While painstakingly reviewing the dystopian near-future novella that destabilized markets early this week, I wrote, of Citrini Research’s “Global Intelligence Crisis,” that in addition to being compelling, the piece is quite lengthy.
In the era of compressed attention spans, it’s surprising that market participants indulged the report. “It’s fantastic,” I said, lauding the effort. “I wish I’d written it but even if I had, no one save my most dedicated fans would’ve read it [because] it’s also fantastically long.”
A couple of hours later, it dawned on me that in fact, most investors probably didn’t read it. Because who needs to read anything anymore? If someone pens a laboriously long article outlining an AI-driven white collar jobs apocalypse and you need to know what it says but can’t conjure the mental fortitude to read more than 200 words in a sitting, you could just ask AI to summarize it.
Sure enough, SocGen’s Manish Kabra sent around the following summary, courtesy of ChatGPT.

Citrini’s report “is comprehensive on the left-tail scenarios [but] takes about 40 minutes to read,” Kabra wrote. “Copilot has summarized the macro impact in 10 points.”
This is supremely ironic in the context of my observation (from the above-linked article) that although ChatGPT didn’t write the Citrini report, it certainly could’ve. If it did, Kabra’s bullet points would be ChatGPT summarizing its own handy work.
“The long story short for macro investors is a white collar recession,” Kabra said, adding that although this isn’t SocGen’s house view, “if software was disinflationary and AI eats up software,” the result “could be more than simply disinflation.”
In other words: We could be staring at outright deflation, in which case the “simple AI trade [is] Long US bonds.”


Ride the equity price increases while margins gain (Step 4) and then jump off into long bonds at Step 6?
There you go. It’s so easy!
Of course, timing is everything! Fortunately, some of us (maybe most here) are no longer white collar “workers”, at least only as avocation rather than need – so, the question is: If we already have enough to be comfortable, who needs the risks from equity or bond markets — but, then what? No such thing as a riskless asset class, is there? Cash? Gold? Property (I suppose that one works – if to live and farm on)? Petrol (Mad Max scenario)? The Spice (ask Frank Herbert)???
Pondering.
“Long US Bonds” ? Long term bonds, or, Long all US bonds?
Ah, Some Lazy Bum answered my question!
I pay for only two market focused subscription services: Heisenberg Report and Citrini. Never did I anticipate the two intersecting like this!
Is Citrini worth paying for? Actual question.
The future is notoriously hard to predict.
I thought Copilot was the Microsoft AI.
Why doesn’t somebody ask ChatGPT the timing of these events?
I have a prediction about SaaS that will probably take some time to prove out but will likely come to fruition. Those who are building their startups with AI as their primary employee(s) are going to disrupt entrenched and bloated businesses. While those businesses are just now trying to figure out how much work they can offload to AI, the startup is AI first, they are building a leaner operating model because it behooves them to keep their OpEx and CapEx as close to zero as possible (thanks to higher rates). If a startup can produce a viable product that undercuts a competitor’s price significantly enough (again owing to ultra low OpEx) to incentivize switching the snowball begins rolling down the mountain. As the product proves its capability and can showcase some premium clients more firms will be incentivized to switch.
At this point, it is just a matter of time before the foundation models outperform all white collar workers at every job, but only if you know how to properly deploy, measure, and adapt them. The founders building with AI know how to do this better than everyone else who’s looking for a partner to help them get there.
I saw a few days ago that Uber is buying the parking app, Spot Hero. Spot Hero is a great resource for finding reasonably priced parking spots for short to medium terms. Everything is done online. Very few employees needed.
I have used Spot Hero on several occasions- the only thing that I do not like, is that if you find a monthly parking spots, they force you into renting by calendar month. A previous generation of the Spot Hero software would allow you to start on any day and end 30 days later. So not as economical for me, as it used to be.
How has no one referenced Karl Marx’s work on the tendency of the rate of profit to fall.
Isn’t a lo of that report just a repackaged Marxist argument?
Pretty much everything anyone says about macro is repackaged Marx. We just don’t realize it.