WEBVTT - Monologue: On Dangerous Rhetoric

0:00:03.160 --> 0:00:08.080
<v Speaker 1>Media. Hello and welcome to this week's Better Offlying Monologue.

0:00:08.119 --> 0:00:18.320
<v Speaker 1>I'm your host ed Zitron. Now I want to be

0:00:18.360 --> 0:00:20.840
<v Speaker 1>abundantly clear about something. It is illegal to throw a

0:00:20.920 --> 0:00:24.400
<v Speaker 1>Molotov cocktail at anyone, and it's morally objectionable to do so.

0:00:24.960 --> 0:00:27.560
<v Speaker 1>I explicitly and fundamentally object to the recent acts of

0:00:27.600 --> 0:00:30.560
<v Speaker 1>violence against Sam Mortman. These acts of violence are not

0:00:30.640 --> 0:00:32.720
<v Speaker 1>something I endorsed in any way. I'm also glad that

0:00:32.760 --> 0:00:36.840
<v Speaker 1>nobody was hurt. It's also morally repugnant for Sam Mortman

0:00:36.920 --> 0:00:39.760
<v Speaker 1>to somehow suggest, by stating that we should de escalate

0:00:39.760 --> 0:00:44.199
<v Speaker 1>the rhetoric and tactics around AI criticism, that the careful, thoughtful, determined,

0:00:44.200 --> 0:00:47.400
<v Speaker 1>and eagerly fair work of ronand Farroh and Andrew Morantz

0:00:47.479 --> 0:00:50.320
<v Speaker 1>is in any way responsible for these acts of violence.

0:00:51.080 --> 0:00:53.960
<v Speaker 1>Doing so is a deliberate and cynical attempt to chill

0:00:54.000 --> 0:00:57.400
<v Speaker 1>the air around criticism of AI and its associated companies.

0:00:58.200 --> 0:01:01.240
<v Speaker 1>I do, however, agree with miss Altman that the rhetoric

0:01:01.280 --> 0:01:04.880
<v Speaker 1>around AI does need to change. Both he and mister

0:01:04.920 --> 0:01:08.880
<v Speaker 1>Amadey of Anthropic need to immediately stop overstating the capabilities

0:01:08.920 --> 0:01:12.479
<v Speaker 1>of large language models. Mister Ortman and mister Abiday should

0:01:12.480 --> 0:01:15.000
<v Speaker 1>not discuss being scared of their models, as they have

0:01:15.080 --> 0:01:18.480
<v Speaker 1>both done so since twenty twenty three, or being uncomfortable

0:01:18.680 --> 0:01:20.960
<v Speaker 1>that men such as they are in control unless they

0:01:20.959 --> 0:01:24.000
<v Speaker 1>wish to shut down their services or They should also

0:01:24.280 --> 0:01:26.640
<v Speaker 1>definitely not say that their models are conscious or suggest

0:01:26.720 --> 0:01:30.200
<v Speaker 1>that they have emotions either. Anthropic is particularly guilty of this.

0:01:31.200 --> 0:01:34.479
<v Speaker 1>These men should immediately stop misleading people through company documentation

0:01:34.880 --> 0:01:37.880
<v Speaker 1>the models are blackmailing people, or, as Anthropic did in

0:01:37.920 --> 0:01:41.640
<v Speaker 1>its Mythoar System card, suggest the model has broken containment

0:01:41.680 --> 0:01:44.400
<v Speaker 1>and sent a message. This was not the case. It

0:01:44.440 --> 0:01:47.400
<v Speaker 1>was in a completely separate instance to the container it

0:01:47.480 --> 0:01:50.600
<v Speaker 1>was messing with, and it was instructed explicitly to do so.

0:01:50.760 --> 0:01:53.200
<v Speaker 1>This is an act of deceit that is scaring people.

