Commons image
By Dennis Broe
While just three months ago the story in AI was one of overspending and a potential bubble that would soon burst. The story these days is simply about the spending now outpacing the bubble, as the Big 6 AI companies make a case that they, like the banks and investment houses in the 2008 financial bubble are, “too big to fail.”
There is no actual sign that the bubble has been overcome. Instead, everywhere the amount of spending is simply overwhelming the market. For example Amazon, one of the big 6, and one of the richest companies in the world, this year will spend more on AI development than its total earnings.
Erased in this flood of market capitulation is the fact that three of the big 6, Amazon, Google and Microsoft, lost a combined $900 billion in market value in one recent evaluation. That near collapse is being erased by this year’s 60 percent rise in spending from last year’s $410 billion, the idea being to simply overwhelm the market with the sheer might and power of the explosion and thus convince all skeptics that the AI revolution is inevitable.
All this is aided by the fact that it is AI investment—including the soon-to-be public offerings of Meta, Elon Musk’s Space X, Anthropic and Open AI—which is keeping the U.S. economy out of recession.
While AI hallucinations continue to abound with three firms recently having to withdraw AI generated reports because they were based on false data generated by the machines, the magnitude of these investments are having an outsized impact on companies at large with “scale, scale, scale” now replacing the old real estate adage “location, location, location.” Because of the scale needed to power these industries there is a nearly unprecedented wave of mergers, hastening increasing monopolization in the energy sector.
These mergers are becoming the occasion for an argument for deregulating utilities to keep up with this push, just as the AI companies themselves use the technology to outflank regulation in the tech sector, all hastened by Trump’s attack on regulation in any area including the Environmental Protection Agency which would ordinarily have a stake in overseeing the costs to the natural world of this expansion.
Where is the money coming from for the Big 6 to make these investments? Part of it is in retrenching in other areas of the company, with Amazon, Meta, Microsoft/Open AI, and Oracle reducing dividends, borrowing and taking the AI lead in cutting jobs. One of the great fears of the “AI revolution” is the potential job loss and the AI behemoths are showing other companies and sectors just how ruthless and “efficient” this job loss can be.
Amazon, at the same time as its leader Jeff Bezos announced a 300 bn investment in AI, cut 300 employees from The Washington Post. Bezos is buying up small manufacturing companies on the hope that just as he is automating his Amazon warehouses, now called fulfillment centres but soon to be unemployment centres, he can, using AI, automate these companies and thus greatly increase their profitability.
Oracle’s Larry Ellison fired approximately 30,000 employees by email, netting the company between 8 to 10 billion and Meta cut 8,000 jobs this year in its reshaping of the company.
AI also may soon be coming for more medium-sized businesses, businesses the processing service is now making more efficient. A scary recent New York Times Magazine article used as an example a three-person bankruptcy agency which processed claims using AI much more efficiently, but had the secretary fearing for her job and her boss admitting that as yet AI can’t answer the phones, indicating that her fear was real as it soon may be able to. At the end of the article, the company owner admits that at this rate AI may soon also be coming for him and his company in its relentless pursuit of “efficiency.”
AI is ever more quickly encroaching on, or as Martin Scorsese now says “enhancing,” the entertainment industry. Sam Altman of OpenAI which owns ChatGPT and who two years ago shunned advertising being part of AI is now, with the increasing burden of expansion embracing it with the company currently expecting to generate billions in advertising and now joined by the already online advertising expertise of Amazon and Google.
In the creative and cultural industries, this highway to hell is paved not with good intentions but only with greed. The drain on advertising dollars, already concentrated on the digital world will affect the streamers who have swung into advertising as well as cable and network TV. Along with this drain, Amazon MGM studios, which recently unveiled three entirely AI animated kids’ series, is also reportedly on the lookout for a blockbuster project it can produce entirely through AI, thus eliminating production teams completely.
The Times gleefully “interviewed” the first AI actress, Tilly Norwood, a white, cute “girl next door” type, who swings her hair in a beguiling fashion, while in the article sidelining until nearly the last paragraph the Actors’ Guild objection. The Guild said the “process undermines the very ecosystem that makes storytelling possible. It insults the artistry of our performers, assaults our business and threatens the legacy our members’ work creates, in many cases built over generations.” This did not stop The Times from ogling this synthetic creation.
AI on the Battlefield
The Financial Times called the billions being poured into AI expansion the biggest peacetime investment project in history, neglecting to recall that, as the economist Michael Hudson reminds us, these are also war times, with the Iran war whose economic effects have spread throughout Asia, Africa and Latin America as well as occasioning gas prices in California in the U.S. to hit 7 to 8 dollars a gallon, hastening the initiation of what Hudson terms “World War Three.”
