Wednesday, July 1, 2026

Corey Doctorow, the reverse centaur, and AI

I’ve been reading Corey Doctorow’s The Reverse Centaur’s Guide to Life After AI. It’s an easy, breezy (and quite short) read. Doctorow packs a good deal of insight into the book.

I knew that being a growth company was more desirable than being a mature company. Your stock is valued more, giving you the ability to hire top talent and even gobble up other companies. 


What I did not realize is just what lengths corporations would go to in order to convince investors that they were still growth companies.


The Metaverse never made sense to me; Doctorow theorizes that it was never meant to work: it’s only purpose was to convince investors that Meta had a whole new area to grow into, with potential billions of profits. Monopoly money as it turns out. 


But it convinced gullible investors and stock market pundits that Meta was still growing. 


And voila: The Metaverse finally makes sense, something I did not think was possible. 


Google+ was meant to show that Google, which had already cornered over 90% of the search market, had whole new social media arenas to grow into (and then eat Meta’s lunch).


Now, AI is fulfilling this role. 


Doctorow’s theories also explain why the AI companies went after artists so hard: they needed a demo to impress investors with. Spreadsheets wouldn’t do. Slick videos and illustrations, on the other hand, could impress investors. 


So… by releasing the AI image generators to the public, they get free labour (AI prompting and art direction) by millions of people, and while much of the result is slop, there’s gems in there too. Like needles in a haystack. 


And the impressive generations impress investors.


The public is incidental to it all. Providing the ‘service’ is incidental. What matters is impressing venture capitalists and investors. 


The perverse part is that replacing low wage artists with AI generated images will never be profitable. Which suggests how underpaid artists are; after all, replacing work of real artists with facsimiles is impressive enough to move trillions of dollars in investment. If the facsimiles are enough to generate investment, what does that say about the human artists? 


The whole thing is a con.


It’s all about generating investment from the top 1% at the expense of the arts, the public, our jobs.


AI companies cannot make back the billions invested in them without replacing workers. That’s their ultimate end game. It's the only one that makes the slightest financial sense.


All the stuff about curing cancer is propaganda designed to create ‘vocational awe’ and convince AI programmers that they are bringing about a new age of wonder. 


The opposite is more likely. 


Talk about AI taking all jobs, or destroying humanity (unlikely but possible I suppose, just not with LLMs), is also marketing: it’s intimidation, and implicitly threatening: AI tech is so big, so transformative you must invest in us or get left behind in the dust. It’s designed to create fear in the target audience: investors.


The more the public buys into these narratives, the more true they seem, and the more validated venture capitalists feel about investing in AI.


It's all so cynically dystopian, you'd think it was written by William Gibson. 


And if we think things are bad now, just wait till the bubble bursts. 


You can get Doctorow’s book on Amazon (heh). 

Thursday, May 21, 2026

AI update: It’s here. Or... is it?

I did another dive into AI systems recently, mostly instigated by job requirements. I pumped out some new imagery, and I have to say it’s greatly improved from the last time I took a dive. 

People keep saying that the recent stream of mass layoffs is only using AI as an excuse, that no one is actually losing their jobs to AI. The real reason is overhiring during COVID.


Well, it’s 2026 now, and that reasoning just isn’t working any more. 


Companies are cutting staff and forcing remaining workers to use AI to pick up the slack. Entry level hiring in areas affected by AI is down 12%. How is that not automation bias? How is that not related to AI?


By 2030, $7 trillion will have been invested in AI and their multiplying data centres. 


The only way CEOs can make that massive investment pay off is with massive impact. AI company CEOs say flat out that their programs are designed to replace workers. Facebook was tracking employee keystrokes to get a better idea of how humans work, so they could feed it into the AI Maw.


How many salaries would $7 trillion cover? 


AI chips wear out faster than human brains, and consume far more electricity and water. A human can work for 50+ years without needing to be replaced, but an AI chip needs to be replaced every 3-5 years. 


Right now, AI companies are selling the services of their creations under cost. They’re burning billions of dollars. Why? It’s an old tactic: enter a market, sell under cost of production until all competition goes bankrupt, then jack the prices up sky high. Consumers have no choice but to buy your product: there's nothing else.


Who’s being undermined here? 


Humans. 


Put enough humans out of work, make corporations dependent on AI systems, and voila, you have a new dystopian future. 


And I thought I was cynical!


AI started out as a fun toy. It got better, but still flawed enough to require human intervention. 


Now? 


Better than ever. Still flawed, but I can see it replacing workers. Or at least, doing well enough to convince management that they can fire people and replace them with AI (even if, long term, it's going to be a disaster).


And it is.


We’re told AI is inevitable. Get on board or get left behind. 


And you know what? 


That kind of fearmongering works… until you realize it's just a vulgar, obnoxious form of marketing mixed with intimidation. 


Interesting times, indeed.