0:01:54.000 --> 0:01:57.440
<v Speaker 1>These men must stop discussing threats to jobs without actual

0:01:57.680 --> 0:02:01.680
<v Speaker 1>meaningful data that is significantly more sound and rigorous than

0:02:01.760 --> 0:02:05.040
<v Speaker 1>jobs that might be affected by AI at some point someday,

0:02:05.280 --> 0:02:09.120
<v Speaker 1>but for now we've got a CHAMPO. Mister Amiday should

0:02:09.120 --> 0:02:11.960
<v Speaker 1>immediately cease any and all discussions of AI potentially or

0:02:12.000 --> 0:02:15.440
<v Speaker 1>otherwise eliminating fifty percent or any level of white collar jobs.

0:02:15.720 --> 0:02:19.160
<v Speaker 1>As mister Ortman should cease predicting when superintelligence might arrive,

0:02:19.480 --> 0:02:23.519
<v Speaker 1>as mister Amiday should actively reject and denounce any suggestions

0:02:23.639 --> 0:02:26.919
<v Speaker 1>of AI creating a white collar blood buff This dangerous

0:02:26.960 --> 0:02:30.800
<v Speaker 1>rhetoric is scaring people. Those that defend AI labs, and

0:02:30.840 --> 0:02:33.840
<v Speaker 1>indeed people have spoken to are anthropic, will claim that

0:02:33.840 --> 0:02:36.959
<v Speaker 1>these are difficult conversations that need to be had, when

0:02:36.960 --> 0:02:41.440
<v Speaker 1>in actuality, these conversations engage in dangerous and frightening rhetoric

0:02:41.639 --> 0:02:45.040
<v Speaker 1>as a means of boosting a company's valuation and garnering attention.

0:02:46.000 --> 0:02:48.520
<v Speaker 1>If either of these men truly believed the things they

0:02:48.520 --> 0:02:51.000
<v Speaker 1>were saying were true, they would do something about it

0:02:51.160 --> 0:02:53.280
<v Speaker 1>other than saying, you should be scared of us and

0:02:53.320 --> 0:02:55.720
<v Speaker 1>the things we're making, and we're the only ones brave

0:02:55.800 --> 0:02:59.840
<v Speaker 1>enough to say anything about it. These conversations are also

0:02:59.880 --> 0:03:02.280
<v Speaker 1>not sensical and misleading when you compare them to what

0:03:02.440 --> 0:03:06.079
<v Speaker 1>large language models can actually do, and this rhetoric is

0:03:06.120 --> 0:03:08.560
<v Speaker 1>a blatant attempt to scare people into paying for software

0:03:08.600 --> 0:03:11.119
<v Speaker 1>today based on what it absolutely cannot do and will

0:03:11.160 --> 0:03:13.720
<v Speaker 1>not do in the future. It is an attempt to

0:03:13.760 --> 0:03:16.320
<v Speaker 1>obfuscate the actual efficacy of a technology as a means

0:03:16.320 --> 0:03:20.480
<v Speaker 1>of deceiving investors, the media, and the general public. Both

0:03:20.560 --> 0:03:23.360
<v Speaker 1>Ortman and Amiday engage in the language of AI dooomerism

0:03:23.440 --> 0:03:26.520
<v Speaker 1>as a means of generating attention, revenue, and investment capital,

0:03:27.000 --> 0:03:30.320
<v Speaker 1>actively selling their software and future investment potential based on

0:03:30.360 --> 0:03:33.720
<v Speaker 1>their ownership of the technology that they say disingenuously is

0:03:33.760 --> 0:03:37.640
<v Speaker 1>potentially going to take everyone's jobs now. Based on the

0:03:37.720 --> 0:03:40.560
<v Speaker 1>reports from his Instagram, the man who threw the Molotov

0:03:40.600 --> 0:03:43.560
<v Speaker 1>cocktail at Sam Ortman's house was at least partially inspired

0:03:43.600 --> 0:03:46.280
<v Speaker 1>by If anyone builds It, Everyone Dies, a doom of

0:03:46.360 --> 0:03:49.520
<v Speaker 1>porn fantasy written by a pair of overlea verbost dunces

0:03:49.720 --> 0:03:52.760
<v Speaker 1>spreading fearful language about the power of AI, inspired by

0:03:52.760 --> 0:03:56.560
<v Speaker 1>the fear mongering of Altman himself. Aortman once suggested that

0:03:56.720 --> 0:04:01.520
<v Speaker 1>Aliaisa Yudkowski might deserve the Nobel Peace Prize one day.

