Urban75 Home About Offline BrixtonBuzz Contact

Expansion of AI and political / social impacts...

Good listen for some pragmatic cynicism about AI - something not on the hype train of 'we'll all live forever as AIs' or 'SkyNet will murder your mum': Why We Must Resist AI w/ Dan McQuillan - Tech Won’t Save Us

Talking of the hype train though, Sam Altman has been reinstated at OpenAI (ChatGPT, Dall-E, Microsoft 'vassal') and he's an arch fantasist (imo), all over the AI means we're creating God (and that's good) end of things. He's a pretty good model for a lot of the AI industry/Tech Bros - massive privilege mixed in with a load of half read Sci-Fi and zero self restraint. And Open AI is a pretty interesting entity too, prone to 'AI for Good' posturing I still kind of think it's there to take the hits from the inevitable lawsuits GAI generates. That's why Microsoft has pumped money into it, they can implement the tech if they like but not face the legal problems it generates. Previously also backed by Elon Musk and, I think, Peter Thiel.
 
Good listen for some pragmatic cynicism about AI - something not on the hype train of 'we'll all live forever as AIs' or 'SkyNet will murder your mum': Why We Must Resist AI w/ Dan McQuillan - Tech Won’t Save Us

Talking of the hype train though, Sam Altman has been reinstated at OpenAI (ChatGPT, Dall-E, Microsoft 'vassal') and he's an arch fantasist (imo), all over the AI means we're creating God (and that's good) end of things. He's a pretty good model for a lot of the AI industry/Tech Bros - massive privilege mixed in with a load of half read Sci-Fi and zero self restraint. And Open AI is a pretty interesting entity too, prone to 'AI for Good' posturing I still kind of think it's there to take the hits from the inevitable lawsuits GAI generates. That's why Microsoft has pumped money into it, they can implement the tech if they like but not face the legal problems it generates. Previously also backed by Elon Musk and, I think, Peter Thiel.

And talking of Open AI and lawsuits - Google Researchers’ Attack Prompts ChatGPT to Reveal Its Training Data

Basically proves that they've been stealing like kleptomaniacs for their datasets, something which has been obvious for a long time but which they've always been very careful about not confessing to.
 
Talking of the hype train though, Sam Altman has been reinstated at OpenAI (ChatGPT, Dall-E, Microsoft 'vassal') and he's an arch fantasist (imo), all over the AI means we're creating God (and that's good) end of things. He's a pretty good model for a lot of the AI industry/Tech Bros - massive privilege mixed in with a load of half read Sci-Fi and zero self restraint.
Some other Sci-Fi. "Once men turned their thinking over to machines in the hope that this would set them free. But that only permitted other men with machines to enslave them." ― Frank Herbert, Dune
 
The shitty stuff will be the use of it to make political pitches for the worst kind of people, to make very persuasive and targeted arguments to win people over to whatever far right libertarian nonsense the deep-pocketed sponsors want. Think what Cambridge Analytical did but with more depth and power.

Will also be used to improve phishing attacks, no longer will poor English be a barrier for crooked bastards.
 
The shitty stuff will be the use of it to make political pitches for the worst kind of people, to make very persuasive and targeted arguments to win people over to whatever far right libertarian nonsense the deep-pocketed sponsors want. Think what Cambridge Analytical did but with more depth and power.

Will also be used to improve phishing attacks, no longer will poor English be a barrier for crooked bastards.

Tbh I think that's the fairly minimal end of it, the structures of misinformation are already pretty effective at doing all that anyway. As far as political narratives go I'd be more worried about the massive potential for chaotic conspiricism and flat out noise. When AI systems are so completely incomprehensible to most people pretty much any story told about them can seem plausible, plus the general noise of GAI can flood any cultural system. I'm not sure those who'd seek to use it for political gain really have the first clue of how unmanagable that'll be in the long run and while a smart opportunist could capitalise on the mess it's not something really controllable.
 
Anduril’s New Drone Killer Is Locked on to AI-Powered Warfare
Wired. Dec 1, 2023 https://archive.is/OGw0Y
Autonomous drones are rapidly changing combat. Anduril’s new one aims to gain an edge with jet power and AI.
Samuel Bendett, an expert on the military use of drones at the Center for New American Security, a think tank, says Roadrunner could be used in Ukraine to intercept Iranian-made Shahed drones, which have become an effective way for Russian forces to target stationary Ukrainian targets.
Bendett says both Russian and Ukrainian forces are now using drones in a complete “kill chain,” with disposable consumer drones being used for target acquisition and then either short- or long-range suicide drones being used to attack. “There is a lot of experimentation taking place in Ukraine, on both sides,” Bendett says. “And I’m assuming that a lot of US [military] innovations are going to be built with Ukraine in mind.”
 
the latest AI programs for video are incredible - i wouldn't have noticed the CGI and I'm allergic to it. Whats publicly available has already moved on so far since this time last year.

sounds like the biggest parties in India are shamelessly going in with deepfake videos - of course Modi well beyond the ability to feel shame :
 
AI commentators are in a frenzy the last couple of weeks, all the latest signs pointing to ever faster progress and exponential developments beyond expectations.... it all sounds very convincing, totally revolutionary and ever closer in terms of time scale...

