend they care about. And

85
00:04:12,600 --> 00:04:14,680
Speaker 1: also a fallow tech industry has had something to get

86
00:04:14,680 --> 00:04:17,520
Speaker 1: excited about. None of this money has trickled down to

87
00:04:17,560 --> 00:04:20,119
Speaker 1: anybody other than the powerful, nor have any mass market

88
00:04:20,240 --> 00:04:24,040
Speaker 1: productivity gains been realized, nor has society improved as a result.

89
00:04:24,440 --> 00:04:26,480
Speaker 1: The only things that appear to have changed is that

90
00:04:26,520 --> 00:04:28,599
Speaker 1: these companies need more money, and they don't even need

91
00:04:28,640 --> 00:04:31,160
Speaker 1: to tell anybody why. The media just assumes they need

92
00:04:31,200 --> 00:04:34,080
Speaker 1: it to build powerful AI. I think it's fair to

93
00:04:34,120 --> 00:04:36,240
Speaker 1: say at this point that the generative AI boom hasn't

94
00:04:36,240 --> 00:04:38,320
Speaker 1: done much of anything other than create new ways for

95
00:04:38,360 --> 00:04:41,640
Speaker 1: software companies to sell software or access to models. There

96
00:04:41,680 --> 00:04:43,680
Speaker 1: are no killer apps, no major shifts in the way

97
00:04:43,680 --> 00:04:45,839
Speaker 1: we live our lives outside of innovations in fraud and

98
00:04:45,839 --> 00:04:48,240
Speaker 1: harms to our power grid, and whatever use cases there

99
00:04:48,279 --> 00:04:50,840
Speaker 1: may be for large language models. Are miniscule in comparison

100
00:04:50,839 --> 00:04:52,640
Speaker 1: to the way that AI is discussed in the media.

101
00:04:53,480 --> 00:04:56,760
Speaker 1: I am justifiably angry because I watch these bubbles form

102
00:04:56,760 --> 00:04:59,480
Speaker 1: again and again in exactly the same way at every

103
00:04:59,520 --> 00:05:02,120
Speaker 1: single p With the metaverus, with cryptocurrency, and now with

104
00:05:02,160 --> 00:05:04,880
Speaker 1: generative AI, there have been obvious moments to say, yeah,

105
00:05:04,960 --> 00:05:07,279
Speaker 1: I get that this is what you say will happen,

106
00:05:07,440 --> 00:05:09,839
Speaker 1: but what's happening today doesn't suggest that it will happen

107
00:05:09,880 --> 00:05:12,440
Speaker 1: at all. And every time those moments have been missed,

108
00:05:12,800 --> 00:05:14,800
Speaker 1: and the media is opted instead to ask, but what

109
00:05:14,920 --> 00:05:18,200
Speaker 1: if it was true? When the media opts to trust

110
00:05:18,200 --> 00:05:20,359
Speaker 1: whatever comes out of the mouth of a powerful person,

111
00:05:20,400 --> 00:05:24,840
Speaker 1: the beneficiary is always always the powerful person in question. Yet,

112
00:05:24,880 --> 00:05:26,680
Speaker 1: a better world would be one where the altmans and

113
00:05:26,720 --> 00:05:29,520
Speaker 1: Amma days have to actually justify themselves, show what the

114
00:05:29,520 --> 00:05:32,440
Speaker 1: models can do, give realistic projections. Know that they can't

115
00:05:32,440 --> 00:05:35,200
Speaker 1: just say whatever and get quoted automatically, because this kind

116
00:05:35,200 --> 00:05:38,440
Speaker 1: of accountability would make their current work overstanding their models capabilities,

117
00:05:38,440 --> 00:05:42,159
Speaker 1: and running unsustainable and destructive businesses impossible or at least

118
00:05:42,200 --> 00:05:45,280
Speaker 1: much more difficult. A better tech industry is one where

119
00:05:45,320 --> 00:05:48,159
Speaker 1: the products we hear about actually exist, where hype cycles

120
00:05:48,200 --> 00:05:50,640
Speaker 1: are built based on execution and outcomes rather than whether

121
00:05:50,720 --> 00:05:54,000
Speaker 1: something may or may not be in the future. While

122
00:05:54,000 --> 00:05:56,279
Speaker 1: there is always a place in the tech media to dream,

123
00:05:56,279 --> 00:05:58,279
Speaker 1: to guess what might come next, to talk to people

124
00:05:58,320 --> 00:06:00,800
Speaker 1: inventing things, to talk to researchers, and I have no

125
00:06:00,839 --> 00:06:03,560
Speaker 1: problem with that, and it's fine to anticipate what the

126
00:06:03,640 --> 00:06:05,719
Speaker 1: effects of these things could be, so much of this

127
00:06:05,839 --> 00:06:08,200
Speaker 1: industry has become about those dreams, to the point that

128
00:06:08,240 --> 00:06:10,560
Speaker 1: innovation is far less relevant than whether you can convince

129
00:06:10,640 --> 00:06:14,320
Speaker 1: enough people that something might happen. Nevertheless, there is definitely

130
00:06:14,320 --> 00:06:17,320
Speaker 1: something missing in Better Offline, and that's excitement. The feedback

131
00:06:17,320 --> 00:06:19,320
Speaker 1: I got from my last monologue around fitness and tech

132
00:06:19,400 --> 00:06:22,880
Speaker 1: I liked was profound, and I will be looking forward

133
00:06:22,880 --> 00:06:25,920
Speaker 1: to doing more episodes like that, especially these monologues on

134
00:06:25,960 --> 00:06:27,680
Speaker 1: the things that I use that I really enjoy, all

135
00:06:27,720 --> 00:06:29,800
Speaker 1: the cool stuff that tech is actually doing. It's going

136
00:06:29,880 --> 00:06:31,400
Speaker 1: to be a process to get there, but I think

137
00:06:31,440 --> 00:06:33,680
Speaker 1: you're really gonna like what you hear when I start

138
00:06:33,760 --> 00:06:36,480
Speaker 1: doing it. Don't worry, though, there's still so much rots

139
00:06:36,480 --> 00:06:44,240
Speaker 1: economy bullshit to unpack, as you'll hear in tomorrow's episode