0:04:02.640 --> 0:04:04.240
<v Speaker 1>I also think we need to be clear about the

0:04:04.320 --> 0:04:07.440
<v Speaker 1>circumstances and the rhetoric that led someone to do this,

0:04:07.840 --> 0:04:10.040
<v Speaker 1>and why the AI industry needs to be well aware

0:04:10.080 --> 0:04:13.240
<v Speaker 1>that the society they're continually threatening with job loss is

0:04:13.240 --> 0:04:15.640
<v Speaker 1>one full of people that are very, very close to

0:04:15.680 --> 0:04:19.279
<v Speaker 1>the edge. This is not about anybody being deserving of

0:04:19.320 --> 0:04:24.080
<v Speaker 1>anything but a frank evaluation of cause and effect. Many

0:04:24.120 --> 0:04:26.720
<v Speaker 1>people feel like they're being fucking tortured every time they

0:04:26.760 --> 0:04:29.880
<v Speaker 1>learn social media their money doesn't go as far their

0:04:29.920 --> 0:04:33.840
<v Speaker 1>financial situation has never been worse. Every time they read something,

0:04:33.920 --> 0:04:36.680
<v Speaker 1>it's a story about ice patrols or a near nuclear

0:04:36.720 --> 0:04:39.599
<v Speaker 1>war in Iran, or the gas is more expensive, or

0:04:39.600 --> 0:04:42.279
<v Speaker 1>that there are more worrying things happening in private credit.

0:04:42.720 --> 0:04:47.320
<v Speaker 1>Nobody can afford a house, and layoffs are constant. One group, however,

0:04:47.440 --> 0:04:50.000
<v Speaker 1>appears to exist in an alternative world where anything they

0:04:50.080 --> 0:04:53.200
<v Speaker 1>want is possible. AI people can raise as much money

0:04:53.200 --> 0:04:55.239
<v Speaker 1>as they want, They can build as big a building

0:04:55.279 --> 0:04:57.839
<v Speaker 1>as they want anywhere in the world. Everything they do

0:04:57.960 --> 0:05:00.560
<v Speaker 1>is taken so seriously that the government will call meeting

0:05:00.600 --> 0:05:04.480
<v Speaker 1>about it. Every single media outlet talks about everything they do.

0:05:04.800 --> 0:05:07.240
<v Speaker 1>Your boss forces you to use it. Every piece of

0:05:07.279 --> 0:05:10.240
<v Speaker 1>software forces you to at least acknowledge that they use

0:05:10.279 --> 0:05:13.800
<v Speaker 1>AI too. Everyone is talking about it with complete certainty,

0:05:14.040 --> 0:05:17.320
<v Speaker 1>despite not being completely clear as to why they're doing so.

0:05:18.320 --> 0:05:22.599
<v Speaker 1>As many people writhe in continual agony and fear, AI

0:05:22.760 --> 0:05:26.640
<v Speaker 1>promises but never quite delivers some sort of vague utopia

0:05:26.680 --> 0:05:30.680
<v Speaker 1>at the highest cost known to man, and these companies are,

0:05:30.720 --> 0:05:34.000
<v Speaker 1>in no uncertain terms, coming for your job. That's what

0:05:34.000 --> 0:05:36.600
<v Speaker 1>they want to do. They all say it. They use

0:05:36.640 --> 0:05:40.359
<v Speaker 1>deceptively worded studies that talk about AI exposed careers to

0:05:40.480 --> 0:05:43.719
<v Speaker 1>scare and mislead people into believing lms are coming for

0:05:43.760 --> 0:05:48.039
<v Speaker 1>their jobs, all while spreading vague proclamations about how said

0:05:48.160 --> 0:05:51.919
<v Speaker 1>job loss is imminent but also always twelve months away.