I get why people might not be that impressed with what they've seen so far but whats coming close down the line looks likely to be on a different scale... I for one welcome our new AGI overlords, good luck to them
 
AI commentators are in a frenzy the last couple of weeks, all the latest signs pointing to ever faster progress and exponential developments beyond expectations.... it all sounds very convincing, totally revolutionary and ever closer in terms of time scale...

I get why people might not be that impressed with what they've seen so far but whats coming close down the line looks likely to be on a different scale... I for one welcome our new AGI overlords, good luck to them
Yeh you won't welcome them so much when they unleash diseases to clear biological life off the planet to make space for their own plans
 
Yeh you won't welcome them so much when they unleash diseases to clear biological life off the planet to make space for their own plans
There are other planetary issues involved more basic than that - the sheer amount of energy required to ramp up AI, which is already having an impact. Quite a number of governments are now effectively throttling the development of new data centres because of their energy requirements.

I suspect the logical endpoint of this is that we will have revolutionary advances in AI - but it will become expensive and very much a tool for the rich.
 
Yeh you won't welcome them so much when they unleash diseases to clear biological life off the planet to make space for their own plans
tbh im feeling so fatalistic about our collective fate even before AI came along (environmental collapse, military-capitalism total dominance) that a massive technological revolution might be about the only thing that can trigger the deep change needed in time. All thats solid and all that
 
The two key areas of disruption seemingly will be in innovation/science/discovery/medicine and labour/work

On the first there could be breakthroughs not only that currently solve lots of problems and improve our lives but also skewer some of the key motors of capitalism: scarcity and obsolescence

With work the potential for automating so much of labour, whether essential or pure capitalist bullshit job, will likely create a deep crisis for the current order. Even nonsocialists like Sam Altman are publicly grappling with and encouraging the idea of redistributive deep taxation to deal with the coming changes

What seems unavoidable is that a massive disruption is coming and potentially on such a scale across the class spectrum that it seems reasonable to presume it will trigger a major systemic upheaval - I dont think it going to be like post 2008 Austerity, just more inequality, I think it could be much more profound than that
 
I see it as the other way round. Film production is already the preserve of the wealthy. AI opens up access to creativity for the lower orders.
yes there is that too....AGI created technological breakthroughs in areas like renewable green energy, water desalination, god knows what else, could have massively empowering impacts for normal people. That's what I'm saying about potentially breaking away from "scarcity and obsolescence", things that are currently the preserves of the elite could become mass-accessible - be that the ability to make a film or to access clean water
 
yes there is that too....AGI created technological breakthroughs in areas like renewable green energy, water desalination, god knows what else, could have massively empowering impacts for normal people. That's what I'm saying about potentially breaking away from "scarcity and obsolescence", things that are currently the preserves of the elite could become mass-accessible - be that the ability to make a film or to access clean water
I think given the environmental impacts of AI, and the fact we are already exceeding environmental boundaries on some metrics, this is sadly just techno-optimism.
 
I think given the environmental impacts of AI, and the fact we are already exceeding environmental boundaries on some metrics, this is sadly just techno-optimism.
i think as with all technological revolutions there will be positives and negatives... but there will be positives for sure, thats not optimism, thats just fact. We dont know yet exactly what these positives will be, so not worth over speculating on their precise nature, but examples that already exist in medicine and also in the creation of new materials give us an idea of the potential
 
You can thank the collapse of cryptocurrency for the rise in AI. All those mining farms now being put to better use.
This might be true but that's temporarily soaking up some existing capacity, not generating the huge amount of extra energy necessary to scale up AI so it can create movies tailored to your precise need. There are real material limits on this that have to be taken into account. Hence, I don't think it's going to be a creative tool for the lower orders at all - maybe to begin with for the early adopters, but not in the medium term.
 
This might be true but that's temporarily soaking up some existing capacity, not generating the huge amount of extra energy necessary to scale up AI so it can create movies tailored to your precise need. There are real material limits on this that have to be taken into account. Hence, I don't think it's going to be a creative tool for the lower orders at all - maybe to begin with for the early adopters, but not in the medium term.
I’m not entirely convinced that AI will completely replace current processes in major production studios, if at all. One of the challenges it faces is with continuity, which presently is next to hopeless along with all of its other idiosycracies.
It’s better placed in more random usages such as music videos, or advertising, rather than large scale productions.
 
