canjoena - stock.adobe.com
Tech bros beware - Erin Brockovich is coming for you
The campaigning heroine of the eponymous movie has AI datacentres in her sights - just as Big Tech spending on memory chips sends PC and mobile prices spiralling up
Not content with destroying our democracies, our planet and our livelihoods, the AI broligopoly is coming directly for your technology. We are encountering what’s being coined as RAMageddon - an era where there is such a shortage of memory chips that curious decisions are being taken by Big Tech, which at the very least are counterintuitive or, to put it another way, stupid.
According to Big Tech, they promise that the AI they put in our devices will achieve a number of transformative things, such as speed and responsiveness; privacy and data security; and offline capability.
Rather than routing through the cloud we will be able to do tasks directly on our devices, promising a new world of efficiency.
And fancy being able to access everything on your device when on a flight? We are told that this will all be possible because of AI. But this is totally untrue, certainly for the foreseeable future. These companies are so voraciously buying RAM to feed their AI training models and datacentres there will be precious little RAM available for our domestic devices. And what we will be given in the future will not be more RAM but less as the broligarchy keep all the memory for their datacentres and AI large language model training.
This time next year - or perhaps even later this year - you will be paying far more for a laptop and phone that may have less, not more, capabilities.
Staggering investment
The scale of investment by four companies to acquire this RAM and other AI training costs is staggering. According to a report from Bloomberg, Alphabet, Amazon, Meta and Microsoft have budgeted $650bn capital expenditure for 2026.
To put that in perspective, according to Bloomberg “21 other major corporations spanning industries from automating to defence contracting found their combined budgets total just $180bn. The AI infrastructure spend of four Silicon Valley giants dwarfs the capital plans of nearly every other industry on Earth, combined.”
If AI is going to change our lives and make them easier, why are the very companies building the technology asking us to pay their capex costs for higher prices with less memory and functionality?
According to this Tim Green article, the irony is hard to miss: “The technology industry has spent the past two years marketing ‘AI smartphones’ with enhanced on-device AI capabilities, features that typically require more RAM, not less. Now the very infrastructure being built to power the AI models behind those features is cannibalising the memory supply those phones need to run them.”
In its most recent market trends analysis, Global memory shortage crisis: Market analysis and the potential impact on the smartphone and PC markets in 2026, analyst IDC set out a stark picture, particularly for the fate of formerly lower-price electronics:
“Manufacturers, whose business is mainly in the low end of the market, are likely to suffer significantly. The business models of vendors such as TCL, Transsion, Realme, Xiaomi, Lenovo, Oppo, Vivo, Honor or Huawei are based on thin margins. This increase in cost will hit their margins substantially, and they will have no other option but to pass the cost (or part) to end-users,” said IDC.
“In the high end of the market, Apple and Samsung face pressure but are structurally hedged. Cash reserves and long-term supply agreements allow it to secure memory supply 12-24 months in advance. On the other hand, new flagship models in 2026 will likely have no RAM upgrades, sticking to 12GB for Pro models rather than increasing to 16GB. It is also unlikely that current models will see the same price erosion seen after the introduction of the latest model.”
Ranjit Atwal, senior director analyst at Gartner, paints the picture for the future of low-priced PCs in chilling terms: “Ultimately, we expect the sub-$500 entry-level PC segment will disappear by 2028.”
In real terms, analyst Trend Force estimates prices for mainstream laptops typically costing about £667 might increase by as much as 40% in 2026 because of the memory chip shortage and rising costs of other parts, such as motherboards.
Vulnerable at risk
Yet again the actions of the Broligarchy and their voracious greed will put those most vulnerable in our society at risk. According to the CEO of Digital Policy Alliance, Elizabeth Anderson: “New research from RM Technology and the Digital Poverty Alliance highlights that more than half (57%) of low-income households lack reliable access to devices or the internet at home. The consequences of this are far reaching, as children struggle to complete homework, access learning resources, and keep pace with peers. Teachers warn the gap is widening, and without urgent action, digital poverty will continue to hold back the country’s most vulnerable students.”