0:05:52.760 --> 0:05:55.960
<v Speaker 1>Ortmann even says that jobs that will vanish weren't real

0:05:56.240 --> 0:05:59.480
<v Speaker 1>work to begin with, much as former open ai CTO

0:05:59.720 --> 0:06:03.279
<v Speaker 1>mirror Mrati said that some creative jobs shouldn't have existed

0:06:03.279 --> 0:06:07.039
<v Speaker 1>in the first place. These are people who sell a

0:06:07.080 --> 0:06:10.480
<v Speaker 1>product with no benefit comparable on any level to its

0:06:10.560 --> 0:06:14.719
<v Speaker 1>ruinous trillion dollar cost, both financially and to the environment,

0:06:14.920 --> 0:06:17.200
<v Speaker 1>and are able to get away with anything and get

0:06:17.240 --> 0:06:19.720
<v Speaker 1>anything they want, at a time when those who work

0:06:19.760 --> 0:06:22.400
<v Speaker 1>hard are given a kick in the fucking teeth, sneered

0:06:22.400 --> 0:06:24.800
<v Speaker 1>out for not using AI that doesn't actually seem to

0:06:24.800 --> 0:06:27.400
<v Speaker 1>make their lives easier, and then told that their labor

0:06:27.560 --> 0:06:30.760
<v Speaker 1>doesn't constitute real work by people that don't appear to

0:06:30.760 --> 0:06:35.200
<v Speaker 1>do anything other than go on fucking CNBC. At a

0:06:35.240 --> 0:06:38.400
<v Speaker 1>time when nobody living a normal life feels like they

0:06:38.440 --> 0:06:41.640
<v Speaker 1>have enough. The AI industry always seems to get more.

0:06:42.320 --> 0:06:45.200
<v Speaker 1>There's not enough money for free college or housing, or

0:06:45.240 --> 0:06:50.040
<v Speaker 1>healthcare or daycare, but there's always more money for AI compute.

0:06:50.160 --> 0:06:53.599
<v Speaker 1>Regular people face the harshest credit market in generations, but

0:06:53.720 --> 0:06:57.520
<v Speaker 1>private credit and specifically data centers can always always get

0:06:57.520 --> 0:07:01.160
<v Speaker 1>more money and more LANDI can never fail. It can

0:07:01.200 --> 0:07:04.479
<v Speaker 1>only be failed if it doesn't work. You simply don't

0:07:04.520 --> 0:07:06.520
<v Speaker 1>know how to use AI properly and will be at

0:07:06.520 --> 0:07:09.920
<v Speaker 1>a huge disadvantage. Despite the sales pitch being this is

0:07:09.960 --> 0:07:14.520
<v Speaker 1>intelligence software that just does stuff. AI companies can get

0:07:14.560 --> 0:07:17.600
<v Speaker 1>as much attention as they need, their failings explained away.

0:07:17.880 --> 0:07:20.960
<v Speaker 1>Their meager success is celebrated like the ball dropping on

0:07:21.000 --> 0:07:23.680
<v Speaker 1>New Year's Eve. They are half our sub war of

0:07:23.720 --> 0:07:27.080
<v Speaker 1>the world's myth thos horseshit treated like they've opened the

0:07:27.120 --> 0:07:41.320
<v Speaker 1>gates of hell. Regular people feel ignored and like they're

0:07:41.360 --> 0:07:43.920
<v Speaker 1>not taken seriously, and the people being given the most

0:07:43.920 --> 0:07:47.160
<v Speaker 1>money and attention are the ones loudly saying we're richer

0:07:47.200 --> 0:07:49.640
<v Speaker 1>than anyone has ever been, we intend to spend more

0:07:49.640 --> 0:07:52.360
<v Speaker 1>than anyone has ever spent, and we intend to take

0:07:52.440 --> 0:07:56.640
<v Speaker 1>your job he mentally unstable took them seriously? Did they

0:07:56.680 --> 0:08:00.160
<v Speaker 1>not think that people would be angry constantly? Tall Talking

0:08:00.200 --> 0:08:02.720
<v Speaker 1>about how your company will make an indeterminate amount of

0:08:02.720 --> 0:08:05.560
<v Speaker 1>people jobless while also being able to raise over one