Last edited:
Im less interested in us getting to have a go on the biggest AI engine., and more about what the mothership AI engine will create and solve. Once that information is out there its that which will really revolutionise everything.
Im sure there'll be plenty of consumer grade tech toys too, many of which will be good enough to make certain jobs redundant, but its the deeper structural innovations that I think will have the biggest impact
 
Excellent read on the AI hype with emphasis on software engineering automation: Losing the imitation game
I dont dismiss these kinds of analyses, in trying to understand what is happening it is definitely important not to get sucked into hype, but they seem to be talking about AI last year, not 2030. Its the projected trends that seem revolutionary to me...we shall see...
 
I see it as the other way round. Film production is already the preserve of the wealthy. AI opens up access to creativity for the lower orders.

Few points on that...

1: GAI is massively exploitative of workers. Pretty much every publicised instance of it is built on stolen content and beyond that is reliant on low paid workers, often in the Global South, facing shitty conditions to categorise data. Something like Amazon's MTurk for example isn't just exploitative, it's a whole new model of dehumanisation for labour which isolates and eradicates agency for workers both through algorithmic management and the removal of basic human identity within the production process. That's a model that's ubiquitously replicated in building training models. And that's before you get to the exploitation and destruction in the material supply chains to build the infrastructure necessary for AI to function.

2: Sooner or later we will see a surge in copyright claims against GAI models for their blatant theft, those claims will come from corporate media platforms and a minority of wealthier artists, not from your average working creator. Already companies are both expanding their training base and getting ahead of those claims by striking details with platform operators and institutions which completely bypass any notion of consent (or even awareness) to inclusion in these datasets. Reddit is the most recent one to sell off its user's work en masse but within the last few weeks universities have been signing up for it too and there was a whole SAG strike about the potential for Hollywood to sell of artists work without consent or recompense.

3: GAI isn't emerging as a free and open system of tools, it's being built by companies (OpenAI, Meta, Microsoft (through OpenAI) & Google) whose sole purpose is profit. While there's value in offering free iterations to generate investment and popular interest they're still going to monetise those tools ultimately, especially when it comes to robust implementations for commercial use. The cost barriers as well as the access to tech and (to a lesser degree) training will all be barriers to the 'lower orders' getting involved.

4: Even where it does allow access on a fairly open level both the companies supporting GAI and the corporate environment they exist in are built on platform capitalism. Any liberatory quality to these tools (which I'm cynical about anyway) will be usurped by the realities of distribution and the attention economy. Anyone(ish) can make music on their laptop but the Creative Commons/free/indy music models were all still gutted by Spotify and YouTube. The culture of corporate centralisation is as present with GAI as it is with everything else (OpenAI has deals with numerous tech companies, hence stuff like Google Copilot).

5: AI in general, including GAI, is absolutely atrocious at dealing with inclusivity and bias. It's built on oblique training data which is held as corporate secrets and uses algorithms which, even where visible, are usually incomprehensible to the layman. The recent stuff over Gemini is a case in point on that one, it was railroaded into a model of corporate diversity which lead to some pretty absurd results. It was in no way representative, the industry generally isn't, but it did offer plenty of fodder to racists insisting that AI is 'woke'. And even if that was a fairly daft little scandal (of sorts) it still reflects how ill considered GAI generally is.

6: It's wish fulfillment above all else. GAI doesn't enable 'creativity', it enables exploitative requests. It has no creative process or craft, just a demand for immediate gratification fed by underpaid, invisible, largely unpaid labour. It's a glorified version of typing 'big tits' into Google.

7: GAI actively undermines real human creativity. One of the big claims of GAI evangelists is always 'now anyone can be an artist!', which basically translates as 'now everyone can create content which we feel matches up with a commercial defintion of 'proper', which is to say profitable'. GAIs with built in house styles and processes built on theft inevitably narrow the scope of potential by having no space for the originality that emerges in learning about art and crafts as well as demeaning and denying the more universal forms of creativity that genuinely are liberatory. If you're just starting to paint for example you might experiment and discover some new and personal way to do it or you might just look at the GAI and go 'well, shit, I can be a master tomorrow, why bother being bad?'

yes there is that too....AGI created technological breakthroughs in areas like renewable green energy, water desalination, god knows what else, could have massively empowering impacts for normal people. That's what I'm saying about potentially breaking away from "scarcity and obsolescence", things that are currently the preserves of the elite could become mass-accessible - be that the ability to make a film or to access clean water

AGI doesn't exist and there's no sign at all that it will any time soon. Even with that said though - abundance doesn't negate scarcity. We don't have as much poverty/hunger/homelessness as we do because we're incapable of increasing the supply of and more fairly distributing resources/food/homes, we have it because there's less profit in doing so. AI models built by some of the most predatory companies on earth and shilled by some of the most delusional crackpots (Thiel, Altman, Musk, Bezos etc.) on it isn't going to go in any utopian direction.
 
Back
Top Bottom