Further egregious behaviour by Big Tech with an impact on your pocket is this story from security researcher Alexander Hanff, aka That Privacy Guy, outlining how Google Chrome silently installs a 4GB AI model on your device - without asking consent, which he argues is in breach of several parts of GDPR.
Downloading 4GB is not trivial for people on metered connections, mobile hotspots or those with bandwidth caps or in developing countries where data costs remain prohibitive. This kind of unflagged download could very quickly use up your data - as was confirmed by a commenter on a Pieter Arntz article for Malware Labs on the same topic: “I happen to be in a rural area myself and have 150GB per month limit on my connection which I use 70% of every month without even trying”.
Not only will Chrome users have paid the download costs silently and unknowingly installed the file (weights.bin) but the file permanently consumes about 4GB of your device’s local drive storage. A further push to offload RAM costs on to the consumer.
And you can’t just delete the file, since Chrome re-downloads a fresh copy during browser updates. It’s also misleading in that Chrome suggests that once the file is installed AI runs locally on your machine, but if you use certain cloud-based features it will still transfer standard token and prompt data to Google servers.
Hanff concludes with clear questions about the role of regulators and prosecutors: “The fact that the bytes are AI bytes does not exempt them from the law that governs every other byte that gets written to a user's device without permission. The fact that the bytes are ‘small’ relative to the user's disk does not exempt the cumulative carbon footprint from being a real, measurable, ongoing harm to the climate.”
If Google's next Chrome update silently removes the unconsented installs and replaces the behaviour with an explicit opt-in, we will know the company can read the room. If it does not, we will know what the company's published positions on responsible AI and sustainability are actually worth.
People power
In light of what is increasingly becoming default behaviour, one has to ask a very simple question: when will regulators and public prosecutors start to enforce the law which has been in place since 2002 - or are global tech corporations exempt from criminal and civil statutes?”
Don’t hold your breath. Yet again it seems clear the cavalry is not coming to help, but there are encouraging green shoots of people power beginning to bloom in the US.
Most notable is the re-emergence of none other than Erin Brockovich, whom readers may remember was played by Julia Roberts in the movie about her work. She was the successful instigator of the historic legal case against the Pacific Gas and Electric Company for contaminating groundwater in Hinkley, California.
While she has always worked on environmental issues, she has now created the ingenious Brockovich AI Data Center Reporting initiative. Launched last month, it includes a brilliant interactive map which is a call to action for concerned communities.
The map details publicly announced, major AI-focused and hyperscale datacentres running AI workloads across the US. The red dots on the map (and there are many) show “community reported” sites with data submitted by concerned residents across the US about AI datacentres. It's precisely the kind of visibility that has been missing and allows US citizens to get a granular look at the huge scale of planned datacentres. It also cites locations where datacentres are operational, under construction or proposed.
This kind of open data makes it a lot easier for communities to see what is happening and acts as a repository that serves as a backbone for resistance. If I were the Tech Bros I’d be feeling nervous. Might be worth having a look at the movie - Erin Brockovich is not a woman to be messing with. She has taken on corporate America before and won. I for one will be cheering her efforts for real transparency on the massive environmental monster the tech industry is creating, all in the pursuit of greed and market dominance. And in the end for what?
As more time passes, more companies executing AI are beginning to realise that the costs of AI far outweigh its usefulness. Take Uber, for example. As reported by Crypto Briefing, “After deploying Claude Code to 5,000 engineers, Uber reportedly burned through its entire 2026 AI budget of $3.4bn in just four months. Monthly usage rates among Uber’s engineers hit 84-95% by April 2026. Per-engineer API costs ranged between $500 and $2,000 monthly.”
There is a reckoning coming in all this for sure, and as always those who will feel the most pain are those of us who did not create this madness but will surely suffer the economic consequences - from poorer performing tech for higher cost, right down to our pensions.
