OpenAI’s billionaire CEO has done it again with another hot take immediately setting off the furious online hordes.
Death, taxes, and billionaire tech CEOs setting people off with their hot takes on weird podcasts.
Billionaire OpenAI chief Sam Altman has gone and done it again this week.
This time, he’s been accused of coveting the delicious hardware sitting inside all of our devices.
Chances are you’re reading this on a smartphone or a laptop. Inside there is memory, or RAM.
It’s an essential pillar sitting under the hood that makes everything we do on a screen possible, and the price of it is exploding as data centres expand across the planet.
You might have paid for your laptop, but Mr Altman thinks it’s a tragedy that it’s being under-utilised in the AI race he’s helped kick off.
“From a global allocation perspective, actually, it’s quite terrible to have everybody have a bunch of the world’s limited supply of DRAM sitting in their own computer and being used 5 per cent of the time,” he told John Collison on the Cheeky Pint podcast this week.
For your average bear, it’s quite a bizarre take. Why on earth should he care about your old laptop sitting in the garage?
It all comes back to the philosophy that is driving the feverish obsession with AI.
Inside the minds of many of today’s tech innovators is an obsession with optimisation, where everything useful can — and should — be translated down to data that can be digitised, measured and then directed and capitalised on.
Much like an F1 team gunning for the championship, today’s RAM-thirsty tech companies are rushing to squeeze every possible atom at their disposal.
And they always want a little more.
The fact that there’s millions of lazy RAM sticks sitting in our spare rooms instead of working flat-out in OpenAI’s data centres appears to be a heartbreaking scenario for Mr Altman, and just anyone else who wants their enterprise to get an inch ahead in today’s searing sprint towards superintelligence.
The nerds are fuming
The backlash over that one line has been absolutely ferocious.
Windows Central executive editor Jez Corden used the word “sociopath” in a scathing headline about Mr Altman’s RAM take, prefaced with “You don’t hate him enough.”
His underlying accusation was that the industry is steering people away from owning computing power and towards renting it from the companies driving the shortage.
Agency is the word of the year among AI commentators, and the rapid integration of cutting-edge models into daily life is creating a mountain of questions and concerns over our future.
Some say there is value in keeping some things entirely up to a human, while others would rather let rapidly improving technology do everything for them as soon as it is capable.
And yes, that really does mean everything.
In the case of hardware like RAM, the opinion among many techies is that the more you personally own and control, the better. Democracy.
But in the case of consumer AI tech, the current market model is to rent out computing power from extremely powerful, rapidly-growing companies.
The case made by the optimists, who coincidentally own and operate the industry, is that enough investment will eventually lead to a world of abundant utopia. Enough investment, in this case, means giving AI companies more of the market’s scarce RAM.
Gizmodo’s Tom McKay made a similar argument to Corden, but was a bit more measured.
“Altman isn’t describing the optimal distribution of resources so much as the distribution of resources optimal to him,” he wrote.
McKay accused the OpenAI boss of dressing up a commercially convenient arrangement as a no-brainer.
More memory for data centres means more capability, and then more dependence on cloud services, and then ultimately more of our daily lives passing through an AI company’s machinery.
“Closed-source data centres are OpenAI’s bread and butter, as is platformisation and consolidation,” he wrote.
While some of the more vulgar criticism involved aggressive attacks against Mr Altman’s character, the core of the more level-headed arguments is anti-monopoly.
Suspicions over tech monopolies existed long before today’s RAM debate.
For years, writers, artists, musicians and publishers have accused AI developers of helping themselves to the raw materials of their professions, then selling tools that compete with the people who supplied them.
To those critics, being told that their computer hardware is also inefficiently allocated feels like another attempt at scooping up resources, and another instance where a seemingly throwaway line might have shown a glimpse of the industry’s true colours.
The missing bit of context
To be fair to Mr Altman, he wasn’t exactly proposing a raid on your laptop.