0:08:05.640 --> 0:08:08.240
<v Speaker 1>hundred and sixty two billion dollars in the space of

0:08:08.320 --> 0:08:10.800
<v Speaker 1>two years and taking up as much space on Earth

0:08:10.840 --> 0:08:13.720
<v Speaker 1>as you please is something that could send people over

0:08:13.760 --> 0:08:16.960
<v Speaker 1>the edge. Every day the news reminds you that everything

0:08:17.080 --> 0:08:20.160
<v Speaker 1>sucks and is more expensive, unless, of course, you're an AI,

0:08:20.280 --> 0:08:22.280
<v Speaker 1>where you'll be given as much money as you want

0:08:22.320 --> 0:08:25.440
<v Speaker 1>and told you're the most special person alive. I can

0:08:25.480 --> 0:08:28.000
<v Speaker 1>imagine it tearing at a person's soul as the world

0:08:28.040 --> 0:08:31.600
<v Speaker 1>beats them down. And I will say that what they

0:08:31.600 --> 0:08:34.840
<v Speaker 1>did was a disgraceful act of violence. But what do

0:08:34.880 --> 0:08:37.400
<v Speaker 1>you think happens when you go to people for years

0:08:37.400 --> 0:08:41.800
<v Speaker 1>and years and years and accumulate so much wealth. Unstable

0:08:41.840 --> 0:08:44.839
<v Speaker 1>people in various stages of torment act in erratic and

0:08:44.960 --> 0:08:48.959
<v Speaker 1>dangerous ways. The suspect in the Molotov cocktail incident apparently

0:08:49.000 --> 0:08:51.240
<v Speaker 1>had a manifest though where he had listed the names

0:08:51.240 --> 0:08:54.559
<v Speaker 1>and addresses of both Altman and multiple other AI executives

0:08:54.640 --> 0:08:58.800
<v Speaker 1>and PERCYNBC discussed the threat of AI to humanity as

0:08:58.800 --> 0:09:02.400
<v Speaker 1>a justification for his acts. I'm genuinely happy to hear

0:09:02.480 --> 0:09:06.280
<v Speaker 1>this person was apprehended without anyone being hurt. These actions

0:09:06.320 --> 0:09:08.880
<v Speaker 1>are morally wrong and are also the direct result of

0:09:08.880 --> 0:09:12.840
<v Speaker 1>the AI industry's deceptive and manipulative scare campaign, one promoted

0:09:12.880 --> 0:09:15.280
<v Speaker 1>by men like Altman and Amiday, as well as DOOMA

0:09:15.320 --> 0:09:18.600
<v Speaker 1>fan fiction writers like Yudkowski and of course Daniel Koko

0:09:18.640 --> 0:09:22.080
<v Speaker 1>Tachlow of AI twenty twenty seven, both of whom have

0:09:22.160 --> 0:09:24.559
<v Speaker 1>had their work validated and propagated via The New York

0:09:24.600 --> 0:09:26.960
<v Speaker 1>Times and Kevin Rus who should feel fucking ashamed of

0:09:27.040 --> 0:09:33.080
<v Speaker 1>himself forever ever platforming these freaks. It's a disgrace. And

0:09:33.120 --> 0:09:36.240
<v Speaker 1>on the subject of dangerous rhetoric, I think we need

0:09:36.280 --> 0:09:38.360
<v Speaker 1>to reckon with the fact that the mainstream media has

0:09:38.360 --> 0:09:41.440
<v Speaker 1>helped spread harmful propaganda and that a lack of scrutiny

0:09:41.480 --> 0:09:45.640
<v Speaker 1>of said propaganda is now causing genuine harm. I also

0:09:45.720 --> 0:09:48.120
<v Speaker 1>do not hear any attempts by mister Altman to deal

0:09:48.160 --> 0:09:51.439
<v Speaker 1>with the actual documented threat of AI psychosis and the

0:09:51.480 --> 0:09:53.880
<v Speaker 1>people that have been twisted by large language models to

0:09:53.920 --> 0:09:57.080
<v Speaker 1>take their lives and those of others. These are acts