The wars in Ukraine, Iran and Lebanon are being fought from the air, with drones and reconnaissance powered by AI. The Ukraine war, where NATO and the West still claim to be only helping in an advisory capacity, could not have been waged without the use of Elon Musk’s Starlink satellites. The Pentagon uses Palantir’s Maven for wartime data analysis and to support the entire what is called “kill chain,” that is finding and hitting a target in seconds with OpenAI’s ChatGPT used to compile targets. Needless to say, hallucinating AI has deadly consequences, as the U.S. claims the hit on the girls’ school in Minab was caused by faulty AI identification.
Anthropic claims it is the good AI, refusing the Pentagon’s urging to use its security program Mythos for domestic surveillance, and opposing completely autonomous weapons, with that refusal boosting its market position. But just as Google’s now outdated motto was “Don’t do evil,” so too Anthoropic’s dissent only applies to domestic warfare. Its model Claude is intimately involved in real-time analysis in the Iran war.
In the category of not throwing stones if you live in glass houses, the Pentagon recently accused the leading Chinese companies, the electric car maker BYD and Alibaba’s Baidu, the leading Chinese search engine, of being connected to the Chinese military, also a convenient way of attempting to isolate the coming threat of BYD’s cheap electric cars which are now everywhere in the West except the U.S.
Quantum processing, which allows greater computer efficiency, is said to be the key to the autonomous weapons of the future. The FT’s recent article on the development of that phenomenon framed it in entirely bellicose terms and titled “How chips became the new nukes,” as peacetime uses of AI are now giving way to the wars necessary to sustain the U.S’s fading empire.
Environmental Wreckage
A fascinating review of tech centre architecture by the FT’s critic Edwin Heathcoat notes the difference between the sleek design of the tech companies’ headquarters, often called a “campus” which suggests the idea of a quiet New England wooded centre of learning and inquiry, and projects life “lived in the cloud,” with the ugliness of the data centre processing plants.
These plants, which consume so much electricity around them that residents find their electric bills doubling and quadrupling, also give off a loud hum, so that workers need to wear ear protectors. The heat from the massive cooling systems simply dissipates and is used for no social purpose, with the system itself generating carbon emissions and reversing the gains from solar panels and electric cars.
The ultimate example of this dangerous waste of energy is that Microsoft bought Three Mile Island, the site of the worst nuclear disaster in the U.S. and is reviving it to power its AI. Heathcoat compares AI infrastructure to that of previous building frenzies, railroads in the 19th century and fibre-optic cables in the 20th, which may continue to have a use. Here, there are simply warehouses, which if the bubble bursts will be empty locations, and even if it doesn’t will be left for future more powerful warehouses.
It is said by some that to oppose AI is to harken back to the Luddite movement in England which campaigned against the factory system; that AI is here and all we can really do is accommodate to it. But there is another way of conceiving this, where the problem is not AI but the economic, political and cultural system under which it is developed and developing.
AI in China, where the technological growth is planned and managed by the government as much as possible for the benefit of all, instead of to make maximum profit for a few tech lords, is quite different and people in China have generally a positive view of that development. AI is developed for much less money in an open system available to be used by all, in a boon to other Global South developers.
Development is not geared entirely to profit, that is to address the concerns of businesses, but is also focused on where it can most directly benefit the society, for example, in robotics being used in old age facilities. The government is key in managing the propensity toward automation and job loss, with each deployment having to explain how many jobs will be replaced and how those workers will be retained or retrained. There is a brake on mass unemployment, which under American capitalism is simply a problem that along with deregulation and the bubble is swept under the rug.
Finally, considering the aesthetics of AI, there is a general improvement in both expression and efficiency in tasks such as resume writing, pitching television series and book proposals, but this upgrading is countered by an equally predominant homogenization. All of these resumes, pitches and proposals start to sound the same, and take on a mechanical quality to the detriment of creativity. The industry’s term for the elimination of what French theorists call difference or the excess, is “democratization.” This is to say nothing of the schools, another area where AI is outracing teachers, who when plagued continuously with AI-written papers simply throw up their hands and surrender. As do we all to a technology which wants – nay needs – to overwhelm us, and hopes we remain asleep as it not only encroaches on but defines and creates our future.
This article is a preview of Culture & Barbarism 12 available on YouTube, Substack, and Spotify.