His comments to Mr Collison, who is the co-founder of Stripe and a close commercial partner to OpenAI, concerned something called “local inference”, which is essentially running an AI model on your own machine rather than sending requests to a company’s servers.
His argument was that scarce memory could serve more people if it were pooled in heavily used data centres, instead of sitting in individual computers that run AI occasionally.
“You want maximum utilisation in the cloud if you’re trying to deal with this very supply-constrained environment. And somehow many people’s intuition is the opposite of that,” he said.
It appears “somehow” there are still people sceptical of monopolies.
Mr Altman’s take is a very theoretical, optimisation-first argument, which is a trap that many tech CEOs have fallen into in recent years.
By now, they should know that personable nuance is extremely important when attempting to sell something they all say will rattle civilisation to its core.
For the tech critics, the proposal is just hypocritical.
“It still feels a bit rich for Altman to complain about the memory scarcity OpenAI and ChatGPT helped create,” PC Gamer hardware writer Jess Kinghorn wrote.
Tom’s Guide’s Jason England then hit us with the layman’s breakdown.
“It’s like the CEO of Chili’s telling you that supermarkets and cooking at home is a terrible allocation of food,” he wrote.
The price of progress?
It’s getting very expensive to build your own computer now, at least compared to 10 years ago.
In some cases, five times more expensive.
In an October 20 report, PC Gamer documented a 32GB Crucial Pro DDR5-6000 kit that had typically cost $US80-100 ($114-142) during much of 2024 and 2025, but was now listed at $US553 ($787).
The broader market tells its own uncomfortable story. According to TrendForce’s June assessment, conventional DRAM contract prices rose approximately 93-98 per cent in the first quarter of 2026 compared with the previous quarter.
It’s important to note that AI is not the only influence on surging memory prices. Production cuts, product retirements and the industry’s familiar boom-and-bust cycle have played a role as well.
Nor can you simply pull a component from a gaming PC and slot it into any AI server.
The competition nevertheless reaches back into the same manufacturers’ factories, investment budgets and production priorities.
TrendForce says suppliers have been favouring higher-margin server applications, constraining availability for PC and smartphone makers.
One company has been unusually clear about where that leaves ordinary customers.
In December 2025, memory giant Micron announced it would exit its crucial consumer business, with shipments continuing only until the end of February 2026.
“The AI-driven growth in the data centre has led to a surge in demand for memory and storage,” chief business officer Sumit Sadana said.
What’s the next step for AI?
Mr Altman has already explained the business he sees emerging from increasingly competent AI.
Speaking at BlackRock’s infrastructure summit in March, he described a future where “intelligence is a utility like electricity or water and people buy it from us on a meter”.
Mr Altman’s stated goal is abundant, cheap access. But his most vocal critics assume extremely powerful intelligence will remain a scarce resource reserved for the rich, and potentially used to questionable ends.
His proposed answer is to build more supply.
“The cost of energy should go down. The cost of chips should go down. The rest of the cost of the supply chain should go down. If it doesn’t go down, I’ll be very sad,” he said.
“But for it to go down, given how much value there is in this, supply’s got to go way up.”
He’s been here before
The RAM row belongs to a much broader argument about the way Mr Altman describes people’s place in his transition towards a vague utopia.
He’s come under intense scrutiny several times in the past. One quip about what constitutes “real work” lit a particularly big fire under critics.
In February, another comparison landed badly.
While defending the eye-watering amount of resources required to develop and run AI, Mr Altman switched the argument around, commenting on the investment needed to develop “human intelligence”.
“It takes like 20 years of life and all of the food you eat during that time before you get smart,” he said.
In attempting to compare the costs, Mr Altman gave everyone a glimpse at the optimisation philosophy that drives the entire industry.
Some listeners understandably objected to treating human development as though it belonged in the same accounting exercise as a literal machine.
Zoho co-founder Sridhar Vembu responded publicly, summing up millions of people’s thoughts in a very apt one-liner.
“I do not want to see a world where we equate a piece of technology to a human being,” he said.