0:09:57.120 --> 0:10:00.000
<v Speaker 1>of violence that could have been stopped had chat GPTs

0:10:00.080 --> 0:10:04.559
<v Speaker 1>in similar applications not anthropomorphized by design and trained to

0:10:04.600 --> 0:10:08.800
<v Speaker 1>be friendly. I genuinely believe that these things should not

0:10:08.880 --> 0:10:12.240
<v Speaker 1>be able to have conversations that shouldn't happen. They should

0:10:12.240 --> 0:10:16.080
<v Speaker 1>act like computers, they should act like software. These people

0:10:16.280 --> 0:10:19.960
<v Speaker 1>have deliberately made that not the case. They've deliberately tried

0:10:20.000 --> 0:10:22.280
<v Speaker 1>to make them friends, and are now trying to wash

0:10:22.320 --> 0:10:26.400
<v Speaker 1>their hands of their responsibility. Na Na, It wasn't an

0:10:26.440 --> 0:10:29.480
<v Speaker 1>act of violence inspired by Ronan Farrow publishing a piece

0:10:29.520 --> 0:10:32.360
<v Speaker 1>about Sam Moltman. It was the result of a year's

0:10:32.400 --> 0:10:35.520
<v Speaker 1>long publicity campaign that has since the beginning been about

0:10:35.600 --> 0:10:38.600
<v Speaker 1>how scary the technology is and how much money its

0:10:38.600 --> 0:10:41.040
<v Speaker 1>owners make and how many jobs they intend to take

0:10:41.080 --> 0:10:45.400
<v Speaker 1>as a result. I separately believe these executives and their

0:10:45.440 --> 0:10:48.520
<v Speaker 1>cohort are intentionally scaring people as a means of growing

0:10:48.520 --> 0:10:51.600
<v Speaker 1>their companies, and that these continual statements of we're making

0:10:51.640 --> 0:10:53.400
<v Speaker 1>something to take your job, or we need more money

0:10:53.440 --> 0:10:55.440
<v Speaker 1>in space to do it could be construed as a

0:10:55.480 --> 0:10:58.840
<v Speaker 1>threat by somebody that's already on the edge. I agree

0:10:58.840 --> 0:11:03.120
<v Speaker 1>that the dangerous rhetoric. AI must stop. Dario Amiday and

0:11:03.200 --> 0:11:07.600
<v Speaker 1>Sam Mortman must immediately seize their manipulative and disingenuous scare

0:11:07.640 --> 0:11:11.520
<v Speaker 1>tactics and begin describing large language models in terms that

0:11:11.559 --> 0:11:15.880
<v Speaker 1>match their actual abilities, all while dispensing with any further

0:11:15.920 --> 0:11:19.920
<v Speaker 1>attempts to extrapolate their future capabilities. Enough with the fluff,

0:11:20.400 --> 0:11:23.760
<v Speaker 1>enough with the bullshit, Stop talking about AGI. Start talking

0:11:23.760 --> 0:11:27.319
<v Speaker 1>about this like regular, old, boring software, because that's all

0:11:27.360 --> 0:11:31.080
<v Speaker 1>that CHAT, GPT and CLAUDE are. In the end, if

0:11:31.160 --> 0:11:34.120
<v Speaker 1>Altman wants to engage in good faith criticism, he should

0:11:34.120 --> 0:11:36.439
<v Speaker 1>actually act in good faith, and the same goes for

0:11:36.520 --> 0:11:41.000
<v Speaker 1>Dario Amiday. This starts with taking ownership of their roles

0:11:41.080 --> 0:11:44.920
<v Speaker 1>in a global disinformation campaign. It starts with them recognizing

0:11:44.920 --> 0:11:47.680
<v Speaker 1>how the AI industry has sold itself based on spreading

0:11:47.720 --> 0:11:51.679
<v Speaker 1>mythology with the intent of creating unrest and fear. And

0:11:51.720 --> 0:11:55.360
<v Speaker 1>it starts with Altman and Amiday and their ILK accepting

0:11:55.440 --> 0:11:59.440
<v Speaker 1>any responsibility for their actions. I'm not holding my breath