Post

Free newsletter: I go over NVIDIA's earnings, the reality of OpenAI's $100 billion round, and the "analyslop" phenomenon of speculative, lie-ridden fan fiction that gets framed as meaningful research and promoted by the mainstream media. www.wheresyoured.at/on-nvidia-an...

wheresyoured.atOn NVIDIA and AnalyslopEditor's note: a previous version of this newsletter went out with Matt Hughes' name on it, that's my editor who went over it for spelling errors and loaded it into the CMS. Sorry! Hey all! I’m going...

First, NVIDIA: on the surface things looked good, but obligations to suppliers like TSMC have nearly tripled from $30.8bn to $95.2bn, inventories are over $21 billion, and 36% ($77.7bn) of its FY2026 revenue came from two customers. Any capex cut is existential. www.wheresyoured.at/on-nvidia-an...

Which brings me onto the main event: Analyslop, a new annoying brand of speculative fiction that moves markets because the media and investors don't understand anything, the latest example being Citrini Research's lie-filled "global intelligence crisis." www.wheresyoured.at/on-nvidia-an...

It's both deeply embarrassing and a failure of the media and markets' critical thinking that the Citrini Memo was taken seriously. No, Citrini's memo is not "provocative" or "brilliant" - it's slop! It doesn't hinge on any technological or financial evidence! www.wheresyoured.at/on-nvidia-an...

The AI economy has become a grifter economy, with dodginess and chicanery at every level, where CEOs lie, and their lies are solidified into plausible-looking analyslop. And the institutions that should be challenging them are nowhere to be seen. www.wheresyoured.at/on-nvidia-an...

19 reposts112 likes

The problem with calling out AI grift is that the benchmarks are real, the demos work, and the products occasionally do something useful. Makes the chicanery harder to isolate than in, say, crypto.

I think we can't entirely blame this one on the media - there's a bunch of programming "influencers" on twitter/youtube who are propagating nonsense - you also see garbage in OECD "research" (which a lot of policy people sadly still take seriously): bsky.app/profile/arch...

Arvīds Kokins@archo5.bsky.social

also how did OECD manage to find near-constant and universal ~2% worldwide gen"AI" use pre-chatgpt? the data does not even make sense then I looked into the OECD paper and it seems I found the same chart then I opened the source to check if I can find the same data it's not there

all I'm saying is, there's a much larger structure of disinformation, where even seemingly sources with credentials are being taken over by grifters (or maybe always were grifters, it's hard to say) many of them could even hold undisclosed positions in the stock market, creating a motive to lie

let's say you discount all the anons, all the twitter/youtube grifters with "AI" sponsors that still leaves plenty of "coding authorities" you can pick from - e.g. the aforementioned OECD, but on a more personal level people like John Carmack and Linus Torvalds have also become "AI" cheerleaders