Cognitive Warmup. You know they know the game is over, but egos take over. Marc Andreessen, co-founder of venture capital firm Andreessen Horowitz, thought it wise to mock (in a retweet) Pope Leo XIV, the head of the Catholic Church, after he called on AI makers to “develop a moral conscience” as a fundamental part of their work. Of course, since this is Andreessen, there’s an (un)healthy Sydney Sweeney obsession (although it seems he’s also the EU’s living rent-free in the mind of Thierry Breton). It’s nice to see Pope Leo XIV directly addressing the AI brothers and telling them straight out, which of course Silicon Valley people, mostly CEOs and VCs, often don’t like. Although Andreessen tried, it just symbolizes that the bubble is bursting with government guarantees on AI company loans. Things are getting worse rapidly. The disappointment is clearly visible. Has anyone tried to tell the AI brothers “you shouldn’t steal”?
algorithm
This week, we’ve talked about Microsoft’s intentions to build superintelligence that replaces humans at the top of the so-called food chain, Amazon being downright unhappy with Perplexity’s Comet agentic browser behaving like a human on its shopping platform, and talking about the rising costs of electricity due to AI.
superintelligent ai dreams
Microsoft AI, although it is too early to know for sure, suggests that their vision for superintelligence is to “ensure that humanity remains at the top of the food chain”. I wouldn’t give so much credence otherwise, but for the fact that Microsoft is leading AI. Mustafa Suleman This is what he says, while describing what he calls Humanistic Superintelligence, or HSI. Speaking about the newly formed MAI Superintelligence team, he says, “We want to explore and prioritize how the most advanced forms of AI can keep humanity in control as well as accelerate our path towards tackling our most pressing global challenges.” The two global challenges Suleiman talks about are medical superintelligence, and optimizing workflows to make them more energy efficient. An impressive vision, we will celebrate when even a fraction of it is materially achieved, without harming humans. For one, AI companies can start by first defining ‘consciousness’, which is key to finding a balance between the interference of humans and AI.
Electricity costs, and political victory
There is a growing perception that rising electricity prices in the US have determined the outcome of recent elections in New Jersey, Virginia and Georgia. A lot of the credit for rising utility bills goes to AI data centers as well as electric motoring. Ample data suggests as much. Prediction At the beginning of the year the International Energy Agency (IEA) indicated that US electricity demand would grow at an average annual rate of about 2% over the period 2025-2027. The latest census data in the country shows that one in three American households reported giving up necessities like food or medicine to pay household energy bills. There is also data that New Jersey has seen the largest increases, with retail rates increasing by as much as 20 percent this summer. At a time when AI bosses are only talking about investing trillions of dollars in data centres, the citizenry already struggling with rising costs of living is clearly getting fed up of billionaires preaching the benefits of AI. Now, political candidates who win elections will also be held equally accountable if they do not bring reforms.
Champion of AI, not impressed by agentic AI
Amazon has sued Perplexity AI over its new Comet, an AI browser that has autonomous browsing habits and an assistive device that can reportedly do multiple tasks. Amazon alleges that Comet can log into a user’s Amazon account and attempt to place or modify orders while disguising itself as normal human traffic. In legal terms, Amazon lawyers say Perplexity’s trespass involves code rather than lockpicks. Distraction would probably call this some kind of innovation. I don’t know, I strictly maintain a distance from AI browsers – there is little to rely on that way. Interestingly, Amazon has also developed some level of agentic AI on its shopping presence – there’s “Buy for Me” in some markets that lets users make purchases within its app, as well as the Rufus AI assistant (it also features prominently on the app in India) that can recommend items and manage carts.
Thinking
“I want to clarify my comments earlier today. OpenAI is not seeking a government backstop for our infrastructure commitments. I used the word “backstop” and it confused the issue. As the full clip of my answer shows, what I was trying to say was that US strength in technology will come from building real industrial capacity which requires the private sector and the government to do their part. As I said, the US government is incredibly forward-looking and has Really understood that Al is a national strategic asset.” – Sarah Fryer, CFO, OpenAI.
There’s something almost poetic about Silicon Valley’s newfound patriotism. The same industry that once prided itself on being ‘borderless’, and had the aura of being too busy inventing the future to deal with government paperwork or regulatory hassles, is now shaking down the gates of Washington with the subtlety of a brass band. Have a small request to them – please outline our trillion-dollar ambitions. Of course, one should not interfere. Just to catch the inevitable free-fall. They know it’s coming. They know it’s a bubble.
OpenAI’s CFO, no matter how hard she tries to backstop the bubble, wants the US government to guarantee the mountain of financing she (and her creative investment partners like Nvidia, Oracle, and AMD) will need to further inflate the bubble. Or in other words, to explicitly build its future GPU infrastructure empire. The argument we’re hearing loudly from people like OpenAI and Jensen Huang (who paused his ‘Let’s Make a Deal’ world tour to make the argument) goes something like this – AI is too strategically important to fail, the US must lead, and so taxpayers must be morally prepared to pick up the check if the math doesn’t work out. And while we’re here, could Treasury perhaps fill out the loan guarantee form in triplicate? The OpenAI and Nvidia tag-team is very visible.
Context: It appears that the AI bros are channeling a techno-age version of their inner Founding Father vibes. Suddenly, GPUs are the new railroads. AI clusters are national projects. Silicon is a symbol of sovereignty. Marketing copy writes itself. There’s also a little flag-waving flourish at the end – if we don’t build it, China will. You don’t want that, right? The specific fallacy Altman has tried to create for OpenAI is that they are too big to fail. If such a thing ever existed. And all the more so in this case, where the company has not yet started making profits throughout its existence, is creatively squandering investment money as if no one is paying attention, and has completely failed the latest models that were supposed to usher in the “PhD level intelligence” era. They still falter in basic mathematics. Some interesting stats – This AI company lost $13.5 billion in the first half of 2025 on revenues of $4.3 billion.
The AI industry is spending money so fast it would put WeWork’s accountants to shame. The capital expenditure required for model training, inference infrastructure and chip supply is not only high – it is almost vertical. The realization has come that private capital alone will not support a billion-dollar national AI infrastructure. The new plan, which was probably thought up in a coffee shop in San Francisco (that’s how most deals are done in that part of the world) is to pass off the risk as a patriotic duty. This takes me back to 2023 when Altman said at a summit, “It’s frustrating to compete with us”.
A reality check: When an industry truly believes in its future, it takes its own financial risks. In the case of AI, this is not the case at all, and hence there is a subtle movement to hand it over to the unsuspecting masses to soak up. A request for a backstop is not a declaration of confidence. The AI companies, whether it’s OpenAI, Nvidia, Oracle, AMD, Anthropic, Meta, and Microsoft, are essentially writing an admission that they have no idea what they’re doing, that everything they’re doing is expensive, and let’s see where it takes us, unless you split the bill. If this bubble persists, Silicon Valley will be hailed as the leader of a new technological era. If this does not happen, taxpayers will simply have to be left holding the bag. “We tried”, we will be told. They win heads, you pay tails.
I rarely agree with any politician (or AI bro, as things go), but I have to say it’s hard to disagree with Florida Governor Ron DeSantis when he said, “My solution is not to bail out big technology companies.”
